ardour/ardour
ardour
/
ardour
13
1
Fork 0
Code Pull Requests Releases Activity
ardour/libs/ardour/session.cc

6900 lines
183 KiB
C++
Raw Normal View History

rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/*
Update libardour GPL boilerplate and (C) from git log
2019-08-02 22:01:25 -04:00
* Copyright (C) 1999-2019 Paul Davis <paul@linuxaudiosystems.com>
* Copyright (C) 2006-2007 Jesse Chappell <jesse@essej.net>
* Copyright (C) 2006-2009 Sampo Savolainen <v2@iki.fi>
* Copyright (C) 2006-2015 David Robillard <d@drobilla.net>
* Copyright (C) 2006-2016 Tim Mayberry <mojofunk@gmail.com>
* Copyright (C) 2007-2012 Carl Hetherington <carl@carlh.net>
* Copyright (C) 2008-2009 Hans Baier <hansfbaier@googlemail.com>
* Copyright (C) 2012-2019 Robin Gareus <robin@gareus.org>
* Copyright (C) 2013-2017 Nick Mainsbridge <mainsbridge@gmail.com>
* Copyright (C) 2014-2019 Ben Loftis <ben@harrisonconsoles.com>
* Copyright (C) 2015 GZharun <grygoriiz@wavesglobal.com>
* Copyright (C) 2016-2018 Len Ovens <len@ovenwerks.net>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License along
* with this program; if not, write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
perhaps, just possibly, a working solo model. needs to be fixed so that connections to other JACK clients count as "physical" connections, so don't use this with ardour connected to other JACK apps just yet. Oh, this also invalidates existing a3 sessions again git-svn-id: svn://localhost/ardour2/branches/3.0@7033 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-01 11:09:19 -04:00
#include <stdint.h>
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
#include <algorithm>
#include <string>
#include <vector>
#include <sstream>
#include <cstdio> /* sprintf(3) ... grrr */
#include <cmath>
#include <cerrno>
#include <unistd.h>
#include <limits.h>
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
#include <glibmm/threads.h>
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
#include <glibmm/miscutils.h>
#include <glibmm/fileutils.h>
Add session MIDI bundles to enable connection of MIDI tracks from the right-click I/O menus. git-svn-id: svn://localhost/ardour2/branches/3.0@7517 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-27 20:43:15 -04:00
#include <boost/algorithm/string/erase.hpp>
when merging external source files into session folder, if a collision occurs, use the original file suffix in conjunction with md5 hash
2014-07-08 12:25:36 -04:00
#include "pbd/basename.h"
#include "pbd/convert.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "pbd/error.h"
#include "pbd/file_utils.h"
add finite state machine to control/manage transport state
2019-09-17 20:26:03 -04:00
#include "pbd/i18n.h"
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
#include "pbd/md5.h"
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
#include "pbd/pthread_utils.h"
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
#include "pbd/search_path.h"
when merging external source files into session folder, if a collision occurs, use the original file suffix in conjunction with md5 hash
2014-07-08 12:25:36 -04:00
#include "pbd/stacktrace.h"
#include "pbd/stl_delete.h"
bring in more connection-related changes from Tracks
2015-05-09 19:34:50 -04:00
#include "pbd/replace_all.h"
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
#include "pbd/types_convert.h"
when merging external source files into session folder, if a collision occurs, use the original file suffix in conjunction with md5 hash
2014-07-08 12:25:36 -04:00
#include "pbd/unwind.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
implement set-send-gains-to-{track_level,zero,unity} git-svn-id: svn://localhost/ardour2/branches/3.0@5401 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-21 08:05:44 -04:00
#include "ardour/amp.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/analyser.h"
major redesign of MIDI port heirarchy and management (part 2)
2013-08-07 22:22:11 -04:00
#include "ardour/async_midi_port.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/audio_buffer.h"
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
#include "ardour/audio_port.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/audio_track.h"
#include "ardour/audioengine.h"
#include "ardour/audiofilesource.h"
#include "ardour/auditioner.h"
make boost shared_ptr debugging a little easier to turn on/off
2016-05-07 13:32:31 -04:00
#include "ardour/boost_debug.h"
substantive change: use the JACK wait API and provide "thread buffers" separately from session in preparation for parallelization. lots of debug output at present. If using JACK1, requires a very current version of JACK1 SVN (0.119.0) git-svn-id: svn://localhost/ardour2/branches/3.0@6888 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-13 16:48:33 -04:00
#include "ardour/buffer_manager.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/buffer_set.h"
#include "ardour/bundle.h"
Split butler into separate object (partially, just data so far...) git-svn-id: svn://localhost/ardour2/branches/3.0@5900 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-23 19:23:00 -04:00
#include "ardour/butler.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/click.h"
Fix restoration of connections to control surface ports. git-svn-id: svn://localhost/ardour2/branches/3.0@9475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-05-04 11:00:43 -04:00
#include "ardour/control_protocol_manager.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/data_type.h"
carl's wondrous DnD VBox patch - processor boxes are now vboxes and not listviews - which included a couple of minor marker/time axis view patches for opaqueness etc. git-svn-id: svn://localhost/ardour2/branches/3.0@6174 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-25 09:37:20 -05:00
#include "ardour/debug.h"
random changes required to get an audio track created and transport functional
2017-04-10 05:25:00 -04:00
#include "ardour/disk_reader.h"
change the semantics of AudioFileSource::peak_path() and Session::peak_path() a little. If given a full path that points outside the current session but is within another ardour session, attempt to use the peakfiles present in that other session.
2015-05-28 19:08:47 -04:00
#include "ardour/directory_names.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/filename_extensions.h"
move Amp::GainControl out into its own source module and out of Amp
2016-01-12 22:04:18 -05:00
#include "ardour/gain_control.h"
Remove over 500 unnecessary includes (including 54 of session.h). It's slightly possible that this causes trivial build failures on different configurations, but otherwise shouldn't cause any problems (i.e. no actual changes other than include/naming/namespace stuff). I deliberately avoided removing libardour-config.h since this can mysteriously break things, though a few of those do seem to be unnecessary. This commit only targets includes of ardour/*.h. There is also a very large number of unnecessary includes of stuff in gtk2_ardour; tackling that should also give a big improvement in build time when things are modified. git-svn-id: svn://localhost/ardour2/branches/3.0@12420 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-24 02:09:29 -04:00
#include "ardour/graph.h"
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
#include "ardour/luabindings.h"
major redesign of MIDI port heirarchy and management (part 2)
2013-08-07 22:22:11 -04:00
#include "ardour/midiport_manager.h"
merge (squash) with scenechange topic branch to provide MIDI-driven scene change markers
2014-04-28 19:58:24 -04:00
#include "ardour/scene_changer.h"
Decouple Session from MidiPatchManager and reduce parsing of midnam xml files The MidiPatchManager only requires a reference to the session to get the path to the Session midnam directory so change it so that the path is passed to MidiPatchManager::add_search_path on Session construction and removed on Session Destruction. This will also make it easier to test and reduce compile times etc. For the common case where the Session doesn't have a Session specific midnam patch files directory(for instance a new session) it won't cause a refresh and reparsing of all the midnam files. This saves about 2 seconds to load a Session on my machine(fast machine with SSD), or about half the time spent in the Session constructor for a new session. There is still going to be that initial cost of parsing the midnam files when the first session is created after starting Ardour. Options to remove that would be to parse the files asynchronously and or use a faster xml parser(eventually), neither of which seem worth doing at this stage. This change will cause a performance regression for the uncommon case where a Session with Session specific midnam files is unloaded and then another Session with Session specific midnam files is loaded as it will cause the common midnam files in midi_patch_path to be parsed twice(unload and load).
2015-10-19 00:52:48 -04:00
#include "ardour/midi_patch_manager.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/midi_track.h"
major design changes: use glib event loop for MIDI thread/UI; rework design of BaseUI and AbstractUI; solo & mute are both temporarily broken; OSC control up next; may segfault during exit git-svn-id: svn://localhost/ardour2/branches/3.0@6328 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-08 22:05:14 -05:00
#include "ardour/midi_ui.h"
Remove over 500 unnecessary includes (including 54 of session.h). It's slightly possible that this causes trivial build failures on different configurations, but otherwise shouldn't cause any problems (i.e. no actual changes other than include/naming/namespace stuff). I deliberately avoided removing libardour-config.h since this can mysteriously break things, though a few of those do seem to be unnecessary. This commit only targets includes of ardour/*.h. There is also a very large number of unnecessary includes of stuff in gtk2_ardour; tackling that should also give a big improvement in build time when things are modified. git-svn-id: svn://localhost/ardour2/branches/3.0@12420 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-24 02:09:29 -04:00
#include "ardour/operations.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/playlist.h"
Fix Duplicate Track with Copy Playlist option selected Previously two new empty playlists were created. Related to and mentioned in issue #7141
2016-11-27 21:30:38 -05:00
#include "ardour/playlist_factory.h"
add a way to create a new MIDI track with an instrument pre-selected git-svn-id: svn://localhost/ardour2/branches/3.0@11421 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-02-01 23:12:23 -05:00
#include "ardour/plugin.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/plugin_insert.h"
Remove over 500 unnecessary includes (including 54 of session.h). It's slightly possible that this causes trivial build failures on different configurations, but otherwise shouldn't cause any problems (i.e. no actual changes other than include/naming/namespace stuff). I deliberately avoided removing libardour-config.h since this can mysteriously break things, though a few of those do seem to be unnecessary. This commit only targets includes of ardour/*.h. There is also a very large number of unnecessary includes of stuff in gtk2_ardour; tackling that should also give a big improvement in build time when things are modified. git-svn-id: svn://localhost/ardour2/branches/3.0@12420 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-24 02:09:29 -04:00
#include "ardour/process_thread.h"
add the Tracks version of Track::monitoring_state() This drastically-stripped down version of the Ardour original is used only when USE_TRACKS_CODE_FEATURES is defined. It doesn't respond to many aspects/features of libardour.
2015-05-08 13:44:24 -04:00
#include "ardour/profile.h"
* libardour uses ARDOUR::nframes_t and ARDOUR::nframes64_t explicitly in headers * use explicit operator<< and operator>> that in turn use PBD::EnumWriter when serializing and deserializing to/from rc files * adds scrolling in mixer window (from 2.X) * BBT math stuff - untested, but basically operational * move LocaleGuard into its own file(s) in libs/pbd * Tempo now uses nframes64_t everywhere (except for sample rate values) * as in 2.X, use mkstemp and hack to avoid temp file nonsense, and remove erroneous free() from disk stats output git-svn-id: svn://localhost/ardour2/branches/3.0@5961 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-28 17:36:40 -04:00
#include "ardour/rc_configuration.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/recent_sessions.h"
Remove over 500 unnecessary includes (including 54 of session.h). It's slightly possible that this causes trivial build failures on different configurations, but otherwise shouldn't cause any problems (i.e. no actual changes other than include/naming/namespace stuff). I deliberately avoided removing libardour-config.h since this can mysteriously break things, though a few of those do seem to be unnecessary. This commit only targets includes of ardour/*.h. There is also a very large number of unnecessary includes of stuff in gtk2_ardour; tackling that should also give a big improvement in build time when things are modified. git-svn-id: svn://localhost/ardour2/branches/3.0@12420 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-24 02:09:29 -04:00
#include "ardour/region.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/region_factory.h"
save version string with session for informational purposes
2016-06-01 16:27:55 -04:00
#include "ardour/revision.h"
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
#include "ardour/route_graph.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/route_group.h"
Parallel execution of cycle-start/end
2017-10-30 10:58:36 -04:00
#include "ardour/rt_tasklist.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/send.h"
libardour now has CoreSelection object to manage selection status of Stripables and AutomationControls
2017-05-05 07:31:21 -04:00
#include "ardour/selection.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/session.h"
#include "ardour/session_directory.h"
Move some (most) playlist-related code and data into a separate object (SessionPlaylists). Main point being to ensure that, when a Session is destroyed (especially as a result of its constructor throwing an exception) SessionPlaylists::track is disconnected from its signal at the same time as the list of playlists is destroyed. Otherwise problems abound. git-svn-id: svn://localhost/ardour2/branches/3.0@6270 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-03 16:52:10 -05:00
#include "ardour/session_playlists.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/smf_source.h"
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
#include "ardour/solo_isolate_control.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/source_factory.h"
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
#include "ardour/speakers.h"
add header to deal with destruction issues for tempo map
2015-04-01 11:56:59 -04:00
#include "ardour/tempo.h"
plug some memory leaks in libardour
2016-04-24 08:41:07 -04:00
#include "ardour/ticker.h"
add finite state machine to control/manage transport state
2019-09-17 20:26:03 -04:00
#include "ardour/transport_fsm.h"
new transport slave/master implementation, libs/ edition
2018-09-18 18:51:59 -04:00
#include "ardour/transport_master.h"
add finite state machine to control/manage transport state
2019-09-17 20:26:03 -04:00
#include "ardour/transport_master_manager.h"
redesign technique for naming/creating regions for MIDI clone (or other non-capture driven MIDI region creation operations). See comments in Session::new_midi_source_name() for details.
2014-04-10 08:58:04 -04:00
#include "ardour/track.h"
Use PBD string conversion functions in PBD::ConfigurationVariable No longer need a specialization for bool as PBD::to_string/string_to already has specializations for bool Remove template specialization for float as string_to/to_string handles string representations of infinity
2016-08-24 08:32:51 -04:00
#include "ardour/types_convert.h"
keep port bundles in sync with engine
2015-03-08 17:40:27 -04:00
#include "ardour/user_bundle.h"
*** NEW CODING POLICY *** All #include statements that include a header that is a part of a library bundled with ardour MUST use quotes, not angle brackets. Do this: #include "ardour/types.h" NOT this: #include <ardour/types.h> Rationale: This is best practice in general, to ensure we include the local version and not the system version. That quotes mean "local" (in some sense) and angle brackets mean "system" (in some sense) is a ubiquitous convention and IIRC right in the C spec somewhere. More pragmatically, this is required by (my) waf (stuff) for dependencies to work correctly. That is: !!! FAILURE TO DO THIS CAN RESULT IN BROKEN BUILDS !!! Failure to comply is punishable by death by torture. :) P.S. It's not that dramatic in all cases, but this (in combination with some GCC flags specific to the include type) is the best way I have found to be absolutely 100% positive the local ones are being used (and we definitely want to be absolutely 100% positive on that one). git-svn-id: svn://localhost/ardour2/branches/3.0@4655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-25 13:26:51 -05:00
#include "ardour/utils.h"
Session HAS-A VCAManager
2016-01-26 00:18:03 -05:00
#include "ardour/vca_manager.h"
clear all solo state should affect VCAs too
2016-04-21 11:08:47 -04:00
#include "ardour/vca.h"
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Remove non-JACK midi++ ports. git-svn-id: svn://localhost/ardour2/branches/3.0@7377 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-05 20:16:36 -04:00
#include "midi++/port.h"
Clean up MMC transmission a bit, and make sure that it is all done from one thread. git-svn-id: svn://localhost/ardour2/branches/3.0@7324 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-29 09:47:53 -04:00
#include "midi++/mmc.h"
editors for control protocols (generalized); editor for Generic MIDI that allows choosing a MIDI binding map (or none); support banking in binding URLs, and other miscellany related to generic MIDI; save+restore JACK_MidiPort connection state (but cause a crash at shutdown time) git-svn-id: svn://localhost/ardour2/branches/3.0@6411 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-29 16:31:14 -05:00
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
#include "LuaBridge/LuaBridge.h"
Fixes case where audiofiles used wrong peakfiles
2015-08-29 07:48:05 -04:00
#include <glibmm/checksum.h>
Remove over 500 unnecessary includes (including 54 of session.h). It's slightly possible that this causes trivial build failures on different configurations, but otherwise shouldn't cause any problems (i.e. no actual changes other than include/naming/namespace stuff). I deliberately avoided removing libardour-config.h since this can mysteriously break things, though a few of those do seem to be unnecessary. This commit only targets includes of ardour/*.h. There is also a very large number of unnecessary includes of stuff in gtk2_ardour; tackling that should also give a big improvement in build time when things are modified. git-svn-id: svn://localhost/ardour2/branches/3.0@12420 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-24 02:09:29 -04:00
namespace ARDOUR {
class MidiSource;
class Processor;
class Speakers;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
using namespace std;
using namespace ARDOUR;
using namespace PBD;
bool Session::_disable_all_loaded_plugins = false;
Prepare option to disable Plugins completely old behavior to only bypassed plugins is being renamed.
2015-09-01 07:26:31 -04:00
bool Session::_bypass_all_loaded_plugins = false;
add a new counter (for sidechain numbering)
2016-04-12 07:49:50 -04:00
guint Session::_name_id_counter = 0;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
use new syntax for connecting to backend signals that enforces explicit connection scope, plus a few other related matters git-svn-id: svn://localhost/ardour2/branches/3.0@6376 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-19 15:26:31 -05:00
PBD::Signal1<void,std::string> Session::Dialog;
PBD::Signal0<int> Session::AskAboutPendingState;
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
PBD::Signal2<int, samplecnt_t, samplecnt_t> Session::AskAboutSampleRateMismatch;
PBD::Signal2<void, samplecnt_t, samplecnt_t> Session::NotifyAboutSampleRateMismatch;
use new syntax for connecting to backend signals that enforces explicit connection scope, plus a few other related matters git-svn-id: svn://localhost/ardour2/branches/3.0@6376 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-19 15:26:31 -05:00
PBD::Signal0<void> Session::SendFeedback;
initial pass at a missing file dialog and "relocatable" source files. lots more to do here git-svn-id: svn://localhost/ardour2/branches/3.0@7983 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 01:03:51 -05:00
PBD::Signal3<int,Session*,std::string,DataType> Session::MissingFile;
use new syntax for connecting to backend signals that enforces explicit connection scope, plus a few other related matters git-svn-id: svn://localhost/ardour2/branches/3.0@6376 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-19 15:26:31 -05:00
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
PBD::Signal1<void, samplepos_t> Session::StartTimeChanged;
PBD::Signal1<void, samplepos_t> Session::EndTimeChanged;
use new syntax for connecting to backend signals that enforces explicit connection scope, plus a few other related matters git-svn-id: svn://localhost/ardour2/branches/3.0@6376 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-19 15:26:31 -05:00
PBD::Signal2<void,std::string, std::string> Session::Exported;
remove a couple of boost::signals2 trouble spots; fix some --strict compile time warnings git-svn-id: svn://localhost/ardour2/branches/3.0@6378 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-20 11:50:41 -05:00
PBD::Signal1<int,boost::shared_ptr<Playlist> > Session::AskAboutPlaylistDeletion;
Respond to JACK session quit request. Fixes #3491. git-svn-id: svn://localhost/ardour2/branches/3.0@7878 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 08:12:16 -04:00
PBD::Signal0<void> Session::Quit;
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
PBD::Signal0<void> Session::FeedbackDetected;
Change the feedback alert to a flashing button; works around part of #4432. git-svn-id: svn://localhost/ardour2/branches/3.0@10539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-11 08:52:27 -05:00
PBD::Signal0<void> Session::SuccessfulGraphSort;
make notice about mismatched session/ardour versions more prominent, since you can guarantee that people are going to do this after release git-svn-id: svn://localhost/ardour2/branches/3.0@14138 d708f5d6-7413-0410-9779-e7cbd77b26cf
2013-03-04 16:57:29 -05:00
PBD::Signal2<void,std::string,std::string> Session::VersionMismatch;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
const samplecnt_t Session::bounce_chunk_size = 8192;
restore rec-enable, solo & mute functionality; remove all Global*Command cruft; reimplement momentary solo&mute; refactor session RT event stuff to be even cleaner and finally compatible with everything-is-a-UI idea ; make all UIs derive from a primitive type that pprovides only void call_slot (sigc::slot<void>) as a way to execute arbitrary code in the UI thread git-svn-id: svn://localhost/ardour2/branches/3.0@6338 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-10 12:45:18 -05:00
static void clean_up_session_event (SessionEvent* ev) { delete ev; }
switch to use boost::function for UI::call_slot operations, to avoid a serious thread safety issue with libsigc++ git-svn-id: svn://localhost/ardour2/branches/3.0@6355 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-11 18:30:48 -05:00
const SessionEvent::RTeventCallback Session::rt_cleanup (clean_up_session_event);
merge in Tracks-derived session-end-is-N-seconds beyond last region end, but define N as zero for Ardour/Mixbus
2015-05-13 10:39:24 -04:00
const uint32_t Session::session_end_shift = 0;
Fix comment. git-svn-id: svn://localhost/ardour2/branches/3.0@12759 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-18 14:28:26 -04:00
/** @param snapshot_name Snapshot name, without .ardour suffix */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Session::Session (AudioEngine &eng,
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
const string& fullpath,
const string& snapshot_name,
make BusProfile argument to new Session constructor be const (and in associated call tree)
2019-10-10 16:34:20 -04:00
BusProfile const * bus_profile,
Remove some wrongly commited stuff. Sorry\! git-svn-id: svn://localhost/ardour2/branches/3.0@11087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-26 19:57:20 -05:00
string mix_template)
Remove global Session::playlists variable, use getter method (1/2) Global variables that can written by anyone are to be avoided. This also simplifies exposing SessionPlaylists as Lua bindings.
2019-03-19 00:14:00 -04:00
: _playlists (new SessionPlaylists)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _engine (eng)
, process_function (&Session::process_with_events)
process effect-automation when bouncing w/processing
2014-05-26 00:17:49 -04:00
, _bounce_processing_active (false)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, waiting_for_sync_offset (false)
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
, _base_sample_rate (0)
, _nominal_sample_rate (0)
, _current_sample_rate (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _record_status (Disabled)
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
, _transport_sample (0)
reinstate "restart seek" behaviour but use a new Session::_seek_counter member variable for this
2017-08-01 11:13:35 -04:00
, _seek_counter (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _session_range_location (0)
Session-range behavior (libardour part)
2019-02-06 10:54:13 -05:00
, _session_range_is_free (true)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _silent (false)
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
, _remaining_latency_preroll (0)
Move vari-speed into backend (resample ports) Previously Ardour used a /local/ per track vari-speed mechanism. Now that the disk-reader is a latency-compensated processor, the speed of each disk-reader would need to be maintained locally, offset by each disk-reader's output latency. Furthermore each disk-reader may produce a different number of samples, depending on its global alignment. This commit introduces port-data resampling directly at the engine-level: Up/down-sample all input ports at the beginning, and down/up-sample output port-data using the inverse ratio at the end of the session's process cycle. The session itself is unaware of the speed-change, and only needs to handle transport speeds {-1, 0, +1}. This also allows for aligned cue-monitoring and vari-speed recording, and also pitch-shifts synthesized MIDI along.
2017-10-29 13:30:18 -04:00
, _engine_speed (1.0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _transport_speed (0)
, _default_transport_speed (1.0)
, _last_transport_speed (0)
Some session cleanup fixes (delete slave, take lock for lua)
2017-07-30 19:48:28 -04:00
, _signalled_varispeed (0)
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, _target_transport_speed (0.0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, auto_play_legal (false)
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
, _requested_return_sample (-1)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, current_block_size (0)
, _worst_output_latency (0)
, _worst_input_latency (0)
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
, _worst_route_latency (0)
, _send_latency_changes (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _have_captured (false)
, _non_soloed_outs_muted (false)
Change last commit to use it's own variable rather than borrowing solo's
2016-01-16 17:40:03 -05:00
, _listening (false)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _listen_cnt (0)
, _solo_isolated_cnt (0)
, _writable (false)
Make sure every imported file is copied to the session path
2013-03-30 14:12:31 -04:00
, _under_nsm_control (false)
add xrun counter API to session
2015-04-28 21:09:17 -04:00
, _xrun_count (0)
new transport slave/master implementation, libs/ edition
2018-09-18 18:51:59 -04:00
, transport_master_tracking_state (Stopped)
, master_wait_end (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, post_export_sync (false)
, post_export_position (0)
, _exporting (false)
, _export_rolling (false)
refactoring to prepare for real-time export * add a threaded TmpFile Writer * update API calls to that process_export_fw() can be used as process_function The idea is to re-use export infrastructure from normalization: export to a tmp-file and then encode target formats after that.
2016-07-13 22:34:18 -04:00
, _realtime_export (false)
fast region export -- don't call process()
2016-10-17 15:34:40 -04:00
, _region_export (false)
pre-process (silence) before export to flush reverb tails etc.
2016-02-29 07:50:56 -05:00
, _export_preroll (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _pre_export_mmc_enabled (false)
, _name (snapshot_name)
, _is_new (true)
, _send_qf_mtc (false)
, _pframes_since_last_mtc (0)
, play_loop (false)
, loop_changing (false)
, last_loopend (0)
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
, _session_dir (new SessionDirectory (fullpath))
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
, _current_snapshot_name (snapshot_name)
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, state_tree (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, state_was_pending (false)
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
, _state_of_the_state (StateOfTheState (CannotSave | InitialConnecting | Loading))
add Session::StateProtector temp. disable save during batch updates, save once at the end.
2014-06-28 15:27:36 -04:00
, _suspend_save (0)
, _save_queued (false)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _last_roll_location (0)
, _last_roll_or_reversal_location (0)
, _last_record_location (0)
, pending_auto_loop (false)
Increase mempool for session-scripts (bindings alone require ~800kB these days)
2017-03-23 14:05:53 -04:00
, _mempool ("Session", 3145728)
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
, lua (lua_newstate (&PBD::ReallocPool::lalloc, &_mempool))
initialize uninitialized variable
2016-02-29 09:17:12 -05:00
, _n_lua_scripts (0)
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, _butler (new Butler (*this))
replace boost::msm - based FSM for transport with one written in "plain C++" Still need to use boost::intrusive to managed qeued/deferred containers
2019-09-20 00:33:43 -04:00
, _transport_fsm (new TransportFSM (*this))
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, _post_transport_work (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _locations (new Locations (*this))
fix initialization order (sigh)
2015-02-18 12:00:28 -05:00
, _ignore_skips_updates (false)
delegate signal emission to dedicated thread. This works around abysmal performance (~.15ms) of boost::function and boost::bind (in PBD::Signal). The overall load is probably higher but the realtime thread remains unaffected.
2015-04-28 16:18:30 -04:00
, _rt_thread_active (false)
, _rt_emit_pending (false)
Session::_ac_thread_active should be used with atomics to create memory barriers, even though lock ordering is likely correct now
2016-12-19 10:33:54 -05:00
, _ac_thread_active (0)
fix initialization order
2016-07-09 17:53:17 -04:00
, _latency_recompute_pending (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, step_speed (0)
, outbound_mtc_timecode_frame (0)
, next_quarter_frame_to_send (-1)
Clarify "frames" (video, timecode) vs "samples" (audio)
2016-12-08 06:26:41 -05:00
, _samples_per_timecode_frame (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _frames_per_hour (0)
, _timecode_frames_per_hour (0)
, last_timecode_valid (false)
, last_timecode_when (0)
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, _send_timecode_update (false)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, ltc_encoder (0)
initialize ltc_enc_buf to 0 to fix crash in ltc_tx_cleanup at startup
2013-05-14 14:01:20 -04:00
, ltc_enc_buf(0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, ltc_buf_off (0)
, ltc_buf_len (0)
, ltc_speed (0)
, ltc_enc_byte (0)
, ltc_enc_pos (0)
, ltc_enc_cnt (0)
, ltc_enc_off (0)
, restarting (false)
, ltc_prev_cycle (0)
, ltc_timecode_offset (0)
, ltc_timecode_negative_offset (false)
, midi_control_ui (0)
, _tempo_map (0)
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, _all_route_group (new RouteGroup (*this, "all"))
, routes (new RouteList)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _adding_routes_in_progress (false)
initial incorporation of Tracks' automatic reconnect features, based on current Tracks code
2015-05-08 20:30:30 -04:00
, _reconnecting_routes_in_progress (false)
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
, _route_deletion_in_progress (false)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, destructive_index (0)
implement track numbering
2014-06-25 15:16:09 -04:00
, _track_number_decimals(1)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, default_fade_steepness (0)
, default_fade_msecs (0)
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, _total_free_4k_blocks (0)
Handle failure of statfs to return available disc space on Samba shares mounted via GVFS, reporting available space as unknown in this case (#4657). git-svn-id: svn://localhost/ardour2/branches/3.0@12674 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-12 12:41:29 -04:00
, _total_free_4k_blocks_uncertain (false)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, no_questions_about_missing_files (false)
, _playback_load (0)
, _capture_load (0)
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, _bundles (new BundleList)
, _bundle_xml_node (0)
Fix undo nesting. git-svn-id: svn://localhost/ardour2/branches/3.0@8538 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:46 -05:00
, _current_trans (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _clicking (false)
Add option for record-only metronome.
2017-04-01 17:52:43 -04:00
, _click_rec_only (false)
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
, click_data (0)
, click_emphasis_data (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, click_length (0)
, click_emphasis_length (0)
, _clicks_cleared (0)
Implement Count-In (before recording), fixed BPM, up to 2 bars
2017-01-17 14:43:55 -05:00
, _count_in_samples (0)
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
, _play_range (false)
initial version of playback priority design. No GUI control over options yet Conflicts: libs/ardour/ardour/session.h libs/ardour/ardour/types.h libs/ardour/enums.cc libs/ardour/session_transport.cc system_config
2015-01-16 12:17:09 -05:00
, _range_selection (-1,-1)
, _object_selection (-1,-1)
Add alternative preroll record mode: * Start recording at preroll, trim region to skip preroll at rec-stop * Keep old punch-in rec-at-preroll API for tape-tracks (later)
2017-01-19 07:03:57 -05:00
, _preroll_record_trim_len (0)
Prepare record with count-in action
2017-02-13 16:59:20 -05:00
, _count_in_once (false)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, main_outs (0)
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
, first_file_data_format_reset (true)
, first_file_header_format_reset (true)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, have_looped (false)
, _have_rec_enabled_track (false)
Tempo ramps - formatting, ensure correct initial lock style.
2016-04-13 10:30:48 -04:00
, _have_rec_disabled_track (true)
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
, _step_editors (0)
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
, _suspend_timecode_transmission (0)
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
, _speakers (new Speakers)
RAII to postpone processor changes/graph recalculation
2016-11-25 12:07:43 -05:00
, _ignore_route_processor_changes (0)
plug some memory leaks in libardour
2016-04-24 08:41:07 -04:00
, midi_clock (0)
merge (squash) with scenechange topic branch to provide MIDI-driven scene change markers
2014-04-28 19:58:24 -04:00
, _scene_changer (0)
fix missing initialization of two Session members, which fixes a crash during quit-without-finished-session
2013-09-16 10:23:37 -04:00
, _midi_ports (0)
, _mmc (0)
first pass at making the GUI for a VCA visible
2016-02-27 22:16:37 -05:00
, _vca_manager (new VCAManager (*this))
libardour now has CoreSelection object to manage selection status of Stripables and AutomationControls
2017-05-05 07:31:21 -04:00
, _selection (new CoreSelection (*this))
when DiskReader checks to see if a locate is in progress, it must be atomic with respect to all DiskReaders
2017-07-05 17:24:02 -04:00
, _global_locate_pending (false)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
save version string with session for informational purposes
2016-06-01 16:27:55 -04:00
created_with = string_compose ("%1 %2", PROGRAM_NAME, revision);
delegate signal emission to dedicated thread. This works around abysmal performance (~.15ms) of boost::function and boost::bind (in PBD::Signal). The overall load is probably higher but the realtime thread remains unaffected.
2015-04-28 16:18:30 -04:00
pthread_mutex_init (&_rt_emit_mutex, 0);
pthread_cond_init (&_rt_emit_cond, 0);
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
pthread_mutex_init (&_auto_connect_mutex, 0);
pthread_cond_init (&_auto_connect_cond, 0);
add a new counter (for sidechain numbering)
2016-04-12 07:49:50 -04:00
init_name_id_counter (1); // reset for new sessions, start at 1
clear all solo state should affect VCAs too
2016-04-21 11:08:47 -04:00
VCA::set_next_vca_number (1); // reset for new sessions, start at 1
add a new counter (for sidechain numbering)
2016-04-12 07:49:50 -04:00
Add backend support to extract session-archives
2016-12-27 16:24:24 -05:00
pre_engine_init (fullpath); // sets _is_new
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
setup_lua ();
remove responsibility for starting AudioEngine from Session
2019-10-11 12:25:08 -04:00
assert (AudioEngine::instance()->running());
immediately_post_engine ();
remove near-duplicate session constructor; change names from control_outs to monitor_out and control_out to monitor_send; may havebroken creation of new sessions in a wierd way (will fix later) git-svn-id: svn://localhost/ardour2/branches/3.0@6783 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-22 17:35:35 -04:00
if (_is_new) {
Tracks specific fix for SR reverting to previous value while creating new session
2015-05-11 18:51:14 -04:00
pre-seed session-state-version. copying/paste of many objects use ::set_state(, version) to set the the state of the copy. Stateful::loading_state_version is used for state-version in those cases. This fails for a New Session (version==0) or when creating a New Session after loading an Ardour2 session.
2015-08-28 08:07:48 -04:00
Stateful::loading_state_version = CURRENT_SESSION_FILE_VERSION;
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
if (create (mix_template, bus_profile)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
destroy ();
provide details about session-creation failure.
2015-09-05 07:16:26 -04:00
throw SessionException (_("Session initialization failed"));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
fix ordering in session construction so that new sessions and existing ones both work
2013-09-16 09:57:22 -04:00
likely fix for creating new sessions from templates
2013-09-21 12:50:45 -04:00
/* if a mix template was provided, then ::create() will
* have copied it into the session and we need to load it
* so that we have the state ready for ::set_state()
* after the engine is started.
*
* Note that we do NOT try to get the sample rate from
* the template at this time, though doing so would
* be easy if we decided this was an appropriate part
* of a template.
*/
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
if (!mix_template.empty()) {
Catch some exceptions during session loading.
2017-08-17 13:28:14 -04:00
try {
if (load_state (_current_snapshot_name)) {
throw SessionException (_("Failed to load template/snapshot state"));
}
} catch (PBD::unknown_enumeration& e) {
throw SessionException (_("Failed to parse template/snapshot state"));
save recent templates analogously to recent sessions
2015-05-08 12:07:33 -04:00
}
store_recent_templates (mix_template);
likely fix for creating new sessions from templates
2013-09-21 12:50:45 -04:00
}
allow to load/save default session-properties
2014-06-29 09:45:08 -04:00
/* load default session properties - if any */
config.load_state();
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
} else {
fix ordering in session construction so that new sessions and existing ones both work
2013-09-16 09:57:22 -04:00
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
if (load_state (_current_snapshot_name)) {
provide details about session-creation failure.
2015-09-05 07:16:26 -04:00
throw SessionException (_("Failed to load state"));
start reordering of Session initialization the goal is to allow a clear demarcation of where we need the audioengine and have existing data parameters (e.g. sample rate) for the session so that we can coerce the engine (if possible) to match to session values. also, to rationalize construction even more and use C++ constructor syntax as much as possible.
2013-09-09 22:40:54 -04:00
}
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
}
Distinguish error-messages. The vast majority of errors reported by users as "Cannot configure audio/midi engine with session parameters" have nothing to do with engine-parameters.
2017-06-27 14:28:45 -04:00
int err = post_engine_init ();
remove responsibility for starting AudioEngine from Session
2019-10-11 12:25:08 -04:00
Distinguish error-messages. The vast majority of errors reported by users as "Cannot configure audio/midi engine with session parameters" have nothing to do with engine-parameters.
2017-06-27 14:28:45 -04:00
if (err) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
destroy ();
Distinguish error-messages. The vast majority of errors reported by users as "Cannot configure audio/midi engine with session parameters" have nothing to do with engine-parameters.
2017-06-27 14:28:45 -04:00
switch (err) {
case -1:
throw SessionException (string_compose (_("Cannot initialize session/engine: %1"), _("Failed to create background threads.")));
break;
case -2:
case -3:
throw SessionException (string_compose (_("Cannot initialize session/engine: %1"), _("Invalid TempoMap in session-file.")));
break;
case -4:
throw SessionException (string_compose (_("Cannot initialize session/engine: %1"), _("Invalid or corrupt session state.")));
break;
case -5:
throw SessionException (string_compose (_("Cannot initialize session/engine: %1"), _("Port registration failed.")));
break;
default:
throw SessionException (string_compose (_("Cannot initialize session/engine: %1"), _("Unexpected exception during session setup, possibly invalid audio/midi engine parameters. Please see stdout/stderr for details")));
break;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
store_recent_sessions (_name, _path);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
bool was_dirty = dirty();
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
unset_dirty ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
PresentationInfo::Change.connect_same_thread (*this, boost::bind (&Session::notify_presentation_info_change, this));
fully implement and deploy explicit x-thread signal connection syntax (testing comes next) git-svn-id: svn://localhost/ardour2/branches/3.0@6379 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-21 13:23:07 -05:00
Config->ParameterChanged.connect_same_thread (*this, boost::bind (&Session::config_changed, this, _1, false));
config.ParameterChanged.connect_same_thread (*this, boost::bind (&Session::config_changed, this, _1, true));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (was_dirty) {
DirtyChanged (); /* EMIT SIGNAL */
}
don't set _is_new=false too early when creating a new session git-svn-id: svn://localhost/ardour2/branches/3.0@6788 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-23 13:58:40 -04:00
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
StartTimeChanged.connect_same_thread (*this, boost::bind (&Session::start_time_changed, this, _1));
EndTimeChanged.connect_same_thread (*this, boost::bind (&Session::end_time_changed, this, _1));
Add abstract API for latency compensated sends This is in preparation for MixbusSends that are not derived from Delivery : IOProcessor.
2019-09-20 14:25:04 -04:00
LatentSend::ChangedLatency.connect_same_thread (*this, boost::bind (&Session::send_latency_compensation_change, this));
Add Signal to indicate global delay compensation status
2019-09-28 14:53:50 -04:00
Latent::DisableSwitchChanged.connect_same_thread (*this, boost::bind (&Session::queue_latency_recompute, this));
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
delegate signal emission to dedicated thread. This works around abysmal performance (~.15ms) of boost::function and boost::bind (in PBD::Signal). The overall load is probably higher but the realtime thread remains unaffected.
2015-04-28 16:18:30 -04:00
emit_thread_start ();
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
auto_connect_thread_start ();
delegate signal emission to dedicated thread. This works around abysmal performance (~.15ms) of boost::function and boost::bind (in PBD::Signal). The overall load is probably higher but the realtime thread remains unaffected.
2015-04-28 16:18:30 -04:00
*maybe* fix issues with zombification during session loading
2013-09-11 16:33:40 -04:00
/* hook us up to the engine since we are now completely constructed */
BootMessage (_("Connect to engine"));
_engine.set_session (this);
_engine.reset_timebase ();
Add backend support to extract session-archives
2016-12-27 16:24:24 -05:00
ensure_subdirs (); // archived or zipped sessions may lack peaks/ analysis/ etc
Save sample-rate for sessions created from templates
2017-08-09 20:21:52 -04:00
if (!mix_template.empty ()) {
/* ::create() unsets _is_new after creating the session.
* But for templated sessions, the sample-rate is initially unset
* (not read from template), so we need to save it (again).
*/
_is_new = true;
}
add in more Tracks-related auto-(re)connect changes
2015-05-08 20:56:10 -04:00
session_loaded ();
Save sample-rate for sessions created from templates
2017-08-09 20:21:52 -04:00
_is_new = false;
Remove _midi_regions_use_bbt_beats from Session, _start_pulse and _length_pulse from MidiRegion. - _start/length_beats are now quarter notes regardless of loaded session version. - also restores note colour update
2016-09-22 13:39:05 -04:00
*maybe* fix issues with zombification during session loading
2013-09-11 16:33:40 -04:00
BootMessage (_("Session loading complete"));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Session::~Session ()
{
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
#ifdef PT_TIMING
Add some debug code to time things in the process thread(s). git-svn-id: svn://localhost/ardour2/branches/3.0@11246 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-14 17:02:59 -05:00
ST.dump ("ST.dump");
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
#endif
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
destroy ();
}
add a new counter (for sidechain numbering)
2016-04-12 07:49:50 -04:00
unsigned int
Session::next_name_id ()
{
return g_atomic_int_add (&_name_id_counter, 1);
}
unsigned int
Session::name_id_counter ()
{
return g_atomic_int_get (&_name_id_counter);
}
void
Session::init_name_id_counter (guint n)
{
g_atomic_int_set (&_name_id_counter, n);
}
more session construction order changes, and removal of n_physical_{inputs,outputs} members which were (a) not initialized early enough (b) not used anywhere except monitor bus connection. Things almost make sense now.
2013-09-16 12:08:19 -04:00
int
Session::immediately_post_engine ()
{
/* Do various initializations that should take place directly after we
* know that the engine is running, but before we either create a
* session or set state for an existing one.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Parallel execution of cycle-start/end
2017-10-30 10:58:36 -04:00
_rt_tasklist.reset (new RTTaskList ());
fix ordering in session construction so that new sessions and existing ones both work
2013-09-16 09:57:22 -04:00
if (how_many_dsp_threads () > 1) {
/* For now, only create the graph if we are using >1 DSP threads, as
it is a bit slower than the old code with 1 thread.
*/
_process_graph.reset (new Graph (*this));
}
reinstate missing latency initialization code Removed (mistakenly) during work on transportFSM
2019-12-16 18:13:27 -05:00
/* every time we reconnect, recompute worst case output latencies */
_engine.Running.connect_same_thread (*this, boost::bind (&Session::initialize_latencies, this));
add finite state machine to control/manage transport state
2019-09-17 20:26:03 -04:00
/* Restart transport FSM */
more session construction order changes, and removal of n_physical_{inputs,outputs} members which were (a) not initialized early enough (b) not used anywhere except monitor bus connection. Things almost make sense now.
2013-09-16 12:08:19 -04:00
replace boost::msm - based FSM for transport with one written in "plain C++" Still need to use boost::intrusive to managed qeued/deferred containers
2019-09-20 00:33:43 -04:00
_transport_fsm->start ();
add finite state machine to control/manage transport state
2019-09-17 20:26:03 -04:00
/* every time we reconnect, do stuff ... */
_engine.Running.connect_same_thread (*this, boost::bind (&Session::engine_running, this));
more session construction order changes, and removal of n_physical_{inputs,outputs} members which were (a) not initialized early enough (b) not used anywhere except monitor bus connection. Things almost make sense now.
2013-09-16 12:08:19 -04:00
try {
BootMessage (_("Set up LTC"));
setup_ltc ();
BootMessage (_("Set up Click"));
setup_click ();
BootMessage (_("Set up standard connections"));
setup_bundles ();
}
catch (failed_constructor& err) {
return -1;
}
keep port bundles in sync with engine
2015-03-08 17:40:27 -04:00
/* TODO, connect in different thread. (PortRegisteredOrUnregistered may be in RT context)
* can we do that? */
_engine.PortRegisteredOrUnregistered.connect_same_thread (*this, boost::bind (&Session::setup_bundles, this));
nicely named session bundles
2015-03-08 13:39:19 -04:00
remove responsibility for starting AudioEngine from Session
2019-10-11 12:25:08 -04:00
// set samplerate for plugins added early
// e.g from templates or MB channelstrip
set_block_size (_engine.samples_per_cycle());
set_sample_rate (_engine.sample_rate());
don't stop metering thread when session is removed; move engine-setup code into its own method. sorry, ardour build-from-source folk :)
2013-09-12 17:06:21 -04:00
return 0;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::destroy ()
{
carl's wondrous DnD VBox patch - processor boxes are now vboxes and not listviews - which included a couple of minor marker/time axis view patches for opaqueness etc. git-svn-id: svn://localhost/ardour2/branches/3.0@6174 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-25 09:37:20 -05:00
vector<void*> debug_pointers;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* if we got to here, leaving pending capture state around
is a mistake.
*/
remove_pending_capture_state ();
Prevent crash at session-close when analysing. The Analyser only holds a weak-pointer the the Source, session-destruction frees the actual Source, which is fatal for any ongoing audio analysis. This fix simply waits for the current ongoing analysis to complete, ideally TransientDetector::run, EBUr128Analysis::run, OnsetDetector::run and AudioAnalyser::analyse could be interrupted. Alternate option: cancel the Analyser::work thread (and re-create with the every session).
2015-11-24 18:55:41 -05:00
Analyser::flush ();
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
_state_of_the_state = StateOfTheState (CannotSave | Deletion);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Some session cleanup fixes (delete slave, take lock for lua)
2017-07-30 19:48:28 -04:00
{
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
ltc_tx_cleanup();
}
Session should disconnect from all signals during its destruction, early git-svn-id: svn://localhost/ardour2/branches/3.0@13601 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-05 11:07:30 -05:00
/* disconnect from any and all signals that we are connected to */
further tweaks to session close
2016-04-23 20:51:23 -04:00
Port::PortSignalDrop (); /* EMIT SIGNAL */
Session should disconnect from all signals during its destruction, early git-svn-id: svn://localhost/ardour2/branches/3.0@13601 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-05 11:07:30 -05:00
drop_connections ();
explicitly drop control surface protocols before disconnecting from engine. This is not bomb/thread proof yet, because it still requires at least one process callback to function
2015-10-02 19:22:16 -04:00
/* shutdown control surface protocols while we still have ports
improve 1a288ddd, destroy IO after disconnecting from signals. This does not completely fix the race-condition. The GUI (e.g. connection-manager) may still hold a shared-ptr reference.
2017-08-03 12:52:39 -04:00
* and the engine to move data to any devices.
*/
explicitly drop control surface protocols before disconnecting from engine. This is not bomb/thread proof yet, because it still requires at least one process callback to function
2015-10-02 19:22:16 -04:00
ControlProtocolManager::instance().drop_protocols ();
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Unregister ports at session close. This fixes issues when re-loading a session without taking the engine down.
2017-04-01 07:14:17 -04:00
/* stop auto dis/connecting */
move termination of auto connect thread till after control protocols are destroyed. Otherwise, control protocol ports don't actually get deleted
2016-12-17 19:06:44 -05:00
auto_connect_thread_terminate ();
Decouple Session from MidiPatchManager and reduce parsing of midnam xml files The MidiPatchManager only requires a reference to the session to get the path to the Session midnam directory so change it so that the path is passed to MidiPatchManager::add_search_path on Session construction and removed on Session Destruction. This will also make it easier to test and reduce compile times etc. For the common case where the Session doesn't have a Session specific midnam patch files directory(for instance a new session) it won't cause a refresh and reparsing of all the midnam files. This saves about 2 seconds to load a Session on my machine(fast machine with SSD), or about half the time spent in the Session constructor for a new session. There is still going to be that initial cost of parsing the midnam files when the first session is created after starting Ardour. Options to remove that would be to parse the files asynchronously and or use a faster xml parser(eventually), neither of which seem worth doing at this stage. This change will cause a performance regression for the uncommon case where a Session with Session specific midnam files is unloaded and then another Session with Session specific midnam files is loaded as it will cause the common midnam files in midi_patch_path to be parsed twice(unload and load).
2015-10-19 00:52:48 -04:00
MIDI::Name::MidiPatchManager::instance().remove_search_path(session_directory().midi_patch_path());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
_engine.remove_session ();
a putatively better approach to cleaning up ports at session closing git-svn-id: svn://localhost/ardour2/branches/3.0@12807 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-21 11:00:10 -04:00
/* deregister all ports - there will be no process or any other
* callbacks from the engine any more.
*/
Port::PortDrop (); /* EMIT SIGNAL */
move reset (deletion) of click_io and ltc_output objects until after we are disconnected from the engine We use those objects unconditionally and without caching inside process(), which could be invoked by the engine during their deletion
2019-10-28 13:15:07 -04:00
/* remove I/O objects that we (the session) own */
_click_io.reset ();
_ltc_output.reset ();
Clean out session-global controllables This isn't strictly speaking needed, there are only a handful of users (most notably generic-midi ctrl surface, and Selection)
2019-03-23 10:52:36 -04:00
{
Glib::Threads::Mutex::Lock lm (controllables_lock);
for (Controllables::iterator i = controllables.begin(); i != controllables.end(); ++i) {
(*i)->DropReferences (); /* EMIT SIGNAL */
}
controllables.clear ();
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* clear history so that no references to objects are held any more */
_history.clear ();
/* clear state tree so that no references to objects are held any more */
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
most of the 2.X->3.0 commit (up to rev 4299) except for gtk2_ardour/editor_canvas.cc; builds and runs and does a few specific things but expect it to be buggy for a while yet git-svn-id: svn://localhost/ardour2/branches/3.0@4313 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-12-12 09:43:24 -05:00
delete state_tree;
fix a bunch of memory leaks
2014-09-10 20:34:09 -04:00
state_tree = 0;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Some session cleanup fixes (delete slave, take lock for lua)
2017-07-30 19:48:28 -04:00
{
/* unregister all lua functions, drop held references (if any) */
Glib::Threads::Mutex::Lock tm (lua_lock, Glib::Threads::TRY_LOCK);
(*_lua_cleanup)();
lua.do_command ("Session = nil");
delete _lua_run;
delete _lua_add;
delete _lua_del;
delete _lua_list;
delete _lua_save;
delete _lua_load;
delete _lua_cleanup;
lua.collect_garbage ();
}
Split butler into separate object (partially, just data so far...) git-svn-id: svn://localhost/ardour2/branches/3.0@5900 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-23 19:23:00 -04:00
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
/* reset dynamic state version back to default */
rework Stateful::set_state() patch to avoid default version argument git-svn-id: svn://localhost/ardour2/branches/3.0@5787 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-15 14:56:11 -04:00
Stateful::loading_state_version = 0;
Prevent crash in pool destruction during session teardown. git-svn-id: svn://localhost/ardour2/branches/3.0@6905 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-14 19:58:20 -04:00
_butler->drop_references ();
clean up editor mixer strip when session is going away; slightly improved boost debugging git-svn-id: svn://localhost/ardour2/branches/3.0@6397 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-23 23:04:01 -05:00
delete _butler;
Hopefully fix race between engine being halted and butler being destroyed. git-svn-id: svn://localhost/ardour2/branches/3.0@12722 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-14 07:45:10 -04:00
_butler = 0;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
delete _all_route_group;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
delete route-groups before flushing routes.
2014-09-10 14:43:30 -04:00
DEBUG_TRACE (DEBUG::Destruction, "delete route groups\n");
for (list<RouteGroup *>::iterator i = _route_groups.begin(); i != _route_groups.end(); ++i) {
delete *i;
}
most of the 2.X->3.0 commit (up to rev 4299) except for gtk2_ardour/editor_canvas.cc; builds and runs and does a few specific things but expect it to be buggy for a while yet git-svn-id: svn://localhost/ardour2/branches/3.0@4313 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-12-12 09:43:24 -05:00
if (click_data != default_click) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
delete [] click_data;
}
most of the 2.X->3.0 commit (up to rev 4299) except for gtk2_ardour/editor_canvas.cc; builds and runs and does a few specific things but expect it to be buggy for a while yet git-svn-id: svn://localhost/ardour2/branches/3.0@4313 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-12-12 09:43:24 -05:00
if (click_emphasis_data != default_click_emphasis) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
delete [] click_emphasis_data;
}
clear_clicks ();
clean up after monitoring-section on session close
2014-06-29 22:04:35 -04:00
/* need to remove auditioner before monitoring section
Fix crash when unloading a partially loaded session.
2017-09-21 11:35:52 -04:00
* otherwise it is re-connected.
* Note: If a session was never successfully loaded, there
* may not yet be an auditioner.
*/
if (auditioner) {
auditioner->drop_references ();
}
clean up after monitoring-section on session close
2014-06-29 22:04:35 -04:00
auditioner.reset ();
/* drop references to routes held by the monitoring section
* specifically _monitor_out aux/listen references */
remove_monitor_section();
switch Route::fed_by from shared_ptr to weak_ptr and thus begin to allow Route destructors to be called git-svn-id: svn://localhost/ardour2/branches/3.0@6189 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 12:29:27 -05:00
/* clear out any pending dead wood from RCU managed objects */
routes.flush ();
_bundles.flush ();
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
random changes required to get an audio track created and transport functional
2017-04-10 05:25:00 -04:00
DiskReader::free_working_buffers();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
cleanup up cleanup at session destruction; clarify the meaning of 3 signals (DropReferences & Destroyed in libardour ; CatchDeletion in the GTK UI); clarify ownership of objects (session no longer pays attention to DropReferences for objects that it is considered to own, such as routes, sources, etc); fix up MIDI parsing and a couple of other places by correcting syntax for return of values from a boost::signals2::signal (possible danger elsewhere to be checked) git-svn-id: svn://localhost/ardour2/branches/3.0@6389 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-22 15:21:43 -05:00
/* tell everyone who is still standing that we're about to die */
drop_references ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
cleanup up cleanup at session destruction; clarify the meaning of 3 signals (DropReferences & Destroyed in libardour ; CatchDeletion in the GTK UI); clarify ownership of objects (session no longer pays attention to DropReferences for objects that it is considered to own, such as routes, sources, etc); fix up MIDI parsing and a couple of other places by correcting syntax for return of values from a boost::signals2::signal (possible danger elsewhere to be checked) git-svn-id: svn://localhost/ardour2/branches/3.0@6389 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-22 15:21:43 -05:00
/* tell everyone to drop references and delete objects as we go */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
carl's wondrous DnD VBox patch - processor boxes are now vboxes and not listviews - which included a couple of minor marker/time axis view patches for opaqueness etc. git-svn-id: svn://localhost/ardour2/branches/3.0@6174 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-25 09:37:20 -05:00
DEBUG_TRACE (DEBUG::Destruction, "delete regions\n");
remove the session region list; GUI now represents (a relatively unfiltered view of) the raw region list that always contains every single region ever created git-svn-id: svn://localhost/ardour2/branches/3.0@6739 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-06 10:40:42 -05:00
RegionFactory::delete_all_regions ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
reorder destruction of VCAManager and Routes
2016-02-29 16:10:31 -05:00
/* Do this early so that VCAs no longer hold references to routes */
DEBUG_TRACE (DEBUG::Destruction, "delete vcas\n");
delete _vca_manager;
carl's wondrous DnD VBox patch - processor boxes are now vboxes and not listviews - which included a couple of minor marker/time axis view patches for opaqueness etc. git-svn-id: svn://localhost/ardour2/branches/3.0@6174 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-25 09:37:20 -05:00
DEBUG_TRACE (DEBUG::Destruction, "delete routes\n");
Fix track lifetime wrt graph. Fixes #3251. git-svn-id: svn://localhost/ardour2/branches/3.0@7410 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-13 22:23:37 -04:00
tracing and small fixes to improve object destruction pathways git-svn-id: svn://localhost/ardour2/branches/3.0@6195 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 19:49:04 -05:00
/* reset these three references to special routes before we do the usual route delete thing */
_master_out.reset ();
remove near-duplicate session constructor; change names from control_outs to monitor_out and control_out to monitor_send; may havebroken creation of new sessions in a wierd way (will fix later) git-svn-id: svn://localhost/ardour2/branches/3.0@6783 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-22 17:35:35 -04:00
_monitor_out.reset ();
tracing and small fixes to improve object destruction pathways git-svn-id: svn://localhost/ardour2/branches/3.0@6195 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 19:49:04 -05:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
RCUWriter<RouteList> writer (routes);
boost::shared_ptr<RouteList> r = writer.get_copy ();
cleanup up cleanup at session destruction; clarify the meaning of 3 signals (DropReferences & Destroyed in libardour ; CatchDeletion in the GTK UI); clarify ownership of objects (session no longer pays attention to DropReferences for objects that it is considered to own, such as routes, sources, etc); fix up MIDI parsing and a couple of other places by correcting syntax for return of values from a boost::signals2::signal (possible danger elsewhere to be checked) git-svn-id: svn://localhost/ardour2/branches/3.0@6389 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-22 15:21:43 -05:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
carl's wondrous DnD VBox patch - processor boxes are now vboxes and not listviews - which included a couple of minor marker/time axis view patches for opaqueness etc. git-svn-id: svn://localhost/ardour2/branches/3.0@6174 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-25 09:37:20 -05:00
DEBUG_TRACE(DEBUG::Destruction, string_compose ("Dropping for route %1 ; pre-ref = %2\n", (*i)->name(), (*i).use_count()));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
(*i)->drop_references ();
}
cleanup up cleanup at session destruction; clarify the meaning of 3 signals (DropReferences & Destroyed in libardour ; CatchDeletion in the GTK UI); clarify ownership of objects (session no longer pays attention to DropReferences for objects that it is considered to own, such as routes, sources, etc); fix up MIDI parsing and a couple of other places by correcting syntax for return of values from a boost::signals2::signal (possible danger elsewhere to be checked) git-svn-id: svn://localhost/ardour2/branches/3.0@6389 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-22 15:21:43 -05:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
r->clear ();
/* writer goes out of scope and updates master */
}
routes.flush ();
rationalize destruction pathway (some more); tidy-ify some ImageFrame code git-svn-id: svn://localhost/ardour2/branches/3.0@6398 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-25 16:06:52 -05:00
lock source list when destroying it.
2014-01-19 16:37:01 -05:00
{
DEBUG_TRACE (DEBUG::Destruction, "delete sources\n");
Glib::Threads::Mutex::Lock lm (source_lock);
for (SourceMap::iterator i = sources.begin(); i != sources.end(); ++i) {
DEBUG_TRACE(DEBUG::Destruction, string_compose ("Dropping for source %1 ; pre-ref = %2\n", i->second->name(), i->second.use_count()));
i->second->drop_references ();
}
tracing and small fixes to improve object destruction pathways git-svn-id: svn://localhost/ardour2/branches/3.0@6195 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 19:49:04 -05:00
lock source list when destroying it.
2014-01-19 16:37:01 -05:00
sources.clear ();
}
tracing and small fixes to improve object destruction pathways git-svn-id: svn://localhost/ardour2/branches/3.0@6195 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 19:49:04 -05:00
Use a shared_ptr for SessionPlaylists so that it can be explicitly destroyed in ~Session to ease debugging. git-svn-id: svn://localhost/ardour2/branches/3.0@6282 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-04 14:09:08 -05:00
/* not strictly necessary, but doing it here allows the shared_ptr debugging to work */
Remove global Session::playlists variable, use getter method (1/2) Global variables that can written by anyone are to be avoided. This also simplifies exposing SessionPlaylists as Lua bindings.
2019-03-19 00:14:00 -04:00
_playlists.reset ();
Use a shared_ptr for SessionPlaylists so that it can be explicitly destroyed in ~Session to ease debugging. git-svn-id: svn://localhost/ardour2/branches/3.0@6282 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-04 14:09:08 -05:00
delegate signal emission to dedicated thread. This works around abysmal performance (~.15ms) of boost::function and boost::bind (in PBD::Signal). The overall load is probably higher but the realtime thread remains unaffected.
2015-04-28 16:18:30 -04:00
emit_thread_terminate ();
pthread_cond_destroy (&_rt_emit_cond);
pthread_mutex_destroy (&_rt_emit_mutex);
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
pthread_cond_destroy (&_auto_connect_cond);
pthread_mutex_destroy (&_auto_connect_mutex);
merge (squash) with scenechange topic branch to provide MIDI-driven scene change markers
2014-04-28 19:58:24 -04:00
delete _scene_changer; _scene_changer = 0;
fix crash at session close/exit if a midi-control-surface is used The "real" problem was that MidiControlUI is destroyed while there are still MIDIControllables around that still have a signal connection to a Controllable: controllable->Destroyed(..., MidiControlUI::instance()) If a Contrallable is deleted after the MidiControlUI event loop is gone, the Destroyed() signal can create odd situations...
2014-09-11 13:29:55 -04:00
delete midi_control_ui; midi_control_ui = 0;
merge (squash) with scenechange topic branch to provide MIDI-driven scene change markers
2014-04-28 19:58:24 -04:00
fix missing initialization of two Session members, which fixes a crash during quit-without-finished-session
2013-09-16 10:23:37 -04:00
delete _mmc; _mmc = 0;
delete _midi_ports; _midi_ports = 0;
delete _locations; _locations = 0;
Allow markers to be glued to bar/beat time. Fixes #1815. git-svn-id: svn://localhost/ardour2/branches/3.0@7573 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-09 12:40:31 -04:00
plug some memory leaks in libardour
2016-04-24 08:41:07 -04:00
delete midi_clock;
delete tempo map at appropriate times
2015-04-01 10:58:56 -04:00
delete _tempo_map;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
flush pending Session Events at session close This prevents a memory leak: The session is closed. The session-butler drops memory pool trash. The Engine keeps running. Once the AudioEngine is taken down (sample-rate switch, or at exit), the backend process-thread is terminated but there is no trash-can anymore. If there are unprocessed SessionEvents, this calls free_per_thread_pool() -> cp->parent()->add_to_trash (cp). "parent()" in this case the trash-can the butler emptied.
2016-04-28 15:37:18 -04:00
/* clear event queue, the session is gone, nobody is interested in
* those anymore, but they do leak memory if not removed
*/
while (!immediate_events.empty ()) {
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
SessionEvent *ev = immediate_events.front ();
DEBUG_TRACE (DEBUG::SessionEvents, string_compose ("Drop event: %1\n", enum_2_string (ev->type)));
immediate_events.pop_front ();
bool remove = true;
bool del = true;
switch (ev->type) {
case SessionEvent::AutoLoop:
case SessionEvent::Skip:
case SessionEvent::PunchIn:
case SessionEvent::PunchOut:
case SessionEvent::RangeStop:
case SessionEvent::RangeLocate:
remove = false;
del = false;
break;
case SessionEvent::RealTimeOperation:
process_rtop (ev);
del = false;
default:
break;
}
if (remove) {
del = del && !_remove_event (ev);
}
if (del) {
delete ev;
}
}
Unregister ports at session close. This fixes issues when re-loading a session without taking the engine down.
2017-04-01 07:14:17 -04:00
{
/* unregister all dropped ports, process pending port deletion. */
// this may call ARDOUR::Port::drop ... jack_port_unregister ()
// jack1 cannot cope with removing ports while processing
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
AudioEngine::instance()->clear_pending_port_deletions ();
}
Delete CoreSelection at session close
2017-07-28 14:11:55 -04:00
DEBUG_TRACE (DEBUG::Destruction, "delete selection\n");
delete _selection;
_selection = 0;
replace boost::msm - based FSM for transport with one written in "plain C++" Still need to use boost::intrusive to managed qeued/deferred containers
2019-09-20 00:33:43 -04:00
_transport_fsm->stop ();
add finite state machine to control/manage transport state
2019-09-17 20:26:03 -04:00
tracing and small fixes to improve object destruction pathways git-svn-id: svn://localhost/ardour2/branches/3.0@6195 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 19:49:04 -05:00
DEBUG_TRACE (DEBUG::Destruction, "Session::destroy() done\n");
break circular reference between panners and pannables; more destruction debug tracing; a little more boost debug stuff git-svn-id: svn://localhost/ardour2/branches/3.0@8822 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-10 13:13:15 -05:00
Clean out session-global controllables This isn't strictly speaking needed, there are only a handful of users (most notably generic-midi ctrl surface, and Selection)
2019-03-23 10:52:36 -04:00
#ifndef NDEBUG
Controllable::dump_registry ();
#endif
make boost shared_ptr debugging a little easier to turn on/off
2016-05-07 13:32:31 -04:00
BOOST_SHOW_POINTERS ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
void
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
Session::setup_ltc ()
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
XMLNode* child = 0;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make fixed IO ports untranslatable Outdated and fuzzy translations of MTC, MTC, LTC, etc caused various issues (duplicate jack port names, exceptions, crashes). Functionality should not be affected by translations (for now).
2013-10-23 12:02:49 -04:00
_ltc_output.reset (new IO (*this, X_("LTC Out"), IO::Output));
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make fixed IO ports untranslatable Outdated and fuzzy translations of MTC, MTC, LTC, etc caused various issues (duplicate jack port names, exceptions, crashes). Functionality should not be affected by translations (for now).
2013-10-23 12:02:49 -04:00
if (state_tree && (child = find_named_node (*state_tree->root(), X_("LTC Out"))) != 0) {
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
_ltc_output->set_state (*(child->children().front()), Stateful::loading_state_version);
} else {
{
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
_ltc_output->ensure_io (ChanCount (DataType::AUDIO, 1), true, this);
Fix potential deadlock on session-load
2016-12-11 18:09:24 -05:00
// TODO use auto-connect thread
reconnect_ltc_output ();
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
/* fix up names of LTC ports because we don't want the normal
* IO style of NAME/TYPE-{in,out}N
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make fixed IO ports untranslatable Outdated and fuzzy translations of MTC, MTC, LTC, etc caused various issues (duplicate jack port names, exceptions, crashes). Functionality should not be affected by translations (for now).
2013-10-23 12:02:49 -04:00
_ltc_output->nth (0)->set_name (X_("LTC-out"));
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
void
Session::setup_click ()
{
_clicking = false;
change ownership of the AutomationControl used by Amp. It used to be owned by Amp. Now it is owned by Amp's owner
2016-01-12 14:09:24 -05:00
boost::shared_ptr<AutomationList> gl (new AutomationList (Evoral::Parameter (GainAutomation)));
change exposed type of various objects' gain controls; remove Amp::gain() as a shortcut
2016-01-25 11:47:21 -05:00
boost::shared_ptr<GainControl> gain_control = boost::shared_ptr<GainControl> (new GainControl (*this, Evoral::Parameter(GainAutomation), gl));
change ownership of the AutomationControl used by Amp. It used to be owned by Amp. Now it is owned by Amp's owner
2016-01-12 14:09:24 -05:00
make fixed IO ports untranslatable Outdated and fuzzy translations of MTC, MTC, LTC, etc caused various issues (duplicate jack port names, exceptions, crashes). Functionality should not be affected by translations (for now).
2013-10-23 12:02:49 -04:00
_click_io.reset (new ClickIO (*this, X_("Click")));
change ownership of the AutomationControl used by Amp. It used to be owned by Amp. Now it is owned by Amp's owner
2016-01-12 14:09:24 -05:00
_click_gain.reset (new Amp (*this, _("Fader"), gain_control, true));
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
_click_gain->activate ();
fix up crash caused by changes in object creation order and setting the state of the click IO object
2013-10-02 16:17:22 -04:00
if (state_tree) {
fix missing Click outputs for new sessions
2013-10-18 15:20:19 -04:00
setup_click_state (state_tree->root());
} else {
setup_click_state (0);
fix up crash caused by changes in object creation order and setting the state of the click IO object
2013-10-02 16:17:22 -04:00
}
}
void
fix missing Click outputs for new sessions
2013-10-18 15:20:19 -04:00
Session::setup_click_state (const XMLNode* node)
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
{
fix up crash caused by changes in object creation order and setting the state of the click IO object
2013-10-02 16:17:22 -04:00
const XMLNode* child = 0;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
fix missing Click outputs for new sessions
2013-10-18 15:20:19 -04:00
if (node && (child = find_named_node (*node, "Click")) != 0) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
/* existing state for Click */
int c = 0;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
if (Stateful::loading_state_version < 3000) {
c = _click_io->set_state_2X (*child->children().front(), Stateful::loading_state_version, false);
move ownership of LTC I/O ports to Session, and manage as IO objects git-svn-id: svn://localhost/ardour2/branches/3.0@13341 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-10-25 15:46:23 -04:00
} else {
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
const XMLNodeList& children (child->children());
XMLNodeList::const_iterator i = children.begin();
if ((c = _click_io->set_state (**i, Stateful::loading_state_version)) == 0) {
++i;
if (i != children.end()) {
c = _click_gain->set_state (**i, Stateful::loading_state_version);
restore independent gain control over click/metronome git-svn-id: svn://localhost/ardour2/branches/3.0@11370 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-27 17:47:16 -05:00
}
more set_state() tweaks to get a 2.X session to load without aborting git-svn-id: svn://localhost/ardour2/branches/3.0@5788 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-15 15:36:48 -04:00
}
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
if (c == 0) {
_clicking = Config->get_clicking ();
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
} else {
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
error << _("could not setup Click I/O") << endmsg;
_clicking = false;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
more set_state() tweaks to get a 2.X session to load without aborting git-svn-id: svn://localhost/ardour2/branches/3.0@5788 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-15 15:36:48 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
} else {
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
/* default state for Click: dual-mono to first 2 physical outputs */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
vector<string> outs;
_engine.get_physical_outputs (DataType::AUDIO, outs);
Trim get_nth_physical port methods. git-svn-id: svn://localhost/ardour2/branches/3.0@7518 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-27 21:40:36 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
for (uint32_t physport = 0; physport < 2; ++physport) {
if (outs.size() > physport) {
if (_click_io->add_port (outs[physport], this)) {
// relax, even though its an error
Trim get_nth_physical port methods. git-svn-id: svn://localhost/ardour2/branches/3.0@7518 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-27 21:40:36 -04:00
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
if (_click_io->n_ports () > ChanCount::ZERO) {
_clicking = Config->get_clicking ();
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
add Session::get_physical_ports() to centralize "get physical ports but (maybe) ignore control-only (and MIDI Through) ports"
2016-07-09 12:21:32 -04:00
void
when fetching MIDI ports for bundles and auto-connection, ignore control-only (and "virtual") MIDI ports "virtual" is a placeholder name for ALSA sequencer MIDI through ports
2016-10-21 16:32:46 -04:00
Session::get_physical_ports (vector<string>& inputs, vector<string>& outputs, DataType type,
MidiPortFlags include, MidiPortFlags exclude)
add Session::get_physical_ports() to centralize "get physical ports but (maybe) ignore control-only (and MIDI Through) ports"
2016-07-09 12:21:32 -04:00
{
when fetching MIDI ports for bundles and auto-connection, ignore control-only (and "virtual") MIDI ports "virtual" is a placeholder name for ALSA sequencer MIDI through ports
2016-10-21 16:32:46 -04:00
_engine.get_physical_inputs (type, inputs, include, exclude);
_engine.get_physical_outputs (type, outputs, include, exclude);
add Session::get_physical_ports() to centralize "get physical ports but (maybe) ignore control-only (and MIDI Through) ports"
2016-07-09 12:21:32 -04:00
}
many changes relating to session construction and audioengine interaction every session member is now initialized using C++ constructor syntax session construction reordered to clarify the split(s) between work where the engine is not relevant and work where is it is. this split is still not 100% obvious, but is enormously clearer than previously. if engine/backend are not running as session is created, and the SR of the sample rate is known, attempt to force backend to that value.
2013-09-10 15:41:19 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
void
Session::auto_connect_master_bus ()
{
if (!_master_out || !Config->get_auto_connect_standard_busses() || _monitor_out) {
return;
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* if requested auto-connect the outputs to the first N physical ports.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
uint32_t limit = _master_out->n_outputs().n_total();
vector<string> outputs[DataType::num_types];
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
for (uint32_t i = 0; i < DataType::num_types; ++i) {
_engine.get_physical_outputs (DataType (DataType::Symbol (i)), outputs[i]);
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
for (uint32_t n = 0; n < limit; ++n) {
boost::shared_ptr<Port> p = _master_out->output()->nth (n);
string connect_to;
if (outputs[p->type()].size() > n) {
connect_to = outputs[p->type()][n];
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if (!connect_to.empty() && p->connected_to (connect_to) == false) {
if (_master_out->output()->connect (p, connect_to, this)) {
error << string_compose (_("cannot connect master output %1 to %2"), n, connect_to)
<< endmsg;
break;
more work on the new all-Processor-all-The-Time redesign of Route - LOTS OF BREAKAGE STILL EXPECTED ; change all(?) methods that pass a start/end frame in to use sframes_t not nframes_t git-svn-id: svn://localhost/ardour2/branches/3.0@5074 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-13 17:34:09 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
void
Session::remove_monitor_section ()
{
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
if (!_monitor_out) {
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
return;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Don't allow to add/remove monitor-section w/o engine Adding/removing the Monitor section changes port-connections, a port-engine is needed to establish connections. Ideally we'd allow this and queue for connection changes. This will however require deep changes. Also toggling the monitor-section on/off/on w/o engine can later result in failure to register ports. Due to lack of engine-signals not all shared_ptr<> port references are dropped.
2018-11-25 21:06:22 -05:00
/* allow deletion when session is unloaded */
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (!_engine.running() && !deletion_in_progress ()) {
Don't allow to add/remove monitor-section w/o engine Adding/removing the Monitor section changes port-connections, a port-engine is needed to establish connections. Ideally we'd allow this and queue for connection changes. This will however require deep changes. Also toggling the monitor-section on/off/on w/o engine can later result in failure to register ports. Due to lack of engine-signals not all shared_ptr<> port references are dropped.
2018-11-25 21:06:22 -05:00
error << _("Cannot remove monitor section while the engine is offline.") << endmsg;
return;
}
Typo in comment. git-svn-id: svn://localhost/ardour2/branches/3.0@11758 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-03-25 15:38:17 -04:00
/* force reversion to Solo-In-Place */
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
Config->set_solo_control_is_listen_control (false);
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
allow auditioning via the monitor section to work. Ideally, we would feed the monitor section via an internal (aux) send/return, but this is an improvement over what we had before
2014-01-16 17:22:19 -05:00
/* if we are auditioning, cancel it ... this is a workaround
to a problem (auditioning does not execute the process graph,
which is needed to remove routes when using >1 core for processing)
*/
cancel_audition ();
Speed up session deletion No need to reconfigure routes due to removal of monitor or aux send/return. The processors will be removed with the route.
2019-03-17 12:35:10 -04:00
if (!deletion_in_progress ()) {
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* Hold process lock while doing this so that we don't hear bits and
* pieces of audio as we work on each route.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* Connect tracks to monitor section. Note that in an
existing session, the internal sends will already exist, but we want the
routes to notice that they connect to the control out specifically.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
RAII to postpone processor changes/graph recalculation
2016-11-25 12:07:43 -05:00
ProcessorChangeBlocker pcb (this, false);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
for (RouteList::iterator x = r->begin(); x != r->end(); ++x) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if ((*x)->is_monitor()) {
/* relax */
} else if ((*x)->is_master()) {
/* relax */
} else {
fix two major assert failures arising from the optional monitor section commit; separate numbering of aux sends, sends and listens to fix #3671 (still testing, but the assert failures are critical) git-svn-id: svn://localhost/ardour2/branches/3.0@11263 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-18 12:51:57 -05:00
(*x)->remove_aux_or_listen (_monitor_out);
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
}
}
}
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
remove_route (_monitor_out);
NO-OP: prefer to use API to check session-state
2019-03-17 12:04:45 -04:00
if (deletion_in_progress ()) {
further tweaks to session close
2016-04-23 20:51:23 -04:00
return;
}
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
auto_connect_master_bus ();
allow auditioning via the monitor section to work. Ideally, we would feed the monitor section via an internal (aux) send/return, but this is an improvement over what we had before
2014-01-16 17:22:19 -05:00
if (auditioner) {
auditioner->connect ();
}
further tweaks to session close
2016-04-23 20:51:23 -04:00
Revert Monitor-Section to be per session This partially reverts 639dff3a7c7. When loading a session, the monitor-bus that was saved with the session is used. This changes semantics of the monitor-section/config. Config::set_use_monitor_bus(bool) is used to initiate a change! Notification about the change is sent asynchronously by Session::MonitorBusAddedOrRemoved It is no longer possible to directly call add/remove_monitor_section() and leave the session + config in an inconsistent state.
2019-03-01 10:45:42 -05:00
MonitorBusAddedOrRemoved (); /* EMIT SIGNAL */
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
}
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
void
Session::add_monitor_section ()
{
RouteList rl;
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
Don't allow to add/remove monitor-section w/o engine Adding/removing the Monitor section changes port-connections, a port-engine is needed to establish connections. Ideally we'd allow this and queue for connection changes. This will however require deep changes. Also toggling the monitor-section on/off/on w/o engine can later result in failure to register ports. Due to lack of engine-signals not all shared_ptr<> port references are dropped.
2018-11-25 21:06:22 -05:00
if (!_engine.running()) {
error << _("Cannot create monitor section while the engine is offline.") << endmsg;
return;
}
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
if (_monitor_out || !_master_out) {
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
return;
}
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
boost::shared_ptr<Route> r (new Route (*this, _("Monitor"), PresentationInfo::MonitorOut, DataType::AUDIO));
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if (r->init ()) {
return;
}
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
make boost shared_ptr debugging a little easier to turn on/off
2016-05-07 13:32:31 -04:00
BOOST_MARK_ROUTE(r);
fail to create Monitor section if port-names are not unique We should still prevent a user from naming a route "Monitor", but since "Monitor" is a translatable string, this is an i18n safeguard. eg. - create a session with a mon-section and a track Монитор. - send the session to a friend in Russia.
2015-09-11 07:14:27 -04:00
try {
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
r->input()->ensure_io (_master_out->output()->n_ports(), false, this);
r->output()->ensure_io (_master_out->output()->n_ports(), false, this);
fail to create Monitor section if port-names are not unique We should still prevent a user from naming a route "Monitor", but since "Monitor" is a translatable string, this is an i18n safeguard. eg. - create a session with a mon-section and a track Монитор. - send the session to a friend in Russia.
2015-09-11 07:14:27 -04:00
} catch (...) {
error << _("Cannot create monitor section. 'Monitor' Port name is not unique.") << endmsg;
return;
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
}
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
rl.push_back (r);
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
add_routes (rl, false, false, false, 0);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
assert (_monitor_out);
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* AUDIO ONLY as of june 29th 2009, because listen semantics for anything else
are undefined, at best.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
uint32_t limit = _monitor_out->n_inputs().n_audio();
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if (_master_out) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* connect the inputs to the master bus outputs. this
* represents a separate data feed from the internal sends from
* each route. as of jan 2011, it allows the monitor section to
* conditionally ignore either the internal sends or the normal
* input feed, but we should really find a better way to do
* this, i think.
*/
_master_out->output()->disconnect (this);
for (uint32_t n = 0; n < limit; ++n) {
boost::shared_ptr<AudioPort> p = _monitor_out->input()->ports().nth_audio_port (n);
boost::shared_ptr<AudioPort> o = _master_out->output()->ports().nth_audio_port (n);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if (o) {
string connect_to = o->name();
if (_monitor_out->input()->connect (p, connect_to, this)) {
error << string_compose (_("cannot connect control input %1 to %2"), n, connect_to)
<< endmsg;
break;
}
}
}
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* if monitor section is not connected, connect it to physical outs
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Mixbus: auto-connect monitor-section Mixbus does not expose the "auto_connect_standard_busses" preference and forces it to false.
2016-04-20 10:17:22 -04:00
if ((Config->get_auto_connect_standard_busses () || Profile->get_mixbus ()) && !_monitor_out->output()->connected ()) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if (!Config->get_monitor_bus_preferred_bundle().empty()) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
boost::shared_ptr<Bundle> b = bundle_by_name (Config->get_monitor_bus_preferred_bundle());
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if (b) {
change input/output button context menus for mixer strip to be non-additive: when selecting a new IO bundle, disconnect from any existing connections before connecting to the new one (fixes #5087) git-svn-id: svn://localhost/ardour2/branches/3.0@13308 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-10-18 12:15:11 -04:00
_monitor_out->output()->connect_ports_to_bundle (b, true, this);
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
} else {
warning << string_compose (_("The preferred I/O for the monitor bus (%1) cannot be found"),
Config->get_monitor_bus_preferred_bundle())
<< endmsg;
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
} else {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* Monitor bus is audio only */
reinstate solo button for master out ; make solo button show state even if in listen mode ; wire up control outs "better" and add an option to name the bundle to connect it to (not controllable in GUI yet) ; fix up a couple of other solo/listen issues git-svn-id: svn://localhost/ardour2/branches/3.0@5301 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 19:14:27 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
vector<string> outputs[DataType::num_types];
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
for (uint32_t i = 0; i < DataType::num_types; ++i) {
_engine.get_physical_outputs (DataType (DataType::Symbol (i)), outputs[i]);
}
more session construction order changes, and removal of n_physical_{inputs,outputs} members which were (a) not initialized early enough (b) not used anywhere except monitor bus connection. Things almost make sense now.
2013-09-16 12:08:19 -04:00
uint32_t mod = outputs[DataType::AUDIO].size();
uint32_t limit = _monitor_out->n_outputs().get (DataType::AUDIO);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if (mod != 0) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
for (uint32_t n = 0; n < limit; ++n) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
boost::shared_ptr<Port> p = _monitor_out->output()->ports().port(DataType::AUDIO, n);
string connect_to;
if (outputs[DataType::AUDIO].size() > (n % mod)) {
connect_to = outputs[DataType::AUDIO][n % mod];
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if (!connect_to.empty()) {
if (_monitor_out->output()->connect (p, connect_to, this)) {
error << string_compose (
_("cannot connect control output %1 to %2"),
n, connect_to)
<< endmsg;
break;
reinstate solo button for master out ; make solo button show state even if in listen mode ; wire up control outs "better" and add an option to name the bundle to connect it to (not controllable in GUI yet) ; fix up a couple of other solo/listen issues git-svn-id: svn://localhost/ardour2/branches/3.0@5301 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 19:14:27 -04:00
}
first pass (ok, third really) at internal send+return - audio routing inside ardour without JACK. lots still to do, but at least the obvious works git-svn-id: svn://localhost/ardour2/branches/3.0@5202 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-16 10:58:33 -04:00
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
}
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* Hold process lock while doing this so that we don't hear bits and
* pieces of audio as we work on each route.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
/* Connect tracks to monitor section. Note that in an
existing session, the internal sends will already exist, but we want the
routes to notice that they connect to the control out specifically.
*/
fixes from 2.X for latency/capture alignment stuff: don't reverse route list, update latency and capture offsets at transport stop; remove some debugging output git-svn-id: svn://localhost/ardour2/branches/3.0@10110 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-09-21 13:28:13 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
boost::shared_ptr<RouteList> rls = routes.reader ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
RAII to postpone processor changes/graph recalculation
2016-11-25 12:07:43 -05:00
ProcessorChangeBlocker pcb (this, false /* XXX */);
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
for (RouteList::iterator x = rls->begin(); x != rls->end(); ++x) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
if ((*x)->is_monitor()) {
/* relax */
} else if ((*x)->is_master()) {
/* relax */
} else {
(*x)->enable_monitor_send ();
}
}
allow auditioning via the monitor section to work. Ideally, we would feed the monitor section via an internal (aux) send/return, but this is an improvement over what we had before
2014-01-16 17:22:19 -05:00
if (auditioner) {
auditioner->connect ();
}
Revert Monitor-Section to be per session This partially reverts 639dff3a7c7. When loading a session, the monitor-bus that was saved with the session is used. This changes semantics of the monitor-section/config. Config::set_use_monitor_bus(bool) is used to initiate a change! Notification about the change is sent asynchronously by Session::MonitorBusAddedOrRemoved It is no longer possible to directly call add/remove_monitor_section() and leave the session + config in an inconsistent state.
2019-03-01 10:45:42 -05:00
MonitorBusAddedOrRemoved (); /* EMIT SIGNAL */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
merge monitor_section branch
2015-04-21 09:18:10 -04:00
void
Session::reset_monitor_section ()
{
/* Process lock should be held by the caller.*/
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
if (!_monitor_out) {
merge monitor_section branch
2015-04-21 09:18:10 -04:00
return;
}
uint32_t limit = _master_out->n_outputs().n_audio();
/* connect the inputs to the master bus outputs. this
* represents a separate data feed from the internal sends from
* each route. as of jan 2011, it allows the monitor section to
* conditionally ignore either the internal sends or the normal
* input feed, but we should really find a better way to do
* this, i think.
*/
_master_out->output()->disconnect (this);
_monitor_out->output()->disconnect (this);
ignore midi on monitor-section and force strict-i/o
2016-06-12 14:16:16 -04:00
// monitor section follow master bus - except midi
ChanCount mon_chn (_master_out->output()->n_ports());
mon_chn.set_midi (0);
_monitor_out->input()->ensure_io (mon_chn, false, this);
_monitor_out->output()->ensure_io (mon_chn, false, this);
merge monitor_section branch
2015-04-21 09:18:10 -04:00
for (uint32_t n = 0; n < limit; ++n) {
boost::shared_ptr<AudioPort> p = _monitor_out->input()->ports().nth_audio_port (n);
boost::shared_ptr<AudioPort> o = _master_out->output()->ports().nth_audio_port (n);
if (o) {
string connect_to = o->name();
if (_monitor_out->input()->connect (p, connect_to, this)) {
error << string_compose (_("cannot connect control input %1 to %2"), n, connect_to)
<< endmsg;
break;
}
}
}
/* connect monitor section to physical outs
*/
if (Config->get_auto_connect_standard_busses()) {
if (!Config->get_monitor_bus_preferred_bundle().empty()) {
boost::shared_ptr<Bundle> b = bundle_by_name (Config->get_monitor_bus_preferred_bundle());
if (b) {
_monitor_out->output()->connect_ports_to_bundle (b, true, this);
} else {
warning << string_compose (_("The preferred I/O for the monitor bus (%1) cannot be found"),
Config->get_monitor_bus_preferred_bundle())
<< endmsg;
}
} else {
/* Monitor bus is audio only */
vector<string> outputs[DataType::num_types];
for (uint32_t i = 0; i < DataType::num_types; ++i) {
_engine.get_physical_outputs (DataType (DataType::Symbol (i)), outputs[i]);
}
uint32_t mod = outputs[DataType::AUDIO].size();
uint32_t limit = _monitor_out->n_outputs().get (DataType::AUDIO);
if (mod != 0) {
for (uint32_t n = 0; n < limit; ++n) {
boost::shared_ptr<Port> p = _monitor_out->output()->ports().port(DataType::AUDIO, n);
string connect_to;
if (outputs[DataType::AUDIO].size() > (n % mod)) {
connect_to = outputs[DataType::AUDIO][n % mod];
}
if (!connect_to.empty()) {
if (_monitor_out->output()->connect (p, connect_to, this)) {
error << string_compose (
_("cannot connect control output %1 to %2"),
n, connect_to)
<< endmsg;
break;
}
}
}
}
}
}
/* Connect tracks to monitor section. Note that in an
existing session, the internal sends will already exist, but we want the
routes to notice that they connect to the control out specifically.
*/
boost::shared_ptr<RouteList> rls = routes.reader ();
RAII to postpone processor changes/graph recalculation
2016-11-25 12:07:43 -05:00
ProcessorChangeBlocker pcb (this, false);
merge monitor_section branch
2015-04-21 09:18:10 -04:00
for (RouteList::iterator x = rls->begin(); x != rls->end(); ++x) {
if ((*x)->is_monitor()) {
/* relax */
} else if ((*x)->is_master()) {
/* relax */
} else {
(*x)->enable_monitor_send ();
}
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Separate "add master bus" (and add Lua bindings) This is in preparation for "advanced session setup" allow a SessionSetup Lua script to create the master-bus.
2017-08-15 17:17:08 -04:00
int
Session::add_master_bus (ChanCount const& count)
{
if (master_out ()) {
return -1;
}
RouteList rl;
boost::shared_ptr<Route> r (new Route (*this, _("Master"), PresentationInfo::MasterOut, DataType::AUDIO));
if (r->init ()) {
return -1;
}
BOOST_MARK_ROUTE(r);
{
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
r->input()->ensure_io (count, false, this);
r->output()->ensure_io (count, false, this);
}
rl.push_back (r);
add_routes (rl, false, false, false, PresentationInfo::max_order);
return 0;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
remove near-duplicate session constructor; change names from control_outs to monitor_out and control_out to monitor_send; may havebroken creation of new sessions in a wierd way (will fix later) git-svn-id: svn://localhost/ardour2/branches/3.0@6783 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-22 17:35:35 -04:00
Session::hookup_io ()
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
/* stop graph reordering notifications from
causing resorts, etc.
*/
_state_of_the_state = StateOfTheState (_state_of_the_state | InitialConnecting);
more work on the new all-Processor-all-The-Time redesign of Route - LOTS OF BREAKAGE STILL EXPECTED ; change all(?) methods that pass a start/end frame in to use sframes_t not nframes_t git-svn-id: svn://localhost/ardour2/branches/3.0@5074 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-13 17:34:09 -04:00
if (!auditioner) {
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* we delay creating the auditioner till now because
it makes its own connections to ports.
*/
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
try {
Use commented out (correct) Mute/SoloControllable set_value() code; requires adding enable_shared_from_this to Route. git-svn-id: svn://localhost/ardour2/branches/3.0@8930 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-22 16:15:42 -05:00
boost::shared_ptr<Auditioner> a (new Auditioner (*this));
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (a->init()) {
Use commented out (correct) Mute/SoloControllable set_value() code; requires adding enable_shared_from_this to Route. git-svn-id: svn://localhost/ardour2/branches/3.0@8930 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-22 16:15:42 -05:00
throw failed_constructor ();
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
Use commented out (correct) Mute/SoloControllable set_value() code; requires adding enable_shared_from_this to Route. git-svn-id: svn://localhost/ardour2/branches/3.0@8930 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-22 16:15:42 -05:00
auditioner = a;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
catch (failed_constructor& err) {
warning << _("cannot create Auditioner: no auditioning of regions possible") << endmsg;
}
}
/* load bundles, which we may have postponed earlier on */
if (_bundle_xml_node) {
load_bundles (*_bundle_xml_node);
delete _bundle_xml_node;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* Tell all IO objects to connect themselves together */
IO::enable_connecting ();
major redesign of MIDI port heirarchy and management (part 2)
2013-08-07 22:22:11 -04:00
/* Now tell all "floating" ports to connect to whatever
they should be connected to.
*/
AudioEngine::instance()->reconnect_ports ();
add finite state machine to control/manage transport state
2019-09-17 20:26:03 -04:00
TransportMasterManager::instance().reconnect_ports ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* Anyone who cares about input state, wake up and do something */
IOConnectionsComplete (); /* EMIT SIGNAL */
_state_of_the_state = StateOfTheState (_state_of_the_state & ~InitialConnecting);
/* now handle the whole enchilada as if it was one
graph reorder event.
*/
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
graph_reordered (false);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
/* update the full solo state, which can't be
correctly determined on a per-route basis, but
needs the global overview that only the session
has.
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
update_route_solo_state ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
void
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
Session::track_playlist_changed (boost::weak_ptr<Track> wp)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
boost::shared_ptr<Track> track = wp.lock ();
if (!track) {
Fix uninitialised variable and remove bind of shared_ptr to a slot. git-svn-id: svn://localhost/ardour2/branches/3.0@6057 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-10 18:54:04 -05:00
return;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
boost::shared_ptr<Playlist> playlist;
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
if ((playlist = track->playlist()) != 0) {
Fix session range location by hooking into the undo system. Fixes #3654. git-svn-id: svn://localhost/ardour2/branches/3.0@8539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:56 -05:00
playlist->RegionAdded.connect_same_thread (*this, boost::bind (&Session::playlist_region_added, this, _1));
playlist->RangesMoved.connect_same_thread (*this, boost::bind (&Session::playlist_ranges_moved, this, _1));
Update session range on trimming regions. git-svn-id: svn://localhost/ardour2/branches/3.0@9004 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-01 11:23:31 -05:00
playlist->RegionsExtended.connect_same_thread (*this, boost::bind (&Session::playlist_regions_extended, this, _1));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
bool
Session::record_enabling_legal () const
{
if (Config->get_all_safe()) {
return false;
}
return true;
}
fix monitoring so that MIDI tracks don't work the same way as audio (basically, they are always in "ardour does monitoring" mode git-svn-id: svn://localhost/ardour2/branches/3.0@9081 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-05 18:16:32 -05:00
void
Session::set_track_monitor_input_status (bool yn)
{
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
boost::shared_ptr<RouteList> rl = routes.reader ();
for (RouteList::iterator i = rl->begin(); i != rl->end(); ++i) {
boost::shared_ptr<AudioTrack> tr = boost::dynamic_pointer_cast<AudioTrack> (*i);
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
if (tr && tr->rec_enable_control()->get_value()) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
//cerr << "switching to input = " << !auto_input << __FILE__ << __LINE__ << endl << endl;
more purging of JACK as an explicit name from libardour
2013-08-09 13:56:23 -04:00
tr->request_input_monitoring (yn);
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
}
}
fix monitoring so that MIDI tracks don't work the same way as audio (basically, they are always in "ardour does monitoring" mode git-svn-id: svn://localhost/ardour2/branches/3.0@9081 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-05 18:16:32 -05:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::auto_punch_start_changed (Location* location)
{
move Session::Event into SessionEvent class; add SessionEventManager (Session IS-A SessionEventManager); make session ops to toggle all track rec-enable be atomic with respect to process() git-svn-id: svn://localhost/ardour2/branches/3.0@6273 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-03 21:15:12 -05:00
replace_event (SessionEvent::PunchIn, location->start());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Align punch in/out recording with latency-compensation
2017-09-29 20:45:13 -04:00
if (get_record_enabled() && config.get_punch_in() && !actively_recording ()) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* capture start has been changed, so save new pending state */
save_state ("", true);
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::auto_punch_end_changed (Location* location)
{
Align punch in/out recording with latency-compensation
2017-09-29 20:45:13 -04:00
replace_event (SessionEvent::PunchOut, location->end());
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::auto_punch_changed (Location* location)
{
Align punch in/out recording with latency-compensation
2017-09-29 20:45:13 -04:00
auto_punch_start_changed (location);
auto_punch_end_changed (location);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::auto_loop_changed (Location* location)
{
fix startup crash if no loop range is defined
2019-11-23 15:37:00 -05:00
if (!location) {
return;
}
move Session::Event into SessionEvent class; add SessionEventManager (Session IS-A SessionEventManager); make session ops to toggle all track rec-enable be atomic with respect to process() git-svn-id: svn://localhost/ardour2/branches/3.0@6273 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-03 21:15:12 -05:00
replace_event (SessionEvent::AutoLoop, location->end(), location->start());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
const bool rolling = transport_rolling ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
more tweaks to correctly (or more correctly) reload disk reader buffers when loop fade choice changes
2019-11-23 02:09:46 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
make 3.0 catch up with transport and other changes in 2.X (hand applied, not merged) git-svn-id: svn://localhost/ardour2/branches/3.0@5989 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-30 14:14:25 -04:00
more tweaks to correctly (or more correctly) reload disk reader buffers when loop fade choice changes
2019-11-23 02:09:46 -05:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
(*i)->reload_loop ();
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
changes to use overwrite-buffers when loop is disabled or loop range changed
2019-12-05 15:00:24 -05:00
more tweaks to correctly (or more correctly) reload disk reader buffers when loop fade choice changes
2019-11-23 02:09:46 -05:00
if (rolling) {
redesign of declicking and fades around loop boundaries
2019-11-23 01:41:56 -05:00
more tweaks to correctly (or more correctly) reload disk reader buffers when loop fade choice changes
2019-11-23 02:09:46 -05:00
if (play_loop) {
redesign of declicking and fades around loop boundaries
2019-11-23 01:41:56 -05:00
avoid a locate-to-loop-start from cancelling looping
2019-12-08 19:00:30 -05:00
if (_transport_sample < location->start() || _transport_sample > location->end()) {
changes to use overwrite-buffers when loop is disabled or loop range changed
2019-12-05 15:00:24 -05:00
avoid a locate-to-loop-start from cancelling looping
2019-12-08 19:00:30 -05:00
/* new loop range excludes current transport
* sample => relocate to beginning of loop and roll.
*/
changes to use overwrite-buffers when loop is disabled or loop range changed
2019-12-05 15:00:24 -05:00
avoid a locate-to-loop-start from cancelling looping
2019-12-08 19:00:30 -05:00
/* Set this so that when/if we have to stop the
* transport for a locate, we know that it is triggered
* by loop-changing, and we do not cancel play loop
*/
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
avoid a locate-to-loop-start from cancelling looping
2019-12-08 19:00:30 -05:00
loop_changing = true;
changes to use overwrite-buffers when loop is disabled or loop range changed
2019-12-05 15:00:24 -05:00
request_locate (location->start(), true);
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
avoid a locate-to-loop-start from cancelling looping
2019-12-08 19:00:30 -05:00
} else {
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
avoid a locate-to-loop-start from cancelling looping
2019-12-08 19:00:30 -05:00
// schedule a buffer overwrite to refill buffers with the new loop.
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
revert to single buffer for disk playback, and 5.x-style overwrite Also address issues with MIDI and also atomicity of _pending_overwrite
2019-12-10 11:28:41 -05:00
request_overwrite_buffer (boost::shared_ptr<Track>(), LoopChanged);
avoid a locate-to-loop-start from cancelling looping
2019-12-08 19:00:30 -05:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
fix transport loop sub-state events Loading a session with a loop-range, or setting a loop-range adds LoopDeclick events. Normal play back (no loop) never processes/clears the events. Ardour was constantly doing split cycles for nothing. This lead to stuck midi-notes: Split cycle calls increment_global_port_buffer_offset() but the midi_diskstream itself was not aware of any loop positions or offsets. Entering loop playback adds those events explicitly in Session::set_play_loop(). auto_loop_changed() only needs to add/replace session-events when actively looping.
2015-06-11 09:05:46 -04:00
} else {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
/* possibly move playhead if not rolling; if we are rolling we'll move
to the loop start on stop if that is appropriate.
*/
follow various events related to playhead priority (loop changes, parameter changes); remove debug output
2015-01-16 17:50:10 -05:00
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
samplepos_t pos;
follow various events related to playhead priority (loop changes, parameter changes); remove debug output
2015-01-16 17:50:10 -05:00
fix the way Session::auto_loop_changed() works to put back various things removed since 5.x
2019-11-01 23:14:40 -04:00
if (select_playhead_priority_target (pos)) {
if (pos == location->start()) {
request_locate (pos);
}
follow various events related to playhead priority (loop changes, parameter changes); remove debug output
2015-01-16 17:50:10 -05:00
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
last_loopend = location->end();
mark session dirty when loop location is changed
2015-02-17 10:50:54 -05:00
set_dirty ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
void
Session::set_auto_punch_location (Location* location)
{
Location* existing;
Allow markers to be glued to bar/beat time. Fixes #1815. git-svn-id: svn://localhost/ardour2/branches/3.0@7573 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-09 12:40:31 -04:00
if ((existing = _locations->auto_punch_location()) != 0 && existing != location) {
switch to using boost::signals2 instead of sigc++, at least for libardour. not finished yet, but compiles, loads sessions, records and can close a session without a crash git-svn-id: svn://localhost/ardour2/branches/3.0@6372 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-17 13:24:23 -05:00
punch_connections.drop_connections();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
existing->set_auto_punch (false, this);
Align punch in/out recording with latency-compensation
2017-09-29 20:45:13 -04:00
clear_events (SessionEvent::PunchIn);
move Session::Event into SessionEvent class; add SessionEventManager (Session IS-A SessionEventManager); make session ops to toggle all track rec-enable be atomic with respect to process() git-svn-id: svn://localhost/ardour2/branches/3.0@6273 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-03 21:15:12 -05:00
clear_events (SessionEvent::PunchOut);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
auto_punch_location_changed (0);
}
set_dirty();
if (location == 0) {
return;
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (location->end() <= location->start()) {
error << _("Session: you can't use that location for auto punch (start <= end)") << endmsg;
return;
}
switch to using boost::signals2 instead of sigc++, at least for libardour. not finished yet, but compiles, loads sessions, records and can close a session without a crash git-svn-id: svn://localhost/ardour2/branches/3.0@6372 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-17 13:24:23 -05:00
punch_connections.drop_connections ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
location->StartChanged.connect_same_thread (punch_connections, boost::bind (&Session::auto_punch_start_changed, this, location));
location->EndChanged.connect_same_thread (punch_connections, boost::bind (&Session::auto_punch_end_changed, this, location));
location->Changed.connect_same_thread (punch_connections, boost::bind (&Session::auto_punch_changed, this, location));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
location->set_auto_punch (true, this);
auto_punch_changed (location);
auto_punch_location_changed (location);
}
add set-session-extents-from-edit-range
2014-12-09 17:17:47 -05:00
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::set_session_extents (samplepos_t start, samplepos_t end)
add set-session-extents-from-edit-range
2014-12-09 17:17:47 -05:00
{
Location* existing;
if ((existing = _locations->session_range_location()) == 0) {
//if there is no existing session, we need to make a new session location (should never happen)
allow all types of range location (loop, start, end etc.) to be glued to bars and beats.
2016-12-28 09:39:31 -05:00
existing = new Location (*this, 0, 0, _("session"), Location::IsSessionRange, 0);
add set-session-extents-from-edit-range
2014-12-09 17:17:47 -05:00
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add set-session-extents-from-edit-range
2014-12-09 17:17:47 -05:00
if (end <= start) {
error << _("Session: you can't use that location for session start/end)") << endmsg;
return;
}
existing->set( start, end );
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add set-session-extents-from-edit-range
2014-12-09 17:17:47 -05:00
set_dirty();
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::set_auto_loop_location (Location* location)
{
Location* existing;
Allow markers to be glued to bar/beat time. Fixes #1815. git-svn-id: svn://localhost/ardour2/branches/3.0@7573 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-09 12:40:31 -04:00
if ((existing = _locations->auto_loop_location()) != 0 && existing != location) {
switch to using boost::signals2 instead of sigc++, at least for libardour. not finished yet, but compiles, loads sessions, records and can close a session without a crash git-svn-id: svn://localhost/ardour2/branches/3.0@6372 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-17 13:24:23 -05:00
loop_connections.drop_connections ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
existing->set_auto_loop (false, this);
move Session::Event into SessionEvent class; add SessionEventManager (Session IS-A SessionEventManager); make session ops to toggle all track rec-enable be atomic with respect to process() git-svn-id: svn://localhost/ardour2/branches/3.0@6273 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-03 21:15:12 -05:00
remove_event (existing->end(), SessionEvent::AutoLoop);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
auto_loop_location_changed (0);
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
set_dirty();
if (location == 0) {
return;
}
if (location->end() <= location->start()) {
better error msg about autoloop location git-svn-id: svn://localhost/ardour2/branches/3.0@13842 d708f5d6-7413-0410-9779-e7cbd77b26cf
2013-01-15 09:15:01 -05:00
error << _("You cannot use this location for auto-loop because it has zero or negative length") << endmsg;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
return;
}
last_loopend = location->end();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
switch to using boost::signals2 instead of sigc++, at least for libardour. not finished yet, but compiles, loads sessions, records and can close a session without a crash git-svn-id: svn://localhost/ardour2/branches/3.0@6372 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-17 13:24:23 -05:00
loop_connections.drop_connections ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
location->StartChanged.connect_same_thread (loop_connections, boost::bind (&Session::auto_loop_changed, this, location));
location->EndChanged.connect_same_thread (loop_connections, boost::bind (&Session::auto_loop_changed, this, location));
location->Changed.connect_same_thread (loop_connections, boost::bind (&Session::auto_loop_changed, this, location));
when loop range is set, if loop-is-mode and seamless looping, set loop for all tracks
2015-05-10 13:15:46 -04:00
location->FlagsChanged.connect_same_thread (loop_connections, boost::bind (&Session::auto_loop_changed, this, location));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
location->set_auto_loop (true, this);
remove seamless looping as an option (it's now the only kind of looping we support)
2019-11-01 16:04:16 -04:00
if (Config->get_loop_is_mode() && play_loop) {
when loop range is set, if loop-is-mode and seamless looping, set loop for all tracks
2015-05-10 13:15:46 -04:00
// set all tracks to use internal looping
boost::shared_ptr<RouteList> rl = routes.reader ();
for (RouteList::iterator i = rl->begin(); i != rl->end(); ++i) {
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
remove Track::hidden(); replace with Stripable::is_private_route()
2017-07-25 11:26:14 -04:00
if (tr && !tr->is_private_route()) {
when loop range is set, if loop-is-mode and seamless looping, set loop for all tracks
2015-05-10 13:15:46 -04:00
tr->set_loop (location);
}
}
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* take care of our stuff first */
auto_loop_changed (location);
/* now tell everyone else */
auto_loop_location_changed (location);
}
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
void
Session::update_marks (Location*)
{
set_dirty ();
}
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
void
Session::update_skips (Location* loc, bool consolidate)
{
noop: indentation
2015-05-10 13:19:02 -04:00
if (_ignore_skips_updates) {
return;
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
Locations::LocationList skips;
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
if (consolidate) {
PBD::Unwinder<bool> uw (_ignore_skips_updates, true);
consolidate_skips (loc);
}
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
sync_locations_to_skips ();
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
set_dirty ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
void
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
Session::consolidate_skips (Location* loc)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
Locations::LocationList all_locations = _locations->list ();
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
for (Locations::LocationList::iterator l = all_locations.begin(); l != all_locations.end(); ) {
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
if (!(*l)->is_skip ()) {
++l;
continue;
}
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
/* don't test against self */
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
if (*l == loc) {
++l;
continue;
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
switch (Evoral::coverage ((*l)->start(), (*l)->end(), loc->start(), loc->end())) {
case Evoral::OverlapInternal:
case Evoral::OverlapExternal:
case Evoral::OverlapStart:
case Evoral::OverlapEnd:
/* adjust new location to cover existing one */
loc->set_start (min (loc->start(), (*l)->start()));
loc->set_end (max (loc->end(), (*l)->end()));
/* we don't need this one any more */
_locations->remove (*l);
/* the location has been deleted, so remove reference to it in our local list */
l = all_locations.erase (l);
break;
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
case Evoral::OverlapNone:
++l;
break;
}
}
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
}
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
void
Session::sync_locations_to_skips ()
{
/* This happens asynchronously (in the audioengine thread). After the clear is done, we will call
* Session::_sync_locations_to_skips() from the audioengine thread.
*/
clear_events (SessionEvent::Skip, boost::bind (&Session::_sync_locations_to_skips, this));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
void
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
Session::_sync_locations_to_skips ()
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
/* called as a callback after existing Skip events have been cleared from a realtime audioengine thread */
Locations::LocationList const & locs (_locations->list());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
for (Locations::LocationList::const_iterator i = locs.begin(); i != locs.end(); ++i) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
Location* location = *i;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
if (location->is_skip() && location->is_skipping()) {
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
SessionEvent* ev = new SessionEvent (SessionEvent::Skip, SessionEvent::Add, location->start(), location->end(), 1.0);
queue_event (ev);
}
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
modify behaviour of session when updating skips to use new SessionEvent API Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc
2015-02-05 16:21:01 -05:00
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
void
Session::location_added (Location *location)
{
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
if (location->is_auto_punch()) {
set_auto_punch_location (location);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
if (location->is_auto_loop()) {
set_auto_loop_location (location);
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
if (location->is_session_range()) {
/* no need for any signal handling or event setting with the session range,
because we keep a direct reference to it and use its start/end directly.
*/
_session_range_location = location;
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
if (location->is_mark()) {
/* listen for per-location signals that require us to do any * global updates for marks */
mark session dirty when markers/ranges are changed
2015-05-08 12:20:25 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
location->StartChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
location->EndChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
location->Changed.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
location->FlagsChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
mark session dirty when moving range markers or altering location lock style.
2017-01-26 11:07:29 -05:00
location->PositionLockStyleChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
}
mark session dirty when moving range markers or altering location lock style.
2017-01-26 11:07:29 -05:00
if (location->is_range_marker()) {
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
/* listen for per-location signals that require us to do any * global updates for marks */
mark session dirty when moving range markers or altering location lock style.
2017-01-26 11:07:29 -05:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
location->StartChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
location->EndChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
location->Changed.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
location->FlagsChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
mark session dirty when moving range markers or altering location lock style.
2017-01-26 11:07:29 -05:00
location->PositionLockStyleChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
}
mark session dirty when markers/ranges are changed
2015-05-08 12:20:25 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
if (location->is_skip()) {
/* listen for per-location signals that require us to update skip-locate events */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
location->StartChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_skips, this, location, true));
location->EndChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_skips, this, location, true));
location->Changed.connect_same_thread (skip_update_connections, boost::bind (&Session::update_skips, this, location, true));
location->FlagsChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_skips, this, location, false));
mark session dirty when moving range markers or altering location lock style.
2017-01-26 11:07:29 -05:00
location->PositionLockStyleChanged.connect_same_thread (skip_update_connections, boost::bind (&Session::update_marks, this, location));
initial attempt to push certain kinds of Locations into LocateRoll events on the timeline for skip lists
2014-09-16 21:32:58 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
update_skips (location, true);
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
set_dirty ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial attempt to push certain kinds of Locations into LocateRoll events on the timeline for skip lists
2014-09-16 21:32:58 -04:00
void
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
Session::location_removed (Location *location)
initial attempt to push certain kinds of Locations into LocateRoll events on the timeline for skip lists
2014-09-16 21:32:58 -04:00
{
NO-OP: whitespace
2018-10-05 15:47:43 -04:00
if (location->is_auto_loop()) {
set_auto_loop_location (0);
Properly disable looping when deleting loop-range
2018-10-05 15:48:41 -04:00
if (!play_loop) {
set_track_loop (false);
}
unset_play_loop ();
NO-OP: whitespace
2018-10-05 15:47:43 -04:00
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
NO-OP: whitespace
2018-10-05 15:47:43 -04:00
if (location->is_auto_punch()) {
set_auto_punch_location (0);
}
initial attempt to push certain kinds of Locations into LocateRoll events on the timeline for skip lists
2014-09-16 21:32:58 -04:00
NO-OP: whitespace
2018-10-05 15:47:43 -04:00
if (location->is_session_range()) {
/* this is never supposed to happen */
error << _("programming error: session range removed!") << endl;
}
initial attempt to push certain kinds of Locations into LocateRoll events on the timeline for skip lists
2014-09-16 21:32:58 -04:00
NO-OP: whitespace
2018-10-05 15:47:43 -04:00
if (location->is_skip()) {
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
NO-OP: whitespace
2018-10-05 15:47:43 -04:00
update_skips (location, false);
}
initial attempt to push certain kinds of Locations into LocateRoll events on the timeline for skip lists
2014-09-16 21:32:58 -04:00
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
set_dirty ();
}
initial attempt to push certain kinds of Locations into LocateRoll events on the timeline for skip lists
2014-09-16 21:32:58 -04:00
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
void
Session::locations_changed ()
{
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
_locations->apply (*this, &Session::_locations_changed);
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
}
void
Session::_locations_changed (const Locations::LocationList& locations)
{
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
/* There was some mass-change in the Locations object.
add finite state machine to control/manage transport state
2019-09-17 20:26:03 -04:00
*
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
* We might be re-adding a location here but it doesn't actually matter
* for all the locations that the Session takes an interest in.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
fix whitespace/indentation and remove references to ScopedConnectionLists that were not used even in Tracks
2015-05-10 13:44:23 -04:00
{
PBD::Unwinder<bool> protect_ignore_skip_updates (_ignore_skips_updates, true);
for (Locations::LocationList::const_iterator i = locations.begin(); i != locations.end(); ++i) {
location_added (*i);
}
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
fix whitespace/indentation and remove references to ScopedConnectionLists that were not used even in Tracks
2015-05-10 13:44:23 -04:00
update_skips (NULL, false);
initial attempt to push certain kinds of Locations into LocateRoll events on the timeline for skip lists
2014-09-16 21:32:58 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::enable_record ()
{
fix stupid thinko in varispeed limitation for record enable git-svn-id: svn://localhost/ardour2/branches/3.0@9980 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-11 13:36:42 -04:00
if (_transport_speed != 0.0 && _transport_speed != 1.0) {
don't use MIDI tracks when adding internal sends to auxes; no varispeed when recording and the corollary - no recording when varispeeding git-svn-id: svn://localhost/ardour2/branches/3.0@9974 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 16:21:18 -04:00
/* no recording at anything except normal speed */
attempt to prevent dropping into record if rolling in reverse (should fix #4254) git-svn-id: svn://localhost/ardour2/branches/3.0@9971 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-09 17:17:55 -04:00
return;
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
while (1) {
RecordState rs = (RecordState) g_atomic_int_get (&_record_status);
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (rs == Recording) {
break;
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (g_atomic_int_compare_and_exchange (&_record_status, rs, Recording)) {
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
_last_record_location = _transport_sample;
fix delivery of MMC events under split process cycle conditions
2014-10-22 17:06:53 -04:00
send_immediate_mmc (MIDI::MachineControlCommand (MIDI::MachineControl::cmdRecordStrobe));
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (Config->get_monitoring_model() == HardwareMonitoring && config.get_auto_input()) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
set_track_monitor_input_status (true);
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
RecordStateChanged ();
break;
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Add all_tracks_rec actions
2016-01-19 15:00:03 -05:00
void
Session::set_all_tracks_record_enabled (bool enable )
{
boost::shared_ptr<RouteList> rl = routes.reader();
use new record safe control in libardour
2016-05-27 16:11:47 -04:00
set_controls (route_list_to_control_list (rl, &Stripable::rec_enable_control), enable, Controllable::NoGroup);
Add all_tracks_rec actions
2016-01-19 15:00:03 -05:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::disable_record (bool rt_context, bool force)
{
RecordState rs;
if ((rs = (RecordState) g_atomic_int_get (&_record_status)) != Disabled) {
Session::disable_record() should work whether we are in loop mode or not
2015-09-22 13:18:08 -04:00
if (!Config->get_latched_record_enable () || force) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
g_atomic_int_set (&_record_status, Disabled);
fix delivery of MMC events under split process cycle conditions
2014-10-22 17:06:53 -04:00
send_immediate_mmc (MIDI::MachineControlCommand (MIDI::MachineControl::cmdRecordExit));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
} else {
if (rs == Recording) {
g_atomic_int_set (&_record_status, Enabled);
}
}
First stage of options rework. - Split Configuration into RCConfiguration and SessionConfiguration; the first for options which are saved to .rc files and the second for options which are saved in a session file. - Move some options from the old `master' Configuration object into SessionConfiguration; this needs more refinement. - Reflect many RCConfiguration options in an expanded Edit->Preferences dialog; my intention is to remove the corresponding menu items eventually. git-svn-id: svn://localhost/ardour2/branches/3.0@5075 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-13 20:13:27 -04:00
if (Config->get_monitoring_model() == HardwareMonitoring && config.get_auto_input()) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
set_track_monitor_input_status (false);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
RecordStateChanged (); /* emit signal */
}
}
void
Session::step_back_from_record ()
{
alignment/capture/positioning changes ported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@6886 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-12 18:35:06 -04:00
if (g_atomic_int_compare_and_exchange (&_record_status, Recording, Enabled)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
First stage of options rework. - Split Configuration into RCConfiguration and SessionConfiguration; the first for options which are saved to .rc files and the second for options which are saved in a session file. - Move some options from the old `master' Configuration object into SessionConfiguration; this needs more refinement. - Reflect many RCConfiguration options in an expanded Edit->Preferences dialog; my intention is to remove the corresponding menu items eventually. git-svn-id: svn://localhost/ardour2/branches/3.0@5075 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-13 20:13:27 -04:00
if (Config->get_monitoring_model() == HardwareMonitoring && config.get_auto_input()) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
set_track_monitor_input_status (false);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
notify about commit change when we slip out of rec-enabled due to auto-punch git-svn-id: svn://localhost/ardour2/branches/3.0@10274 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-10-21 19:12:19 -04:00
RecordStateChanged (); /* emit signal */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
void
Match API: add rt_context option for en/disable_record() This allows to call maybe_enable_record () from Lua session or DSP scripts
2017-01-01 18:26:21 -05:00
Session::maybe_enable_record (bool rt_context)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (_step_editors > 0) {
return;
}
step editing now accessed via a context menu on MIDI track rec-enable buttons. along the way, fixed up confusion between set_rec_enable() and set_rec_enabled() methods in Route/Track/Diskstream git-svn-id: svn://localhost/ardour2/branches/3.0@7482 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-24 12:40:56 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
g_atomic_int_set (&_record_status, Enabled);
Revert previous commit, adding comment as to why. git-svn-id: svn://localhost/ardour2/branches/3.0@7822 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-09-21 11:02:45 -04:00
/* This function is currently called from somewhere other than an RT thread.
Match API: add rt_context option for en/disable_record() This allows to call maybe_enable_record () from Lua session or DSP scripts
2017-01-01 18:26:21 -05:00
* (except maybe lua scripts, which can use rt_context = true)
* This save_state() call therefore doesn't impact anything. Doing it here
* means that we save pending state of which sources the next record will use,
* which gives us some chance of recovering from a crash during the record.
*/
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Match API: add rt_context option for en/disable_record() This allows to call maybe_enable_record () from Lua session or DSP scripts
2017-01-01 18:26:21 -05:00
if (!rt_context) {
save_state ("", true);
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (_transport_speed) {
Remove unused punch+preroll API This API was not used, also superseded by record w/preroll.
2017-09-29 15:01:50 -04:00
if (!config.get_punch_in()) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
enable_record ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
} else {
fix delivery of MMC events under split process cycle conditions
2014-10-22 17:06:53 -04:00
send_immediate_mmc (MIDI::MachineControlCommand (MIDI::MachineControl::cmdRecordPause));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
RecordStateChanged (); /* EMIT SIGNAL */
}
set_dirty();
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
samplepos_t
Session::audible_sample (bool* latent_locate) const
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Add an API to query if the audible frame is latent pending a locate eg. at the end of a loop, the session may already be playing the beginning of the loop. The TransportLooped signal was emitted. Yet due to playback latency, the audible frame is still at the end of the loop. To interpolate the playhead position the UI needs to be able to know: Relying on the TransportLooped signal is not sufficient because it does not take playback latency into account.
2017-03-31 06:59:44 -04:00
if (latent_locate) {
*latent_locate = false;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
samplepos_t ret;
switch from JACK_Slave to Engine_Slave
2013-09-19 17:34:23 -04:00
if (synced_to_engine()) {
tweak implementation of Session::audible_frame() to be a little more defensible, and correct problems with record align/playhead position while disconnected from physical outputs
2014-10-12 23:00:14 -04:00
/* Note: this is basically just sync-to-JACK */
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
ret = _engine.transport_sample();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
} else {
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
ret = _transport_sample;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
assert (ret >= 0);
most of the 2.X->3.0 commit (up to rev 4299) except for gtk2_ardour/editor_canvas.cc; builds and runs and does a few specific things but expect it to be buggy for a while yet git-svn-id: svn://localhost/ardour2/branches/3.0@4313 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-12-12 09:43:24 -05:00
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
if (!transport_rolling()) {
return ret;
}
most of the 2.X->3.0 commit (up to rev 4299) except for gtk2_ardour/editor_canvas.cc; builds and runs and does a few specific things but expect it to be buggy for a while yet git-svn-id: svn://localhost/ardour2/branches/3.0@4313 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-12-12 09:43:24 -05:00
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
#if 0 // TODO looping
if (_transport_speed > 0.0f) {
if (play_loop && have_looped) {
/* the play-position wrapped at the loop-point
* ardour is already playing the beginning of the loop,
* but due to playback latency, the "audible frame"
* is still at the end of the loop.
*/
Location *location = _locations->auto_loop_location();
sampleoffset_t lo = location->start() - ret;
if (lo > 0) {
ret = location->end () - lo;
if (latent_locate) {
*latent_locate = true;
another attempt to fix playhead position with varispeed.
2016-07-25 13:46:19 -04:00
}
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
}
most of the 2.X->3.0 commit (up to rev 4299) except for gtk2_ardour/editor_canvas.cc; builds and runs and does a few specific things but expect it to be buggy for a while yet git-svn-id: svn://localhost/ardour2/branches/3.0@4313 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-12-12 09:43:24 -05:00
}
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
} else if (_transport_speed < 0.0f) {
/* XXX wot? no backward looping? */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
#endif
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
return std::max ((samplepos_t)0, ret);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
samplecnt_t
Session::preroll_samples (samplepos_t pos) const
Move preroll duration calc into libardour
2017-01-19 06:39:15 -05:00
{
const float pr = Config->get_preroll_seconds();
if (pos >= 0 && pr < 0) {
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
const Tempo& tempo = _tempo_map->tempo_at_sample (pos);
const Meter& meter = _tempo_map->meter_at_sample (pos);
return meter.samples_per_bar (tempo, sample_rate()) * -pr;
Move preroll duration calc into libardour
2017-01-19 06:39:15 -05:00
}
if (pr < 0) {
return 0;
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
return pr * sample_rate();
Move preroll duration calc into libardour
2017-01-19 06:39:15 -05:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::set_sample_rate (samplecnt_t frames_per_second)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
/** \fn void Session::set_sample_size(samplecnt_t)
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
the AudioEngine object that calls this guarantees
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
that it will not be called while we are also in
::process(). Its fine to do things that block
here.
*/
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
if (_base_sample_rate == 0) {
_base_sample_rate = frames_per_second;
never ever change sample-rate saved with the session after creation.
2016-04-18 08:50:09 -04:00
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
else if (_base_sample_rate != frames_per_second && frames_per_second != _nominal_sample_rate) {
NotifyAboutSampleRateMismatch (_base_sample_rate, frames_per_second);
add notification if Engine Rate changes and mismatches session's SR
2016-04-18 09:16:27 -04:00
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
_nominal_sample_rate = frames_per_second;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
sync_time_vars();
clear_clicks ();
after changing session SR, reset all write sources to use new rate
2015-05-12 10:16:26 -04:00
reset_write_sources (false);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
redesign of declicking and fades around loop boundaries
2019-11-23 01:41:56 -05:00
DiskReader::alloc_loop_declick (nominal_sample_rate());
Location* loc = _locations->auto_loop_location ();
DiskReader::reset_loop_declick (loc, nominal_sample_rate());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
// XXX we need some equivalent to this, somehow
// SndFileSource::setup_standard_crossfades (frames_per_second);
set_dirty();
/* XXX need to reset/reinstantiate all LADSPA plugins */
}
void
Remove all use of nframes_t. git-svn-id: svn://localhost/ardour2/branches/3.0@8166 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-03 17:26:29 -05:00
Session::set_block_size (pframes_t nframes)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
/* the AudioEngine guarantees
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
that it will not be called while we are also in
::process(). It is therefore fine to do things that block
here.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
{
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
current_block_size = nframes;
substantive change: use the JACK wait API and provide "thread buffers" separately from session in preparation for parallelization. lots of debug output at present. If using JACK1, requires a very current version of JACK1 SVN (0.119.0) git-svn-id: svn://localhost/ardour2/branches/3.0@6888 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-13 16:48:33 -04:00
ensure_buffers ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
boost::shared_ptr<RouteList> r = routes.reader ();
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
(*i)->set_block_size (nframes);
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
boost::shared_ptr<RouteList> rl = routes.reader ();
for (RouteList::iterator i = rl->begin(); i != rl->end(); ++i) {
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
if (tr) {
tr->set_block_size (nframes);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
set_worst_io_latencies ();
}
}
static void
Fix solo logic, connection/feed-information Assume the following connections Audio -> Bus1 --aux-send--> Bus2 Prior to this commit, Audio -> Bus2 was marked as "direct feed", even though the connection was only indirect via an aux-send. This caused issues with implicit solo and could lead to stuck solo.
2019-05-08 13:51:20 -04:00
trace_terminal (boost::shared_ptr<Route> r1, boost::shared_ptr<Route> rbase, bool sends_only)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
boost::shared_ptr<Route> r2;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
if (r1->feeds (rbase) && rbase->feeds (r1)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
info << string_compose(_("feedback loop setup between %1 and %2"), r1->name(), rbase->name()) << endmsg;
return;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* make a copy of the existing list of routes that feed r1 */
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
Route::FedBy existing (r1->fed_by());
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* for each route that feeds r1, recurse, marking it as feeding
rbase as well.
*/
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
for (Route::FedBy::iterator i = existing.begin(); i != existing.end(); ++i) {
if (!(r2 = i->r.lock ())) {
switch Route::fed_by from shared_ptr to weak_ptr and thus begin to allow Route destructors to be called git-svn-id: svn://localhost/ardour2/branches/3.0@6189 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 12:29:27 -05:00
/* (*i) went away, ignore it */
continue;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* r2 is a route that feeds r1 which somehow feeds base. mark
base as being fed by r2
*/
Fix solo logic, connection/feed-information Assume the following connections Audio -> Bus1 --aux-send--> Bus2 Prior to this commit, Audio -> Bus2 was marked as "direct feed", even though the connection was only indirect via an aux-send. This caused issues with implicit solo and could lead to stuck solo.
2019-05-08 13:51:20 -04:00
rbase->add_fed_by (r2, i->sends_only || sends_only);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (r2 != rbase) {
/* 2nd level feedback loop detection. if r1 feeds or is fed by r2,
stop here.
switch Route::fed_by from shared_ptr to weak_ptr and thus begin to allow Route destructors to be called git-svn-id: svn://localhost/ardour2/branches/3.0@6189 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 12:29:27 -05:00
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
if (r1->feeds (r2) && r2->feeds (r1)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
continue;
}
/* now recurse, so that we can mark base as being fed by
all routes that feed r2
*/
Fix solo logic, connection/feed-information Assume the following connections Audio -> Bus1 --aux-send--> Bus2 Prior to this commit, Audio -> Bus2 was marked as "direct feed", even though the connection was only indirect via an aux-send. This caused issues with implicit solo and could lead to stuck solo.
2019-05-08 13:51:20 -04:00
trace_terminal (r2, rbase, i->sends_only || sends_only);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
}
void
Session::resort_routes ()
{
/* don't do anything here with signals emitted
Don't resort routes at every step during a session's construction; speeds up loading of big sessions. git-svn-id: svn://localhost/ardour2/branches/3.0@10308 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-10-26 11:14:53 -04:00
by Routes during initial setup or while we
are being destroyed.
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
*/
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (inital_connect_or_deletion_in_progress ()) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
return;
}
fix deadlock when removing subgroup bus w/sends. ...and potentially other route removals that involve processor changes.
2015-10-15 07:49:35 -04:00
if (_route_deletion_in_progress) {
return;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
RCUWriter<RouteList> writer (routes);
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
boost::shared_ptr<RouteList> r = writer.get_copy ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
resort_routes_using (r);
/* writer goes out of scope and forces update */
}
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
#ifndef NDEBUG
reduce debug-build overhead.
2016-04-24 11:35:21 -04:00
if (DEBUG_ENABLED(DEBUG::Graph)) {
boost::shared_ptr<RouteList> rl = routes.reader ();
for (RouteList::iterator i = rl->begin(); i != rl->end(); ++i) {
DEBUG_TRACE (DEBUG::Graph, string_compose ("%1 fed by ...\n", (*i)->name()));
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
reduce debug-build overhead.
2016-04-24 11:35:21 -04:00
const Route::FedBy& fb ((*i)->fed_by());
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
reduce debug-build overhead.
2016-04-24 11:35:21 -04:00
for (Route::FedBy::const_iterator f = fb.begin(); f != fb.end(); ++f) {
boost::shared_ptr<Route> sf = f->r.lock();
if (sf) {
DEBUG_TRACE (DEBUG::Graph, string_compose ("\t%1 (sends only ? %2)\n", sf->name(), f->sends_only));
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
}
}
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
#endif
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Do a topological sort of the route list before passing it to the graph, as the graph's feedback detection algorithm depends on the input route list being sorted in such a way. Fixes #3924. git-svn-id: svn://localhost/ardour2/branches/3.0@10471 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-07 11:01:46 -05:00
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
/** This is called whenever we need to rebuild the graph of how we will process
* routes.
* @param r List of routes, in any order.
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
Session::resort_routes_using (boost::shared_ptr<RouteList> r)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
/* We are going to build a directed graph of our routes;
this is where the edges of that graph are put.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
GraphEdges edges;
/* Go through all routes doing two things:
*
* 1. Collect the edges of the route graph. Each of these edges
* is a pair of routes, one of which directly feeds the other
* either by a JACK connection or by an internal send.
*
* 2. Begin the process of making routes aware of which other
* routes directly or indirectly feed them. This information
* is used by the solo code.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Do a topological sort of the route list before passing it to the graph, as the graph's feedback detection algorithm depends on the input route list being sorted in such a way. Fixes #3924. git-svn-id: svn://localhost/ardour2/branches/3.0@10471 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-07 11:01:46 -05:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
/* Clear out the route's list of direct or indirect feeds */
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
(*i)->clear_fed_by ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Do a topological sort of the route list before passing it to the graph, as the graph's feedback detection algorithm depends on the input route list being sorted in such a way. Fixes #3924. git-svn-id: svn://localhost/ardour2/branches/3.0@10471 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-07 11:01:46 -05:00
for (RouteList::iterator j = r->begin(); j != r->end(); ++j) {
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
bool via_sends_only;
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
/* See if this *j feeds *i according to the current state of the JACK
connections and internal sends.
*/
if ((*j)->direct_feeds_according_to_reality (*i, &via_sends_only)) {
/* add the edge to the graph (part #1) */
edges.add (*j, *i, via_sends_only);
/* tell the route (for part #2) */
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
(*i)->add_fed_by (*j, via_sends_only);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
/* Attempt a topological sort of the route graph */
Fix name and a comment. git-svn-id: svn://localhost/ardour2/branches/3.0@10483 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-07 17:39:16 -05:00
boost::shared_ptr<RouteList> sorted_routes = topological_sort (r, edges);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Revert "option to ignore feedback detection - /i know what i'm doing/ mode." This reverts commit e17f75b351356279778e4ea4adbb6c0d544ba755.
2013-04-15 10:02:57 -04:00
if (sorted_routes) {
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
/* We got a satisfactory topological sort, so there is no feedback;
use this new graph.
Note: the process graph rechain does not require a
topologically-sorted list, but hey ho.
*/
For now, only use the multi-threaded process code if we are using >1 processor for DSP; this involves making the DSP use setting only take effect on a restart of Ardour. git-svn-id: svn://localhost/ardour2/branches/3.0@11302 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-22 07:28:49 -05:00
if (_process_graph) {
_process_graph->rechain (sorted_routes, edges);
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
_current_route_graph = edges;
/* Complete the building of the routes' lists of what directly
or indirectly feeds them.
*/
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
Fix solo logic, connection/feed-information Assume the following connections Audio -> Bus1 --aux-send--> Bus2 Prior to this commit, Audio -> Bus2 was marked as "direct feed", even though the connection was only indirect via an aux-send. This caused issues with implicit solo and could lead to stuck solo.
2019-05-08 13:51:20 -04:00
trace_terminal (*i, *i, false);
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
}
Write the topologically sorted route list back to the session's list so that the 1-processor process code sees it. Should fix #4937. git-svn-id: svn://localhost/ardour2/branches/3.0@12784 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-19 11:51:56 -04:00
*r = *sorted_routes;
move route_graph->rechain( r ); into Session::resort_routes_using git-svn-id: svn://localhost/ardour2/branches/3.0@7227 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-03 16:38:01 -04:00
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
#ifndef NDEBUG
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
DEBUG_TRACE (DEBUG::Graph, "Routes resorted, order follows:\n");
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
simplify PresentationInfo concept of order so that it is always global Cases that need more complex sort orders (e.g. all routes, then all vcas then master) need to take of this themselves
2016-06-03 15:14:23 -04:00
DEBUG_TRACE (DEBUG::Graph, string_compose ("\t%1 presentation order %2\n", (*i)->name(), (*i)->presentation_info().order()));
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
#endif
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Change the feedback alert to a flashing button; works around part of #4432. git-svn-id: svn://localhost/ardour2/branches/3.0@10539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-11 08:52:27 -05:00
SuccessfulGraphSort (); /* EMIT SIGNAL */
Revert "option to ignore feedback detection - /i know what i'm doing/ mode." This reverts commit e17f75b351356279778e4ea4adbb6c0d544ba755.
2013-04-15 10:02:57 -04:00
} else {
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
/* The topological sort failed, so we have a problem. Tell everyone
and stick to the old graph; this will continue to be processed, so
until the feedback is fixed, what is played back will not quite
reflect what is actually connected. Note also that we do not
do trace_terminal here, as it would fail due to an endless recursion,
so the solo code will think that everything is still connected
as it was before.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Alert the user if a connection is made which causes feedback, and preserve the route graph in the state that it was in before the feedback was introduced. The intent being to simplify the code, reduce the number of areas of code which must consider feedback, and fix a few bugs. git-svn-id: svn://localhost/ardour2/branches/3.0@10510 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-09 12:44:39 -05:00
FeedbackDetected (); /* EMIT SIGNAL */
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Allow naming of new tracks/busses in the add route dialogue (#3376). git-svn-id: svn://localhost/ardour2/branches/3.0@8976 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-26 20:59:04 -05:00
/** Find a route name starting with \a base, maybe followed by the
* lowest \a id. \a id will always be added if \a definitely_add_number
* is true on entry; otherwise it will only be added if required
* to make the name unique.
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
*
Allow naming of new tracks/busses in the add route dialogue (#3376). git-svn-id: svn://localhost/ardour2/branches/3.0@8976 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-26 20:59:04 -05:00
* Names are constructed like e.g. "Audio 3" for base="Audio" and id=3.
* The available route name with the lowest ID will be used, and \a id
* will be set to the ID.
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
*
Allow naming of new tracks/busses in the add route dialogue (#3376). git-svn-id: svn://localhost/ardour2/branches/3.0@8976 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-26 20:59:04 -05:00
* \return false if a route name could not be found, and \a track_name
* and \a id do not reflect a free route name.
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
*/
bool
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
Session::find_route_name (string const & base, uint32_t& id, string& name, bool definitely_add_number)
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
{
use PortManager::port_name_prefix_is_unique to check for new route names
2015-09-10 16:27:14 -04:00
/* the base may conflict with ports that do not belong to existing
routes, but hidden objects like the click track. So check port names
before anything else.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Check for reserved i/o that are not routes, fixes #7171
2016-12-10 18:03:44 -05:00
for (map<string,bool>::const_iterator reserved = reserved_io_names.begin(); reserved != reserved_io_names.end(); ++reserved) {
if (base == reserved->first) {
don't force rename of reserved route/IO names if the name isn't taken yet
2015-11-07 22:15:59 -05:00
/* Check if this reserved name already exists, and if
so, disallow it without a numeric suffix.
*/
Check for reserved i/o that are not routes, fixes #7171
2016-12-10 18:03:44 -05:00
if (!reserved->second || route_by_name (reserved->first)) {
don't force rename of reserved route/IO names if the name isn't taken yet
2015-11-07 22:15:59 -05:00
definitely_add_number = true;
if (id < 1) {
id = 1;
}
use PortManager::port_name_prefix_is_unique to check for new route names
2015-09-10 16:27:14 -04:00
}
use an explicit black list of route names that require numbering. This replaces an attempt to check port names which fails for a number of reasons.
2015-09-14 10:47:21 -04:00
break;
}
use PortManager::port_name_prefix_is_unique to check for new route names
2015-09-10 16:27:14 -04:00
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make auto track naming make a bit more sense
2016-11-07 14:08:00 -05:00
/* if we have "base 1" already, it doesn't make sense to add "base"
* if "base 1" has been deleted, adding "base" is no worse than "base 1"
*/
if (!definitely_add_number && route_by_name (base) == 0 && (route_by_name (string_compose("%1 1", base)) == 0)) {
/* just use the base */
use an explicit black list of route names that require numbering. This replaces an attempt to check port names which fails for a number of reasons.
2015-09-14 10:47:21 -04:00
name = base;
Allow naming of new tracks/busses in the add route dialogue (#3376). git-svn-id: svn://localhost/ardour2/branches/3.0@8976 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-26 20:59:04 -05:00
return true;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
do {
use an explicit black list of route names that require numbering. This replaces an attempt to check port names which fails for a number of reasons.
2015-09-14 10:47:21 -04:00
name = string_compose ("%1 %2", base, id);
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
if (route_by_name (name) == 0) {
return true;
}
++id;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
} while (id < (UINT_MAX-1));
return false;
}
fix all manner of things relating to io connections, setting capture alignment, and so on. still needs more tests of actual precise placement of newly recorded material git-svn-id: svn://localhost/ardour2/branches/3.0@9155 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-15 15:32:21 -04:00
/** Count the total ins and outs of all non-hidden tracks in the session and return them in in and out */
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
void
fix all manner of things relating to io connections, setting capture alignment, and so on. still needs more tests of actual precise placement of newly recorded material git-svn-id: svn://localhost/ardour2/branches/3.0@9155 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-15 15:32:21 -04:00
Session::count_existing_track_channels (ChanCount& in, ChanCount& out)
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
{
in = ChanCount::ZERO;
out = ChanCount::ZERO;
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
boost::shared_ptr<RouteList> r = routes.reader ();
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
if (!tr) {
continue;
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
}
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
assert (!tr->is_auditioner()); // XXX remove me
in += tr->n_inputs();
out += tr->n_outputs();
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
}
}
new Session::default_track_name_pattern (DataType) method, based on an idea in Tracks
2015-05-19 15:54:52 -04:00
string
Session::default_track_name_pattern (DataType t)
{
switch (t) {
case DataType::AUDIO:
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
return _("Audio ");
new Session::default_track_name_pattern (DataType) method, based on an idea in Tracks
2015-05-19 15:54:52 -04:00
break;
case DataType::MIDI:
return _("MIDI ");
}
return "";
}
Allow naming of new tracks/busses in the add route dialogue (#3376). git-svn-id: svn://localhost/ardour2/branches/3.0@8976 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-26 20:59:04 -05:00
/** Caller must not hold process lock
Fix naming / numbering of new MIDI tracks (#4453). git-svn-id: svn://localhost/ardour2/branches/3.0@10487 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-07 19:31:17 -05:00
* @param name_template string to use for the start of the name, or "" to use "MIDI".
add a way to create a new MIDI track with an instrument pre-selected git-svn-id: svn://localhost/ardour2/branches/3.0@11421 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-02-01 23:12:23 -05:00
* @param instrument plugin info for the instrument to insert pre-fader, if any
Allow naming of new tracks/busses in the add route dialogue (#3376). git-svn-id: svn://localhost/ardour2/branches/3.0@8976 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-26 20:59:04 -05:00
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
list<boost::shared_ptr<MidiTrack> >
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
Session::new_midi_track (const ChanCount& input, const ChanCount& output, bool strict_io,
manually fix rebase/merge issues caused by addition of plugin presets to methods which add a new track/bus
2016-05-17 08:21:05 -04:00
boost::shared_ptr<PluginInfo> instrument, Plugin::PresetRecord* pset,
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
RouteGroup* route_group, uint32_t how_many,
string name_template, PresentationInfo::order_t order,
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
TrackMode mode)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
string track_name;
Fix adding multiple tracks (regression from r6711). git-svn-id: svn://localhost/ardour2/branches/3.0@6712 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-26 18:35:58 -05:00
uint32_t track_id = 0;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
string port;
RouteList new_routes;
list<boost::shared_ptr<MidiTrack> > ret;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
new Session::default_track_name_pattern (DataType) method, based on an idea in Tracks
2015-05-19 15:54:52 -04:00
const string name_pattern = default_track_name_pattern (DataType::MIDI);
bool const use_number = (how_many != 1) || name_template.empty () || (name_template == name_pattern);
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
while (how_many) {
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
if (!find_route_name (name_template.empty() ? _("MIDI") : name_template, ++track_id, track_name, use_number)) {
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
error << "cannot find name for new midi track" << endmsg;
goto failed;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
boost::shared_ptr<MidiTrack> track;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
try {
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
track.reset (new MidiTrack (*this, track_name, mode));
would you look at that! all those changes just to make the auditioner work again .... git-svn-id: svn://localhost/ardour2/branches/3.0@6794 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-24 23:40:07 -04:00
Use commented out (correct) Mute/SoloControllable set_value() code; requires adding enable_shared_from_this to Route. git-svn-id: svn://localhost/ardour2/branches/3.0@8930 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-22 16:15:42 -05:00
if (track->init ()) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
goto failed;
}
would you look at that! all those changes just to make the auditioner work again .... git-svn-id: svn://localhost/ardour2/branches/3.0@6794 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-24 23:40:07 -04:00
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
if (strict_io) {
special case mixbus routing
2016-04-15 08:52:19 -04:00
track->set_strict_io (true);
}
make boost shared_ptr debugging a little easier to turn on/off
2016-05-07 13:32:31 -04:00
BOOST_MARK_TRACK (track);
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
{
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
provide an easier way to add mixed (audio+MIDI) tracks to a session, to facilitate the use of plugins like Reaktor which accept both audio & MIDI data. there's probably still a paradigm shift waiting to happen here but its likely post-3.0. git-svn-id: svn://localhost/ardour2/branches/3.0@12816 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-21 21:45:16 -04:00
if (track->input()->ensure_io (input, false, this)) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
error << "cannot configure " << input << " out configuration for new midi track" << endmsg;
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
goto failed;
}
the big Route structure refactor. !!!! THIS WILL ***NOT LOAD*** PRIOR 3.0 or 2.X SESSIONS !!!! BREAKAGE IS EXPECTED !!!! IF YOU HAVE AND NEED A WORKING 3.0 DO **NOT** UPDATE. !!!! otherwise, update and enjoy the steadily emerging joys of this major reworking of ardour internals git-svn-id: svn://localhost/ardour2/branches/3.0@5137 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-09 16:21:19 -04:00
provide an easier way to add mixed (audio+MIDI) tracks to a session, to facilitate the use of plugins like Reaktor which accept both audio & MIDI data. there's probably still a paradigm shift waiting to happen here but its likely post-3.0. git-svn-id: svn://localhost/ardour2/branches/3.0@12816 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-21 21:45:16 -04:00
if (track->output()->ensure_io (output, false, this)) {
error << "cannot configure " << output << " out configuration for new midi track" << endmsg;
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
goto failed;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
fix recently-added crash when adding new tracks+busses git-svn-id: svn://localhost/ardour2/branches/3.0@6339 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-10 13:33:54 -05:00
if (route_group) {
route_group->add (track);
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
new_routes.push_back (track);
ret.push_back (track);
}
catch (failed_constructor &err) {
error << _("Session: could not create new midi track.") << endmsg;
goto failed;
}
catch (AudioEngine::PortRegistrationFailure& pfe) {
A few more typo fixes in libardour. git-svn-id: svn://localhost/ardour2/branches/3.0@13813 d708f5d6-7413-0410-9779-e7cbd77b26cf
2013-01-09 10:05:04 -05:00
error << string_compose (_("No more JACK ports are available. You will need to stop %1 and restart JACK with more ports if you need this many tracks."), PROGRAM_NAME) << endmsg;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
goto failed;
}
--how_many;
}
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
failed:
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (!new_routes.empty()) {
use state Session::StateProtector
2014-06-28 15:36:13 -04:00
StateProtector sp (this);
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
add_routes (new_routes, true, true, false, order);
add a way to create a new MIDI track with an instrument pre-selected git-svn-id: svn://localhost/ardour2/branches/3.0@11421 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-02-01 23:12:23 -05:00
if (instrument) {
for (RouteList::iterator r = new_routes.begin(); r != new_routes.end(); ++r) {
PluginPtr plugin = instrument->load (*this);
Fix crash when instrument instantiation fails on track creation.
2016-10-30 19:35:53 -04:00
if (!plugin) {
warning << "Failed to add Synth Plugin to newly created track." << endmsg;
continue;
}
add API to load Instrument Preset when creating a Midi track
2016-05-06 10:02:04 -04:00
if (pset) {
plugin->load_preset (*pset);
}
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
boost::shared_ptr<PluginInsert> pi (new PluginInsert (*this, plugin));
if (strict_io) {
pi->set_strict_io (true);
}
Instrument insert options: * allow to directly fan-out when adding a multi-channel instrument * Mixbus: move multi-channel instruments after Comp & EQ.
2017-01-20 18:32:15 -05:00
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
(*r)->add_processor (pi, PreFader);
Instrument insert options: * allow to directly fan-out when adding a multi-channel instrument * Mixbus: move multi-channel instruments after Comp & EQ.
2017-01-20 18:32:15 -05:00
if (Profile->get_mixbus () && pi->configured () && pi->output_streams().n_audio() > 2) {
(*r)->move_instrument_down (false);
}
add a way to create a new MIDI track with an instrument pre-selected git-svn-id: svn://localhost/ardour2/branches/3.0@11421 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-02-01 23:12:23 -05:00
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
return ret;
}
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
RouteList
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
Session::new_midi_route (RouteGroup* route_group, uint32_t how_many, string name_template, bool strict_io,
boost::shared_ptr<PluginInfo> instrument, Plugin::PresetRecord* pset,
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
PresentationInfo::Flag flag, PresentationInfo::order_t order)
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
{
string bus_name;
uint32_t bus_id = 0;
string port;
RouteList ret;
bool const use_number = (how_many != 1) || name_template.empty () || name_template == _("Midi Bus");
while (how_many) {
if (!find_route_name (name_template.empty () ? _("Midi Bus") : name_template, ++bus_id, bus_name, use_number)) {
error << "cannot find name for new midi bus" << endmsg;
goto failure;
}
clean up debug output for OrderKeys
2016-05-17 09:36:28 -04:00
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
try {
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
boost::shared_ptr<Route> bus (new Route (*this, bus_name, flag, DataType::AUDIO)); // XXX Editor::add_routes is not ready for ARDOUR::DataType::MIDI
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
if (bus->init ()) {
goto failure;
}
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
if (strict_io) {
special case mixbus routing
2016-04-15 08:52:19 -04:00
bus->set_strict_io (true);
}
make boost shared_ptr debugging a little easier to turn on/off
2016-05-07 13:32:31 -04:00
BOOST_MARK_ROUTE(bus);
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
{
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
if (bus->input()->ensure_io (ChanCount(DataType::MIDI, 1), false, this)) {
error << _("cannot configure new midi bus input") << endmsg;
goto failure;
}
if (bus->output()->ensure_io (ChanCount(DataType::MIDI, 1), false, this)) {
error << _("cannot configure new midi bus output") << endmsg;
goto failure;
}
}
if (route_group) {
route_group->add (bus);
}
Make MIDI busses possible targets of internal sends The session only added an internal return to new audio busses, and so only those were proposed in the "New Aux Send" list. Also add the return to new midi busses, now that internal sends know how to deal with midi.
2016-07-08 20:17:00 -04:00
bus->add_internal_return ();
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
ret.push_back (bus);
}
catch (failed_constructor &err) {
error << _("Session: could not create new audio route.") << endmsg;
goto failure;
}
catch (AudioEngine::PortRegistrationFailure& pfe) {
error << pfe.what() << endmsg;
goto failure;
}
--how_many;
}
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
failure:
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
if (!ret.empty()) {
StateProtector sp (this);
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
add_routes (ret, false, false, false, order);
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
if (instrument) {
for (RouteList::iterator r = ret.begin(); r != ret.end(); ++r) {
PluginPtr plugin = instrument->load (*this);
Fix crash when instrument instantiation fails on track creation.
2016-10-30 19:35:53 -04:00
if (!plugin) {
warning << "Failed to add Synth Plugin to newly created track." << endmsg;
continue;
}
add API to load Instrument Preset when creating a Midi track
2016-05-06 10:02:04 -04:00
if (pset) {
plugin->load_preset (*pset);
}
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
boost::shared_ptr<PluginInsert> pi (new PluginInsert (*this, plugin));
if (strict_io) {
pi->set_strict_io (true);
}
Instrument insert options: * allow to directly fan-out when adding a multi-channel instrument * Mixbus: move multi-channel instruments after Comp & EQ.
2017-01-20 18:32:15 -05:00
Instrument channel option when adding track By setting strict-io on the Instrument plugin early during track creation, adding the plugin will trigger a PluginSetup Dialog for multi-out instruments in a strict i/o track.
2017-01-20 15:46:47 -05:00
(*r)->add_processor (pi, PreFader);
Instrument insert options: * allow to directly fan-out when adding a multi-channel instrument * Mixbus: move multi-channel instruments after Comp & EQ.
2017-01-20 18:32:15 -05:00
if (Profile->get_mixbus () && pi->configured () && pi->output_streams().n_audio() > 2) {
(*r)->move_instrument_down (false);
}
Midi Busses? why yes!
2016-03-15 23:41:19 -04:00
}
}
}
return ret;
}
fixup auto-connection of new MIDI tracks and MIDI tracks with newly-added audio ports git-svn-id: svn://localhost/ardour2/branches/3.0@9170 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-20 18:09:46 -04:00
void
Session::midi_output_change_handler (IOChange change, void * /*src*/, boost::weak_ptr<Route> wmt)
{
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
boost::shared_ptr<Route> midi_track (wmt.lock());
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
if (!midi_track) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
return;
}
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
if ((change.type & IOChange::ConfigurationChanged) && Config->get_output_auto_connect() != ManualConnect) {
fixup auto-connection of new MIDI tracks and MIDI tracks with newly-added audio ports git-svn-id: svn://localhost/ardour2/branches/3.0@9170 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-20 18:09:46 -04:00
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
if (change.after.n_audio() <= change.before.n_audio()) {
return;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
/* new audio ports: make sure the audio goes somewhere useful,
* unless the user has no-auto-connect selected.
*
* The existing ChanCounts don't matter for this call as they are only
* to do with matching input and output indices, and we are only changing
* outputs here.
*/
auto_connect_route (midi_track, false, ChanCount(), change.before);
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Add method to ensure Stripable sort constrains (for UI use)
2017-06-16 20:31:03 -04:00
bool
Session::ensure_stripable_sort_order ()
{
StripableList sl;
get_stripables (sl);
sl.sort (Stripable::Sorter ());
bool change = false;
PresentationInfo::order_t order = 0;
for (StripableList::iterator si = sl.begin(); si != sl.end(); ++si) {
boost::shared_ptr<Stripable> s (*si);
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
assert (!s->is_auditioner ()); // XXX remove me
if (s->is_monitor ()) {
Add method to ensure Stripable sort constrains (for UI use)
2017-06-16 20:31:03 -04:00
continue;
}
if (order != s->presentation_info().order()) {
s->set_presentation_order (order);
change = true;
}
++order;
}
return change;
}
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
void
change name of a Session method to makes its intended function clear
2016-06-02 11:47:19 -04:00
Session::ensure_route_presentation_info_gap (PresentationInfo::order_t first_new_order, uint32_t how_many)
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
{
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
DEBUG_TRACE (DEBUG::OrderKeys, string_compose ("ensure order gap starting at %1 for %2\n", first_new_order, how_many));
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
if (first_new_order == PresentationInfo::max_order) {
/* adding at end, no worries */
return;
}
change name of a Session method to makes its intended function clear
2016-06-02 11:47:19 -04:00
/* create a gap in the presentation info to accomodate @param how_many
* new objects.
*/
fix behaviour when opening up a gap to insert new Stripables. Need to include all stripables, not just routes, and need to include master since editor (and maybe others) opt to use the presentation_info().order() value for it
2016-07-03 13:22:47 -04:00
StripableList sl;
get_stripables (sl);
change name of a Session method to makes its intended function clear
2016-06-02 11:47:19 -04:00
fix behaviour when opening up a gap to insert new Stripables. Need to include all stripables, not just routes, and need to include master since editor (and maybe others) opt to use the presentation_info().order() value for it
2016-07-03 13:22:47 -04:00
for (StripableList::iterator si = sl.begin(); si != sl.end(); ++si) {
boost::shared_ptr<Stripable> s (*si);
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
if (s->presentation_info().special (false)) {
continue;
}
if (!s->presentation_info().order_set()) {
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
continue;
}
fix behaviour when opening up a gap to insert new Stripables. Need to include all stripables, not just routes, and need to include master since editor (and maybe others) opt to use the presentation_info().order() value for it
2016-07-03 13:22:47 -04:00
if (s->presentation_info().order () >= first_new_order) {
s->set_presentation_order (s->presentation_info().order () + how_many);
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
}
}
}
Allow naming of new tracks/busses in the add route dialogue (#3376). git-svn-id: svn://localhost/ardour2/branches/3.0@8976 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-26 20:59:04 -05:00
/** Caller must not hold process lock
* @param name_template string to use for the start of the name, or "" to use "Audio".
*/
Fix adding multiple tracks (regression from r6711). git-svn-id: svn://localhost/ardour2/branches/3.0@6712 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-26 18:35:58 -05:00
list< boost::shared_ptr<AudioTrack> >
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Session::new_audio_track (int input_channels, int output_channels, RouteGroup* route_group,
uint32_t how_many, string name_template, PresentationInfo::order_t order,
TrackMode mode)
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
{
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
string track_name;
Fix adding multiple tracks (regression from r6711). git-svn-id: svn://localhost/ardour2/branches/3.0@6712 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-26 18:35:58 -05:00
uint32_t track_id = 0;
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
string port;
RouteList new_routes;
list<boost::shared_ptr<AudioTrack> > ret;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
new Session::default_track_name_pattern (DataType) method, based on an idea in Tracks
2015-05-19 15:54:52 -04:00
const string name_pattern = default_track_name_pattern (DataType::AUDIO);
bool const use_number = (how_many != 1) || name_template.empty () || (name_template == name_pattern);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
while (how_many) {
Tracks calls tracks "Tracks"; Ardour uses type-names
2015-05-08 15:29:19 -04:00
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
if (!find_route_name (name_template.empty() ? _(name_pattern.c_str()) : name_template, ++track_id, track_name, use_number)) {
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
error << "cannot find name for new audio track" << endmsg;
goto failed;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
boost::shared_ptr<AudioTrack> track;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
try {
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
track.reset (new AudioTrack (*this, track_name, mode));
would you look at that! all those changes just to make the auditioner work again .... git-svn-id: svn://localhost/ardour2/branches/3.0@6794 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-24 23:40:07 -04:00
Use commented out (correct) Mute/SoloControllable set_value() code; requires adding enable_shared_from_this to Route. git-svn-id: svn://localhost/ardour2/branches/3.0@8930 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-22 16:15:42 -05:00
if (track->init ()) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
goto failed;
}
would you look at that! all those changes just to make the auditioner work again .... git-svn-id: svn://localhost/ardour2/branches/3.0@6794 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-24 23:40:07 -04:00
special case mixbus routing
2016-04-15 08:52:19 -04:00
if (Profile->get_mixbus ()) {
track->set_strict_io (true);
}
make boost shared_ptr debugging a little easier to turn on/off
2016-05-07 13:32:31 -04:00
BOOST_MARK_TRACK (track);
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
{
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
if (track->input()->ensure_io (ChanCount(DataType::AUDIO, input_channels), false, this)) {
error << string_compose (
_("cannot configure %1 in/%2 out configuration for new audio track"),
input_channels, output_channels)
<< endmsg;
goto failed;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
if (track->output()->ensure_io (ChanCount(DataType::AUDIO, output_channels), false, this)) {
error << string_compose (
_("cannot configure %1 in/%2 out configuration for new audio track"),
input_channels, output_channels)
<< endmsg;
goto failed;
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
fix recently-added crash when adding new tracks+busses git-svn-id: svn://localhost/ardour2/branches/3.0@6339 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-10 13:33:54 -05:00
if (route_group) {
route_group->add (track);
}
Allow edit group creation from the route UI's menu. Allow new tracks to be put in particular edit groups. git-svn-id: svn://localhost/ardour2/branches/3.0@5226 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-20 11:40:26 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
new_routes.push_back (track);
ret.push_back (track);
}
catch (failed_constructor &err) {
error << _("Session: could not create new audio track.") << endmsg;
goto failed;
}
catch (AudioEngine::PortRegistrationFailure& pfe) {
first half of the mega-by-hand-commit-by-commit merge from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4980 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-15 14:04:23 -04:00
error << pfe.what() << endmsg;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
goto failed;
}
--how_many;
}
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
failed:
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (!new_routes.empty()) {
use state Session::StateProtector
2014-06-28 15:36:13 -04:00
StateProtector sp (this);
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
add_routes (new_routes, true, true, false, order);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
return ret;
}
Allow naming of new tracks/busses in the add route dialogue (#3376). git-svn-id: svn://localhost/ardour2/branches/3.0@8976 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-26 20:59:04 -05:00
/** Caller must not hold process lock.
* @param name_template string to use for the start of the name, or "" to use "Bus".
*/
first pass at internal sends. this is a very tentative work in progress, and it is possible that major changes may follow in the near future. it is certainly not complete, but the fundamental changes to Port/Buffer operation merit a commit at this point git-svn-id: svn://localhost/ardour2/branches/3.0@4464 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-01-30 02:40:13 -05:00
RouteList
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Session::new_audio_route (int input_channels, int output_channels, RouteGroup* route_group, uint32_t how_many, string name_template,
PresentationInfo::Flag flags, PresentationInfo::order_t order)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
string bus_name;
Fix adding multiple tracks (regression from r6711). git-svn-id: svn://localhost/ardour2/branches/3.0@6712 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-26 18:35:58 -05:00
uint32_t bus_id = 0;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
string port;
RouteList ret;
Always use numeric prefixes for default-named tracks / busses (#4123). git-svn-id: svn://localhost/ardour2/branches/3.0@9762 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-22 19:37:02 -04:00
bool const use_number = (how_many != 1) || name_template.empty () || name_template == _("Bus");
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
while (how_many) {
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
if (!find_route_name (name_template.empty () ? _("Bus") : name_template, ++bus_id, bus_name, use_number)) {
Factor out massive code duplication in route creation stuff. Fix auto-connect logic for multiply-typed routes. git-svn-id: svn://localhost/ardour2/branches/3.0@6711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-23 17:45:07 -05:00
error << "cannot find name for new audio bus" << endmsg;
goto failure;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
try {
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
boost::shared_ptr<Route> bus (new Route (*this, bus_name, flags, DataType::AUDIO));
would you look at that! all those changes just to make the auditioner work again .... git-svn-id: svn://localhost/ardour2/branches/3.0@6794 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-24 23:40:07 -04:00
Use commented out (correct) Mute/SoloControllable set_value() code; requires adding enable_shared_from_this to Route. git-svn-id: svn://localhost/ardour2/branches/3.0@8930 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-22 16:15:42 -05:00
if (bus->init ()) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
goto failure;
}
would you look at that! all those changes just to make the auditioner work again .... git-svn-id: svn://localhost/ardour2/branches/3.0@6794 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-24 23:40:07 -04:00
special case mixbus routing
2016-04-15 08:52:19 -04:00
if (Profile->get_mixbus ()) {
bus->set_strict_io (true);
}
make boost shared_ptr debugging a little easier to turn on/off
2016-05-07 13:32:31 -04:00
BOOST_MARK_ROUTE(bus);
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
{
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
the big Route structure refactor. !!!! THIS WILL ***NOT LOAD*** PRIOR 3.0 or 2.X SESSIONS !!!! BREAKAGE IS EXPECTED !!!! IF YOU HAVE AND NEED A WORKING 3.0 DO **NOT** UPDATE. !!!! otherwise, update and enjoy the steadily emerging joys of this major reworking of ardour internals git-svn-id: svn://localhost/ardour2/branches/3.0@5137 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-09 16:21:19 -04:00
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
if (bus->input()->ensure_io (ChanCount(DataType::AUDIO, input_channels), false, this)) {
error << string_compose (_("cannot configure %1 in/%2 out configuration for new audio track"),
input_channels, output_channels)
<< endmsg;
goto failure;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Prevent removal of route inputs when the plugins cannot be configured with the new number. Rework locking so that the process lock is held from the point that a route input is removed until after the processors are reconfigured; fixes #3548. git-svn-id: svn://localhost/ardour2/branches/3.0@8089 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-25 18:46:24 -05:00
if (bus->output()->ensure_io (ChanCount(DataType::AUDIO, output_channels), false, this)) {
error << string_compose (_("cannot configure %1 in/%2 out configuration for new audio track"),
input_channels, output_channels)
<< endmsg;
goto failure;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
fix recently-added crash when adding new tracks+busses git-svn-id: svn://localhost/ardour2/branches/3.0@6339 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-10 13:33:54 -05:00
if (route_group) {
route_group->add (bus);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Remove direct busses so that all busses have internal returns. git-svn-id: svn://localhost/ardour2/branches/3.0@8447 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-05 10:52:27 -05:00
bus->add_internal_return ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
ret.push_back (bus);
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
catch (failed_constructor &err) {
error << _("Session: could not create new audio route.") << endmsg;
goto failure;
}
catch (AudioEngine::PortRegistrationFailure& pfe) {
first half of the mega-by-hand-commit-by-commit merge from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4980 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-15 14:04:23 -04:00
error << pfe.what() << endmsg;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
goto failure;
}
--how_many;
}
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
failure:
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (!ret.empty()) {
use state Session::StateProtector
2014-06-28 15:36:13 -04:00
StateProtector sp (this);
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
if (flags == PresentationInfo::FoldbackBus) {
Add ListenBus to stripable types
2018-10-22 15:36:59 -04:00
add_routes (ret, false, false, true, order); // no autoconnect
adding routes no longer saves session but leaves it marked dirty (Tracks semantics for add_routes() retained)
2015-05-08 12:48:07 -04:00
} else {
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
add_routes (ret, false, true, true, order); // autoconnect // outputs only
adding routes no longer saves session but leaves it marked dirty (Tracks semantics for add_routes() retained)
2015-05-08 12:48:07 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
return ret;
}
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
RouteList
new libardour API for Session::new_route_from_template to allow specifying insertion point
2016-08-22 09:41:08 -04:00
Session::new_route_from_template (uint32_t how_many, PresentationInfo::order_t insert_at, const std::string& template_path, const std::string& name_base,
PlaylistDisposition pd)
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
{
XMLTree tree;
if (!tree.read (template_path.c_str())) {
substantial part of infrastructure required for track/bus duplication This includes removing the removal of ID values in XML, and its replacement with thread-local forcing of ID resets, implemented in a previous commit
2015-11-13 16:14:09 -05:00
return RouteList();
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
}
new libardour API for Session::new_route_from_template to allow specifying insertion point
2016-08-22 09:41:08 -04:00
return new_route_from_template (how_many, insert_at, *tree.root(), name_base, pd);
substantial part of infrastructure required for track/bus duplication This includes removing the removal of ID values in XML, and its replacement with thread-local forcing of ID resets, implemented in a previous commit
2015-11-13 16:14:09 -05:00
}
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
substantial part of infrastructure required for track/bus duplication This includes removing the removal of ID values in XML, and its replacement with thread-local forcing of ID resets, implemented in a previous commit
2015-11-13 16:14:09 -05:00
RouteList
new libardour API for Session::new_route_from_template to allow specifying insertion point
2016-08-22 09:41:08 -04:00
Session::new_route_from_template (uint32_t how_many, PresentationInfo::order_t insert_at, XMLNode& node, const std::string& name_base, PlaylistDisposition pd)
substantial part of infrastructure required for track/bus duplication This includes removing the removal of ID values in XML, and its replacement with thread-local forcing of ID resets, implemented in a previous commit
2015-11-13 16:14:09 -05:00
{
RouteList ret;
uint32_t number = 0;
const uint32_t being_added = how_many;
/* This will prevent the use of any existing XML-provided PBD::ID
values by Stateful.
*/
Stateful::ForceIDRegeneration force_ids;
make adding tracks/busses from templates just like adding them at the start of a session, in that IO connections are prevented until they are all added, and then we connect. this should fix the locking issue in #4643 git-svn-id: svn://localhost/ardour2/branches/3.0@11329 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-24 08:43:48 -05:00
IO::disable_connecting ();
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
while (how_many) {
substantial part of infrastructure required for track/bus duplication This includes removing the removal of ID values in XML, and its replacement with thread-local forcing of ID resets, implemented in a previous commit
2015-11-13 16:14:09 -05:00
/* We're going to modify the node contents a bit so take a
* copy. The node may be re-used when duplicating more than once.
*/
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
substantial part of infrastructure required for track/bus duplication This includes removing the removal of ID values in XML, and its replacement with thread-local forcing of ID resets, implemented in a previous commit
2015-11-13 16:14:09 -05:00
XMLNode node_copy (node);
working but still confusing fix for ensuring that playlist shared-with data is correct when duplicating tracks. Confusing because orig_track_id is still being switched to the NEW track, and shared_with refers to the OLD one
2019-03-21 00:16:57 -04:00
std::vector<boost::shared_ptr<Playlist> > shared_playlists;
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
Slightly unpleasant fix for creation of tracks from templates; it would be nice if we could set things up using the Route's logic for setting names of its children, rather than repeating the same logic in XML-land (#4303). git-svn-id: svn://localhost/ardour2/branches/3.0@10655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-16 18:03:59 -05:00
try {
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
string name;
fix #5424: routes created from templates do not get names based on the user-supplied text from the add route dialog
2013-04-01 20:45:57 -04:00
if (!name_base.empty()) {
/* if we're adding more than one routes, force
* all the names of the new routes to be
* numbered, via the final parameter.
*/
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
if (!find_route_name (name_base.c_str(), ++number, name, (being_added > 1))) {
fix #5424: routes created from templates do not get names based on the user-supplied text from the add route dialog
2013-04-01 20:45:57 -04:00
fatal << _("Session: UINT_MAX routes? impossible!") << endmsg;
Consistent use of abort() /* NOTREACHED */ This fixes some static analysis warnings: PBD::fatal transmitter needs to be connected to a function that aborts. This is usually the case with GUI
2019-09-17 23:57:26 -04:00
abort(); /*NOTREACHED*/
fix #5424: routes created from templates do not get names based on the user-supplied text from the add route dialog
2013-04-01 20:45:57 -04:00
}
} else {
string const route_name = node_copy.property(X_("name"))->value ();
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
fix #5424: routes created from templates do not get names based on the user-supplied text from the add route dialog
2013-04-01 20:45:57 -04:00
/* generate a new name by adding a number to the end of the template name */
superior fix for limits on track name length
2015-05-12 10:43:37 -04:00
if (!find_route_name (route_name.c_str(), ++number, name, true)) {
fix #5424: routes created from templates do not get names based on the user-supplied text from the add route dialog
2013-04-01 20:45:57 -04:00
fatal << _("Session: UINT_MAX routes? impossible!") << endmsg;
add abort() to non-reached code This cleans up a lot of false-positives in static analysis and also helps compilers to optimize code paths in general. (tagging the fatal stingstream operator as ‘noreturn’ is far less trivial)
2014-11-14 04:47:43 -05:00
abort(); /*NOTREACHED*/
fix #5424: routes created from templates do not get names based on the user-supplied text from the add route dialog
2013-04-01 20:45:57 -04:00
}
Reset both IO names rather than just one when setting up the use of a track template. git-svn-id: svn://localhost/ardour2/branches/3.0@7085 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-08 20:47:16 -04:00
}
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
/* figure out the appropriate playlist setup. The track
* (if the Route we're creating is a track) will find
* playlists via ID.
*/
Fix Duplicate Track with Copy Playlist option selected Previously two new empty playlists were created. Related to and mentioned in issue #7141
2016-11-27 21:30:38 -05:00
if (pd == CopyPlaylist) {
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
PBD::ID playlist_id;
if (node_copy.get_property (X_("audio-playlist"), playlist_id)) {
boost::shared_ptr<Playlist> playlist = _playlists->by_id (playlist_id);
playlist = PlaylistFactory::create (playlist, string_compose ("%1.1", name));
playlist->reset_shares ();
node_copy.set_property (X_("audio-playlist"), playlist->id());
}
if (node_copy.get_property (X_("midi-playlist"), playlist_id)) {
boost::shared_ptr<Playlist> playlist = _playlists->by_id (playlist_id);
Proposed fix managing shared playlists (see #7150) Actually, when duplicating a track with "share playlist", the current playlist is owned by the new created track(orig-track-id). The sharing mecanism is made by diskstreams pointing on the same(shared) playlist. Since playlist now owned by the new track, selecting another playlist in the original track "forgets" the playlist for this track.You can't select the shared playlist anymore from the original track. This commit adds a way to keep trace of shared playlist between tracks.
2016-11-30 11:02:20 -05:00
playlist = PlaylistFactory::create (playlist, string_compose ("%1.1", name));
playlist->reset_shares ();
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
node_copy.set_property (X_("midi-playlist"), playlist->id());
Proposed fix managing shared playlists (see #7150) Actually, when duplicating a track with "share playlist", the current playlist is owned by the new created track(orig-track-id). The sharing mecanism is made by diskstreams pointing on the same(shared) playlist. Since playlist now owned by the new track, selecting another playlist in the original track "forgets" the playlist for this track.You can't select the shared playlist anymore from the original track. This commit adds a way to keep trace of shared playlist between tracks.
2016-11-30 11:02:20 -05:00
}
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
Proposed fix managing shared playlists (see #7150) Actually, when duplicating a track with "share playlist", the current playlist is owned by the new created track(orig-track-id). The sharing mecanism is made by diskstreams pointing on the same(shared) playlist. Since playlist now owned by the new track, selecting another playlist in the original track "forgets" the playlist for this track.You can't select the shared playlist anymore from the original track. This commit adds a way to keep trace of shared playlist between tracks.
2016-11-30 11:02:20 -05:00
} else if (pd == SharePlaylist) {
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
PBD::ID playlist_id;
if (node_copy.get_property (X_("audio-playlist"), playlist_id)) {
boost::shared_ptr<Playlist> playlist = _playlists->by_id (playlist_id);
working but still confusing fix for ensuring that playlist shared-with data is correct when duplicating tracks. Confusing because orig_track_id is still being switched to the NEW track, and shared_with refers to the OLD one
2019-03-21 00:16:57 -04:00
shared_playlists.push_back (playlist);
Fix Duplicate Track with Copy Playlist option selected Previously two new empty playlists were created. Related to and mentioned in issue #7141
2016-11-27 21:30:38 -05:00
}
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
if (node_copy.get_property (X_("midi-playlist"), playlist_id)) {
boost::shared_ptr<Playlist> playlist = _playlists->by_id (playlist_id);
working but still confusing fix for ensuring that playlist shared-with data is correct when duplicating tracks. Confusing because orig_track_id is still being switched to the NEW track, and shared_with refers to the OLD one
2019-03-21 00:16:57 -04:00
shared_playlists.push_back (playlist);
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
}
} else { /* NewPlaylist */
PBD::ID pid;
if (node_copy.get_property (X_("audio-playlist"), pid)) {
boost::shared_ptr<Playlist> playlist = PlaylistFactory::create (DataType::AUDIO, *this, name, false);
node_copy.set_property (X_("audio-playlist"), playlist->id());
}
if (node_copy.get_property (X_("midi-playlist"), pid)) {
boost::shared_ptr<Playlist> playlist = PlaylistFactory::create (DataType::MIDI, *this, name, false);
node_copy.set_property (X_("midi-playlist"), playlist->id());
}
substantial part of infrastructure required for track/bus duplication This includes removing the removal of ID values in XML, and its replacement with thread-local forcing of ID resets, implemented in a previous commit
2015-11-13 16:14:09 -05:00
}
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
/* Fix up new name in the XML node */
Fix Duplicate Track with Copy Playlist option selected Previously two new empty playlists were created. Related to and mentioned in issue #7141
2016-11-27 21:30:38 -05:00
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
Route::set_name_in_state (node_copy, name);
make track templates work, including a fix for the MidiTrack XML constructor git-svn-id: svn://localhost/ardour2/branches/3.0@4735 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-05 12:27:05 -05:00
Slightly unpleasant fix for creation of tracks from templates; it would be nice if we could set things up using the Route's logic for setting names of its children, rather than repeating the same logic in XML-land (#4303). git-svn-id: svn://localhost/ardour2/branches/3.0@10655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-16 18:03:59 -05:00
/* trim bitslots from listen sends so that new ones are used */
XMLNodeList children = node_copy.children ();
for (XMLNodeList::iterator i = children.begin(); i != children.end(); ++i) {
if ((*i)->name() == X_("Processor")) {
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
/* ForceIDRegeneration does not catch the following */
OMNIBUS COMMIT: prefer const XMLNode::property method (and provide a real one)
2016-05-04 23:09:37 -04:00
XMLProperty const * role = (*i)->property (X_("role"));
XMLProperty const * type = (*i)->property (X_("type"));
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
if (role && role->value() == X_("Aux")) {
/* check if the target bus exists.
* we should not save aux-sends in templates.
*/
OMNIBUS COMMIT: prefer const XMLNode::property method (and provide a real one)
2016-05-04 23:09:37 -04:00
XMLProperty const * target = (*i)->property (X_("target"));
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
if (!target) {
Use XMLNode::get/set_property API in ARDOUR::Session class
2016-08-28 07:12:16 -04:00
(*i)->set_property ("type", "dangling-aux-send");
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
continue;
}
boost::shared_ptr<Route> r = route_by_id (target->value());
if (!r || boost::dynamic_pointer_cast<Track>(r)) {
Use XMLNode::get/set_property API in ARDOUR::Session class
2016-08-28 07:12:16 -04:00
(*i)->set_property ("type", "dangling-aux-send");
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
continue;
}
}
Slightly unpleasant fix for creation of tracks from templates; it would be nice if we could set things up using the Route's logic for setting names of its children, rather than repeating the same logic in XML-land (#4303). git-svn-id: svn://localhost/ardour2/branches/3.0@10655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-16 18:03:59 -05:00
if (role && role->value() == X_("Listen")) {
(*i)->remove_property (X_("bitslot"));
}
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
else if (role && (role->value() == X_("Send") || role->value() == X_("Aux"))) {
Delivery::Role xrole;
uint32_t bitslot = 0;
xrole = Delivery::Role (string_2_enum (role->value(), xrole));
std::string name = Send::name_and_id_new_send(*this, xrole, bitslot, false);
(*i)->remove_property (X_("bitslot"));
(*i)->remove_property (X_("name"));
Use XMLNode::get/set_property API in ARDOUR::Session class
2016-08-28 07:12:16 -04:00
(*i)->set_property ("bitslot", bitslot);
(*i)->set_property ("name", name);
Fix route-templates with external sends
2019-03-27 22:01:53 -04:00
XMLNodeList io_kids = (*i)->children ();
for (XMLNodeList::iterator j = io_kids.begin(); j != io_kids.end(); ++j) {
if ((*j)->name() != X_("IO")) {
continue;
}
(*j)->remove_property (X_("name"));
(*j)->set_property ("name", name);
}
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
}
fix creating aux-busses from templates, reset return-bitslot
2016-06-02 16:01:36 -04:00
else if (type && type->value() == X_("intreturn")) {
(*i)->remove_property (X_("bitslot"));
Use XMLNode::get/set_property API in ARDOUR::Session class
2016-08-28 07:12:16 -04:00
(*i)->set_property ("ignore-bitslot", "1");
fix creating aux-busses from templates, reset return-bitslot
2016-06-02 16:01:36 -04:00
}
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
else if (type && type->value() == X_("return")) {
// Return::set_state() generates a new one
(*i)->remove_property (X_("bitslot"));
}
else if (type && type->value() == X_("port")) {
when creating a route from a template, use correct call to setup XML state so that name/ports are unique
2018-06-28 19:14:37 -04:00
IOProcessor::prepare_for_reset (**i, name);
allow to duplicate tracks with sends or load templates with sends/inserts
2016-04-26 12:15:53 -04:00
}
Slightly unpleasant fix for creation of tracks from templates; it would be nice if we could set things up using the Route's logic for setting names of its children, rather than repeating the same logic in XML-land (#4303). git-svn-id: svn://localhost/ardour2/branches/3.0@10655 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-11-16 18:03:59 -05:00
}
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
duplicate routes start off unsoloed to avoid issues related to upstream / downstream buses
2017-05-28 16:46:22 -04:00
/* new routes start off unsoloed to avoid issues related to
fix Session::new_route_from_template() to use playlist IDs and not names. This also fixes #7674
2019-03-20 23:09:17 -04:00
upstream / downstream buses.
*/
NO-OP: whitespace
2017-05-29 17:45:16 -04:00
node_copy.remove_node_and_delete (X_("Controllable"), X_("name"), X_("solo"));
duplicate routes start off unsoloed to avoid issues related to upstream / downstream buses
2017-05-28 16:46:22 -04:00
Remove hardcoded session-state versions
2019-12-11 07:55:51 -05:00
boost::shared_ptr<Route> route (XMLRouteFactory (node_copy, Stateful::loading_state_version));
Reset both IO names rather than just one when setting up the use of a track template. git-svn-id: svn://localhost/ardour2/branches/3.0@7085 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-08 20:47:16 -04:00
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
if (route == 0) {
error << _("Session: cannot create track/bus from template description") << endmsg;
goto out;
}
working but still confusing fix for ensuring that playlist shared-with data is correct when duplicating tracks. Confusing because orig_track_id is still being switched to the NEW track, and shared_with refers to the OLD one
2019-03-21 00:16:57 -04:00
/* Fix up sharing of playlists with the new Route/Track */
for (vector<boost::shared_ptr<Playlist> >::iterator sp = shared_playlists.begin(); sp != shared_playlists.end(); ++sp) {
(*sp)->share_with (route->id());
}
make track templates work, including a fix for the MidiTrack XML constructor git-svn-id: svn://localhost/ardour2/branches/3.0@4735 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-05 12:27:05 -05:00
if (boost::dynamic_pointer_cast<Track>(route)) {
/* force input/output change signals so that the new diskstream
picks up the configuration of the route. During session
loading this normally happens in a different way.
*/
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Auto-connect outputs that appear as a result of e.g. instruments being added to MIDI tracks. Fixes #3368. git-svn-id: svn://localhost/ardour2/branches/3.0@7720 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-30 18:34:21 -04:00
IOChange change (IOChange::Type (IOChange::ConfigurationChanged | IOChange::ConnectionsChanged));
change.after = route->input()->n_ports();
route->input()->changed (change, this);
change.after = route->output()->n_ports();
route->output()->changed (change, this);
make track templates work, including a fix for the MidiTrack XML constructor git-svn-id: svn://localhost/ardour2/branches/3.0@4735 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-05 12:27:05 -05:00
}
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
ret.push_back (route);
}
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
catch (failed_constructor &err) {
error << _("Session: could not create new route from template") << endmsg;
goto out;
}
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
catch (AudioEngine::PortRegistrationFailure& pfe) {
first half of the mega-by-hand-commit-by-commit merge from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4980 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-15 14:04:23 -04:00
error << pfe.what() << endmsg;
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
goto out;
}
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
Fix stuck connection block when template load fails. Session::new_route_from_template() unconditionally calls IO::disable_connecting(), and needs to be reverted likewise (no RAII here).
2018-11-26 15:45:06 -05:00
catch (...) {
IO::enable_connecting ();
throw;
}
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
--how_many;
}
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
out:
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
if (!ret.empty()) {
use state Session::StateProtector
2014-06-28 15:36:13 -04:00
StateProtector sp (this);
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
add_routes (ret, true, true, false, insert_at);
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
}
Fix stuck connection block when template load fails. Session::new_route_from_template() unconditionally calls IO::disable_connecting(), and needs to be reverted likewise (no RAII here).
2018-11-26 15:45:06 -05:00
IO::enable_connecting ();
track templates, backported from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4713 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-03-02 11:52:40 -05:00
return ret;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Session::add_routes (RouteList& new_routes, bool input_auto_connect, bool output_auto_connect, bool save, PresentationInfo::order_t order)
dramatically speed up the addition of large numbers of busses + tracks. consists of a backend part (ignore JACK graph/latency callbacks while we're adding tracks) and a GUI side (avoid O(N^N) behaviour while adding each new time axis view) git-svn-id: svn://localhost/ardour2/branches/3.0@13595 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-04 09:32:28 -05:00
{
try {
PBD::Unwinder<bool> aip (_adding_routes_in_progress, true);
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
add_routes_inner (new_routes, input_auto_connect, output_auto_connect, order);
dramatically speed up the addition of large numbers of busses + tracks. consists of a backend part (ignore JACK graph/latency callbacks while we're adding tracks) and a GUI side (avoid O(N^N) behaviour while adding each new time axis view) git-svn-id: svn://localhost/ardour2/branches/3.0@13595 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-04 09:32:28 -05:00
} catch (...) {
error << _("Adding new tracks/busses failed") << endmsg;
}
add explanatory comment
2019-10-28 15:36:38 -04:00
/* During the route additions there will have been potentially several
* signals emitted to indicate the new graph. ::graph_reordered() will
* have ignored all of them because _adding_routes_in_progress was
* true.
*
* We still need the effects of ::graph_reordered(), but we didn't want
* it called multiple times during the addition of multiple routes. Now
* that the addition is done, call it explicitly.
*/
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
graph_reordered (false);
dramatically speed up the addition of large numbers of busses + tracks. consists of a backend part (ignore JACK graph/latency callbacks while we're adding tracks) and a GUI side (avoid O(N^N) behaviour while adding each new time axis view) git-svn-id: svn://localhost/ardour2/branches/3.0@13595 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-04 09:32:28 -05:00
set_dirty();
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
dramatically speed up the addition of large numbers of busses + tracks. consists of a backend part (ignore JACK graph/latency callbacks while we're adding tracks) and a GUI side (avoid O(N^N) behaviour while adding each new time axis view) git-svn-id: svn://localhost/ardour2/branches/3.0@13595 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-04 09:32:28 -05:00
if (save) {
save_state (_current_snapshot_name);
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
[Summary] Changed function name update_route_active_state() on update_route_record_state() [Reviewed] GZharun Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc libs/ardour/session_state.cc
2014-09-19 09:45:01 -04:00
update_route_record_state ();
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
dramatically speed up the addition of large numbers of busses + tracks. consists of a backend part (ignore JACK graph/latency callbacks while we're adding tracks) and a GUI side (avoid O(N^N) behaviour while adding each new time axis view) git-svn-id: svn://localhost/ardour2/branches/3.0@13595 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-04 09:32:28 -05:00
RouteAdded (new_routes); /* EMIT SIGNAL */
}
void
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Session::add_routes_inner (RouteList& new_routes, bool input_auto_connect, bool output_auto_connect, PresentationInfo::order_t order)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
ChanCount existing_inputs;
ChanCount existing_outputs;
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
uint32_t n_routes;
uint32_t added = 0;
New routes are placed after highest selected route.
2013-10-23 10:27:13 -04:00
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
count_existing_track_channels (existing_inputs, existing_outputs);
fix all manner of things relating to io connections, setting capture alignment, and so on. still needs more tests of actual precise placement of newly recorded material git-svn-id: svn://localhost/ardour2/branches/3.0@9155 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-15 15:32:21 -04:00
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
{
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
RCUWriter<RouteList> writer (routes);
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
boost::shared_ptr<RouteList> r = writer.get_copy ();
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
n_routes = r->size();
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
r->insert (r->end(), new_routes.begin(), new_routes.end());
first pass (ok, third really) at internal send+return - audio routing inside ardour without JACK. lots still to do, but at least the obvious works git-svn-id: svn://localhost/ardour2/branches/3.0@5202 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-16 10:58:33 -04:00
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
/* if there is no control out and we're not in the middle of loading,
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
* resort the graph here. if there is a control out, we will resort
* toward the end of this method. if we are in the middle of loading,
* we will resort when done.
*/
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
remove near-duplicate session constructor; change names from control_outs to monitor_out and control_out to monitor_send; may havebroken creation of new sessions in a wierd way (will fix later) git-svn-id: svn://localhost/ardour2/branches/3.0@6783 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-22 17:35:35 -04:00
if (!_monitor_out && IO::connecting_legal) {
first pass (ok, third really) at internal send+return - audio routing inside ardour without JACK. lots still to do, but at least the obvious works git-svn-id: svn://localhost/ardour2/branches/3.0@5202 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-16 10:58:33 -04:00
resort_routes_using (r);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
/* monitor is not part of the order */
Fix initial order-keys when creating tracks/busses .. and leave some notes for the day when UI treemodel won't be canonical to set order-keys.
2017-02-16 18:50:52 -05:00
if (_monitor_out) {
assert (n_routes > 0);
--n_routes;
}
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
{
PresentationInfo::ChangeSuspender cs;
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
ensure_route_presentation_info_gap (order, new_routes.size());
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
for (RouteList::iterator x = new_routes.begin(); x != new_routes.end(); ++x, ++added) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
boost::weak_ptr<Route> wpr (*x);
boost::shared_ptr<Route> r (*x);
change API of Controllable::Changed signal to include (from_self, GroupControlDisposition) This allows the signal to convey more information, which may be required by some handlers of a control's Changed signal
2016-03-11 10:34:51 -05:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
r->solo_control()->Changed.connect_same_thread (*this, boost::bind (&Session::route_solo_changed, this, _1, _2,wpr));
r->solo_isolate_control()->Changed.connect_same_thread (*this, boost::bind (&Session::route_solo_isolated_changed, this, wpr));
r->mute_control()->Changed.connect_same_thread (*this, boost::bind (&Session::route_mute_changed, this));
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
r->output()->changed.connect_same_thread (*this, boost::bind (&Session::set_worst_io_latencies_x, this, _1, _2));
r->processors_changed.connect_same_thread (*this, boost::bind (&Session::route_processors_changed, this, _1));
r->processor_latency_changed.connect_same_thread (*this, boost::bind (&Session::queue_latency_recompute, this));
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
if (r->is_master()) {
_master_out = r;
}
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
if (r->is_monitor()) {
_monitor_out = r;
}
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (r);
if (tr) {
tr->PlaylistChanged.connect_same_thread (*this, boost::bind (&Session::track_playlist_changed, this, boost::weak_ptr<Track> (tr)));
track_playlist_changed (boost::weak_ptr<Track> (tr));
tr->rec_enable_control()->Changed.connect_same_thread (*this, boost::bind (&Session::update_route_record_state, this));
boost::shared_ptr<MidiTrack> mt = boost::dynamic_pointer_cast<MidiTrack> (tr);
if (mt) {
mt->StepEditStatusChange.connect_same_thread (*this, boost::bind (&Session::step_edit_status_change, this, _1));
mt->output()->changed.connect_same_thread (*this, boost::bind (&Session::midi_output_change_handler, this, _1, _2, boost::weak_ptr<Route>(mt)));
mt->presentation_info().PropertyChanged.connect_same_thread (*this, boost::bind (&Session::midi_track_presentation_info_changed, this, _1, boost::weak_ptr<MidiTrack>(mt)));
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
if (!r->presentation_info().special (false)) {
first pass at the big rethink of managing sort order keys for editor and mixer. this appears to work, but remote control IDs are not yet correct (frequently off by one because of the presence of the master bus in the editor) git-svn-id: svn://localhost/ardour2/branches/3.0@12953 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-27 18:57:06 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
DEBUG_TRACE (DEBUG::OrderKeys, string_compose ("checking PI state for %1\n", r->name()));
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
/* presentation info order may already have been set from XML */
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
if (!r->presentation_info().order_set()) {
if (order == PresentationInfo::max_order) {
/* just add to the end */
r->set_presentation_order (n_routes + added);
DEBUG_TRACE (DEBUG::OrderKeys, string_compose ("group order not set, set to NR %1 + %2 = %3\n", n_routes, added, n_routes + added));
} else {
r->set_presentation_order (order + added);
DEBUG_TRACE (DEBUG::OrderKeys, string_compose ("group order not set, set to %1 + %2 = %3\n", order, added, order + added));
}
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
} else {
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
DEBUG_TRACE (DEBUG::OrderKeys, string_compose ("group order already set to %1\n", r->presentation_info().order()));
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
}
first pass at the big rethink of managing sort order keys for editor and mixer. this appears to work, but remote control IDs are not yet correct (frequently off by one because of the presence of the master bus in the editor) git-svn-id: svn://localhost/ardour2/branches/3.0@12953 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-27 18:57:06 -04:00
}
reinstate the windows way for GUIIdle's but limit the scope of the connection to just the time when a long-running action is active (for now, just adding > 8 routes) git-svn-id: svn://localhost/ardour2/branches/3.0@13654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-13 08:44:11 -05:00
Enable build for FreeBSD (part 1/2) Adopted from Michael Beer -- GH pull-request #232 with minor changes: * rebased on master, * removed trailing whitespace, * don't explicitly change saved configuration defaults (wscript) * moved sys/wait (WNOHANG) to header include * separate changes in GUI and lib
2016-06-26 10:43:07 -04:00
#if !defined(__APPLE__) && !defined(__FreeBSD__)
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
/* clang complains: 'operator<<' should be declared prior to the call site or in an associated namespace of one of its
* arguments std::ostream& operator<<(std::ostream& o, ARDOUR::PresentationInfo const& rid)"
*/
DEBUG_TRACE (DEBUG::OrderKeys, string_compose ("added route %1, group order %2 type %3 (summary: %4)\n",
r->name(),
r->presentation_info().order(),
enum_2_string (r->presentation_info().flags()),
r->presentation_info()));
OS X clang build issue fix
2016-06-05 17:59:56 -04:00
#endif
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
if (input_auto_connect || output_auto_connect) {
auto_connect_route (r, input_auto_connect, ChanCount (), ChanCount (), existing_inputs, existing_outputs);
existing_inputs += r->n_inputs();
existing_outputs += r->n_outputs();
}
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
use RAII for class-wide PresentationInfo::Change signal, along with properties to describe what changed
2017-01-26 13:20:36 -05:00
ARDOUR::GUIIdle ();
}
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
ensure_stripable_sort_order ();
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
remove near-duplicate session constructor; change names from control_outs to monitor_out and control_out to monitor_send; may havebroken creation of new sessions in a wierd way (will fix later) git-svn-id: svn://localhost/ardour2/branches/3.0@6783 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-22 17:35:35 -04:00
if (_monitor_out && IO::connecting_legal) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Glib::Threads::Mutex::Lock lm (_engine.process_lock());
dramatically speed up the addition of large numbers of busses + tracks. consists of a backend part (ignore JACK graph/latency callbacks while we're adding tracks) and a GUI side (avoid O(N^N) behaviour while adding each new time axis view) git-svn-id: svn://localhost/ardour2/branches/3.0@13595 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-04 09:32:28 -05:00
for (RouteList::iterator x = new_routes.begin(); x != new_routes.end(); ++x) {
if ((*x)->is_monitor()) {
/* relax */
} else if ((*x)->is_master()) {
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
/* relax */
dramatically speed up the addition of large numbers of busses + tracks. consists of a backend part (ignore JACK graph/latency callbacks while we're adding tracks) and a GUI side (avoid O(N^N) behaviour while adding each new time axis view) git-svn-id: svn://localhost/ardour2/branches/3.0@13595 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-12-04 09:32:28 -05:00
} else {
(*x)->enable_monitor_send ();
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
clean up debug output for OrderKeys
2016-05-17 09:36:28 -04:00
reassign_track_numbers ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
implement set-send-gains-to-{track_level,zero,unity} git-svn-id: svn://localhost/ardour2/branches/3.0@5401 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-21 08:05:44 -04:00
void
Session::globally_set_send_gains_to_zero (boost::shared_ptr<Route> dest)
{
boost::shared_ptr<RouteList> r = routes.reader ();
boost::shared_ptr<Send> s;
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
Make 'copy track gains to sends' and 'set sends gains' work with sends from busses. Partially fixes #3637. git-svn-id: svn://localhost/ardour2/branches/3.0@8369 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-29 16:07:22 -05:00
if ((s = (*i)->internal_send_for (dest)) != 0) {
change Controllable::set_value() API to include grouped control consideration. This also removes Route::group_gain_control() and associated machinery. Not yet tested with Mackie or other surfaces. More work to done to start using the group capabilities, and also potentially to add or derive more controls as RouteAutomationControls
2016-01-02 04:58:23 -05:00
s->amp()->gain_control()->set_value (GAIN_COEFF_ZERO, Controllable::NoGroup);
implement set-send-gains-to-{track_level,zero,unity} git-svn-id: svn://localhost/ardour2/branches/3.0@5401 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-21 08:05:44 -04:00
}
}
}
void
Session::globally_set_send_gains_to_unity (boost::shared_ptr<Route> dest)
{
boost::shared_ptr<RouteList> r = routes.reader ();
boost::shared_ptr<Send> s;
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
Make 'copy track gains to sends' and 'set sends gains' work with sends from busses. Partially fixes #3637. git-svn-id: svn://localhost/ardour2/branches/3.0@8369 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-29 16:07:22 -05:00
if ((s = (*i)->internal_send_for (dest)) != 0) {
change Controllable::set_value() API to include grouped control consideration. This also removes Route::group_gain_control() and associated machinery. Not yet tested with Mackie or other surfaces. More work to done to start using the group capabilities, and also potentially to add or derive more controls as RouteAutomationControls
2016-01-02 04:58:23 -05:00
s->amp()->gain_control()->set_value (GAIN_COEFF_UNITY, Controllable::NoGroup);
implement set-send-gains-to-{track_level,zero,unity} git-svn-id: svn://localhost/ardour2/branches/3.0@5401 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-21 08:05:44 -04:00
}
}
}
void
Session::globally_set_send_gains_from_track(boost::shared_ptr<Route> dest)
{
boost::shared_ptr<RouteList> r = routes.reader ();
boost::shared_ptr<Send> s;
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
Make 'copy track gains to sends' and 'set sends gains' work with sends from busses. Partially fixes #3637. git-svn-id: svn://localhost/ardour2/branches/3.0@8369 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-29 16:07:22 -05:00
if ((s = (*i)->internal_send_for (dest)) != 0) {
change Controllable::set_value() API to include grouped control consideration. This also removes Route::group_gain_control() and associated machinery. Not yet tested with Mackie or other surfaces. More work to done to start using the group capabilities, and also potentially to add or derive more controls as RouteAutomationControls
2016-01-02 04:58:23 -05:00
s->amp()->gain_control()->set_value ((*i)->gain_control()->get_value(), Controllable::NoGroup);
implement set-send-gains-to-{track_level,zero,unity} git-svn-id: svn://localhost/ardour2/branches/3.0@5401 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-21 08:05:44 -04:00
}
}
}
Add options to add sends from buses as well as tracks; remainder of #3637. git-svn-id: svn://localhost/ardour2/branches/3.0@8372 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-29 17:07:34 -05:00
/** @param include_buses true to add sends to buses and tracks, false for just tracks */
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
void
Add options to add sends from buses as well as tracks; remainder of #3637. git-svn-id: svn://localhost/ardour2/branches/3.0@8372 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-29 17:07:34 -05:00
Session::globally_add_internal_sends (boost::shared_ptr<Route> dest, Placement p, bool include_buses)
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
{
boost::shared_ptr<RouteList> r = routes.reader ();
boost::shared_ptr<RouteList> t (new RouteList);
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
don't use MIDI tracks when adding internal sends to auxes; no varispeed when recording and the corollary - no recording when varispeeding git-svn-id: svn://localhost/ardour2/branches/3.0@9974 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 16:21:18 -04:00
/* no MIDI sends because there are no MIDI busses yet */
if (include_buses || boost::dynamic_pointer_cast<AudioTrack>(*i)) {
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
t->push_back (*i);
}
}
processor naming tweaks, processor visibility tweaks and more, trying to finish this audio routing thing... git-svn-id: svn://localhost/ardour2/branches/3.0@5364 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-14 20:47:34 -04:00
add_internal_sends (dest, p, t);
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
}
void
processor naming tweaks, processor visibility tweaks and more, trying to finish this audio routing thing... git-svn-id: svn://localhost/ardour2/branches/3.0@5364 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-14 20:47:34 -04:00
Session::add_internal_sends (boost::shared_ptr<Route> dest, Placement p, boost::shared_ptr<RouteList> senders)
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
{
Add aux sends at the position that the menu was opened, rather than always pre-fader (#4660). git-svn-id: svn://localhost/ardour2/branches/3.0@11387 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-30 13:09:54 -05:00
for (RouteList::iterator i = senders->begin(); i != senders->end(); ++i) {
add_internal_send (dest, (*i)->before_processor_for_placement (p), *i);
}
}
void
Session::add_internal_send (boost::shared_ptr<Route> dest, int index, boost::shared_ptr<Route> sender)
{
add_internal_send (dest, sender->before_processor_for_index (index), sender);
}
void
Session::add_internal_send (boost::shared_ptr<Route> dest, boost::shared_ptr<Processor> before, boost::shared_ptr<Route> sender)
{
if (sender->is_monitor() || sender->is_master() || sender == dest || dest->is_monitor() || dest->is_master()) {
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
return;
}
if (!dest->internal_return()) {
Add aux sends at the position that the menu was opened, rather than always pre-fader (#4660). git-svn-id: svn://localhost/ardour2/branches/3.0@11387 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-30 13:09:54 -05:00
dest->add_internal_return ();
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
}
add new Graph debug bit ; make adding aux sends really work git-svn-id: svn://localhost/ardour2/branches/3.0@6132 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-19 12:06:00 -05:00
Add aux sends at the position that the menu was opened, rather than always pre-fader (#4660). git-svn-id: svn://localhost/ardour2/branches/3.0@11387 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-30 13:09:54 -05:00
sender->add_aux_send (dest, before);
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
graph_reordered (false);
startup assistant patch from tinman; cleanup fix backported from 2.X ; easy(ier) ways to create aux sends ; facility to subgroup (route via bus) for a route group ; fix up internal send/return operation ; fix internal send naming since it doesn't need to be unique - no JACK ports involved git-svn-id: svn://localhost/ardour2/branches/3.0@5272 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-25 16:46:39 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
Session::remove_routes (boost::shared_ptr<RouteList> routes_to_remove)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Emit Session::MuteChanged() if a mutes route is removed.
2017-04-08 17:44:18 -04:00
bool mute_changed = false;
when removing a Route from Session, remove it from CoreSelection and signal that change if necessary
2017-05-08 05:36:46 -04:00
bool send_selected = false;
Emit Session::MuteChanged() if a mutes route is removed.
2017-04-08 17:44:18 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
{ // RCU Writer scope
fix deadlock when removing subgroup bus w/sends. ...and potentially other route removals that involve processor changes.
2015-10-15 07:49:35 -04:00
PBD::Unwinder<bool> uw_flag (_route_deletion_in_progress, true);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
RCUWriter<RouteList> writer (routes);
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
boost::shared_ptr<RouteList> rs = writer.get_copy ();
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
for (RouteList::iterator iter = routes_to_remove->begin(); iter != routes_to_remove->end(); ++iter) {
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
when removing a Route from Session, remove it from CoreSelection and signal that change if necessary
2017-05-08 05:36:46 -04:00
if (_selection->selected (*iter)) {
send_selected = true;
}
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
if (*iter == _master_out) {
continue;
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
Speed up unloading large sessions.
2016-12-07 06:50:35 -05:00
/* speed up session deletion, don't do the solo dance */
NO-OP: prefer to use API to check session-state
2019-03-17 12:04:45 -04:00
if (!deletion_in_progress ()) {
Speed up unloading large sessions.
2016-12-07 06:50:35 -05:00
(*iter)->solo_control()->set_value (0.0, Controllable::NoGroup);
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
Emit Session::MuteChanged() if a mutes route is removed.
2017-04-08 17:44:18 -04:00
if ((*iter)->mute_control()->muted ()) {
mute_changed = true;
}
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
rs->remove (*iter);
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
/* deleting the master out seems like a dumb
idea, but its more of a UI policy issue
than our concern.
*/
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
if (*iter == _master_out) {
_master_out = boost::shared_ptr<Route> ();
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
if (*iter == _monitor_out) {
_monitor_out.reset ();
Remove internal sends to an aux bus when it is removed. git-svn-id: svn://localhost/ardour2/branches/3.0@6228 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-30 18:34:48 -05:00
}
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
// We need to disconnect the route's inputs and outputs
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
(*iter)->input()->disconnect (0);
(*iter)->output()->disconnect (0);
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
/* if the route had internal sends sending to it, remove them */
Speed up session deletion No need to reconfigure routes due to removal of monitor or aux send/return. The processors will be removed with the route.
2019-03-17 12:35:10 -04:00
if (!deletion_in_progress () && (*iter)->internal_return()) {
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
boost::shared_ptr<RouteList> r = routes.reader ();
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
boost::shared_ptr<Send> s = (*i)->internal_send_for (*iter);
if (s) {
(*i)->remove_processor (s);
}
}
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
/* if the monitoring section had a pointer to this route, remove it */
if (_monitor_out && !(*iter)->is_master() && !(*iter)->is_monitor()) {
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
RAII to postpone processor changes/graph recalculation
2016-11-25 12:07:43 -05:00
ProcessorChangeBlocker pcb (this, false);
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
(*iter)->remove_aux_or_listen (_monitor_out);
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
boost::shared_ptr<MidiTrack> mt = boost::dynamic_pointer_cast<MidiTrack> (*iter);
if (mt && mt->step_editing()) {
if (_step_editors > 0) {
_step_editors--;
}
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
/* writer goes out of scope, forces route list update */
step editing now accessed via a context menu on MIDI track rec-enable buttons. along the way, fixed up confusion between set_rec_enable() and set_rec_enabled() methods in Route/Track/Diskstream git-svn-id: svn://localhost/ardour2/branches/3.0@7482 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-24 12:40:56 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
} // end of RCU Writer scope
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
Emit Session::MuteChanged() if a mutes route is removed.
2017-04-08 17:44:18 -04:00
if (mute_changed) {
MuteChanged (); /* EMIT SIGNAL */
}
when removing routes, don't do potentially expensive work once for each route. We can update solo state and tell interested parties about the removal once the actual removal is done
2015-09-10 11:35:25 -04:00
update_route_solo_state ();
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
update_latency_compensation (false, false);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
set_dirty();
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
/* Re-sort routes to remove the graph's current references to the one that is
Fix track lifetime wrt graph. Fixes #3251. git-svn-id: svn://localhost/ardour2/branches/3.0@7410 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-13 22:23:37 -04:00
* going away, then flush old references out of the graph.
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
*/
clear activation_sets component of Graph, thus restoring full Route deletion git-svn-id: svn://localhost/ardour2/branches/3.0@7298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-24 12:20:32 -04:00
goodbye USE_TRACKS_CODE_FEATURES and is_tracks_build
2019-09-25 14:16:13 -04:00
routes.flush (); // maybe unsafe, see below.
resort_routes ();
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (_process_graph && !deletion_in_progress() && _engine.running()) {
For now, only use the multi-threaded process code if we are using >1 processor for DSP; this involves making the DSP use setting only take effect on a restart of Ardour. git-svn-id: svn://localhost/ardour2/branches/3.0@11302 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-22 07:28:49 -05:00
_process_graph->clear_other_chain ();
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* get rid of it from the dead wood collection in the route list manager */
/* XXX i think this is unsafe as it currently stands, but i am not sure. (pd, october 2nd, 2006) */
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
routes.flush ();
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
fix logic when removing routes Selection change and re-order signals are distinct.
2017-05-08 05:56:24 -04:00
/* remove these routes from the selection if appropriate, and signal
* the change *before* we call DropReferences for them.
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
when removing a Route from Session, remove it from CoreSelection and signal that change if necessary
2017-05-08 05:36:46 -04:00
if (send_selected && !deletion_in_progress()) {
for (RouteList::iterator iter = routes_to_remove->begin(); iter != routes_to_remove->end(); ++iter) {
_selection->remove_stripable_by_id ((*iter)->id());
}
fix logic when removing routes Selection change and re-order signals are distinct.
2017-05-08 05:56:24 -04:00
PropertyChange pc;
pc.add (Properties::selected);
PresentationInfo::Change (pc);
when removing a Route from Session, remove it from CoreSelection and signal that change if necessary
2017-05-08 05:36:46 -04:00
}
fix logic when removing routes Selection change and re-order signals are distinct.
2017-05-08 05:56:24 -04:00
/* try to cause everyone to drop their references
* and unregister ports from the backend
*/
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
for (RouteList::iterator iter = routes_to_remove->begin(); iter != routes_to_remove->end(); ++iter) {
(*iter)->drop_references ();
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
when removing a Route from Session, remove it from CoreSelection and signal that change if necessary
2017-05-08 05:36:46 -04:00
if (deletion_in_progress()) {
further tweaks to session close
2016-04-23 20:51:23 -04:00
return;
}
fix logic when removing routes Selection change and re-order signals are distinct.
2017-05-08 05:56:24 -04:00
PropertyChange pc;
fix copy-n-paste error
2017-05-08 06:03:35 -04:00
pc.add (Properties::order);
fix logic when removing routes Selection change and re-order signals are distinct.
2017-05-08 05:56:24 -04:00
PresentationInfo::Change (pc);
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* save the new state of the world */
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (save_state (_current_snapshot_name)) {
save_history (_current_snapshot_name);
}
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
[Summary] Changed function name update_route_active_state() on update_route_record_state() [Reviewed] GZharun Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc libs/ardour/session_state.cc
2014-09-19 09:45:01 -04:00
update_route_record_state ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
integrate Session::remove_routes() from Tracks and replace Session::remove_route() internal implementation. The singular case replicated (and was the model for) the plural case.
2015-05-08 16:04:15 -04:00
void
Session::remove_route (boost::shared_ptr<Route> route)
{
boost::shared_ptr<RouteList> rl (new RouteList);
rl->push_back (route);
remove_routes (rl);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
change API of Controllable::Changed signal to include (from_self, GroupControlDisposition) This allows the signal to convey more information, which may be required by some handlers of a control's Changed signal
2016-03-11 10:34:51 -05:00
Session::route_mute_changed ()
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Add session signal to track mute-changes In preparation for Faderport8 "Mute Clear" LED Button.
2017-04-07 17:26:59 -04:00
MuteChanged (); /* EMIT SIGNAL */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
set_dirty ();
}
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
void
change API of Controllable::Changed signal to include (from_self, GroupControlDisposition) This allows the signal to convey more information, which may be required by some handlers of a control's Changed signal
2016-03-11 10:34:51 -05:00
Session::route_listen_changed (Controllable::GroupControlDisposition group_override, boost::weak_ptr<Route> wpr)
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
{
change API of Controllable::Changed signal to include (from_self, GroupControlDisposition) This allows the signal to convey more information, which may be required by some handlers of a control's Changed signal
2016-03-11 10:34:51 -05:00
boost::shared_ptr<Route> route (wpr.lock());
if (!route) {
return;
}
remove Route::listening_via_monitor() This is better tested with direct use of the solo_control and Config->get_solo_controls_are_listen_controls()
2016-04-21 02:02:13 -04:00
assert (Config->get_solo_control_is_listen_control());
if (route->solo_control()->soloed_by_self_or_masters()) {
make "exclusive solo" apply to listen as well as SiP git-svn-id: svn://localhost/ardour2/branches/3.0@7064 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-05 11:47:51 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (Config->get_exclusive_solo()) {
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
allow grouped AFL/PFL with excl. solo
2015-10-04 18:27:48 -04:00
RouteGroup* rg = route->route_group ();
fix backend logic for PFL/AFL when mastered
2016-04-21 01:45:09 -04:00
const bool group_already_accounted_for = (group_override == Controllable::ForGroup);
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
if ((*i) == route) {
/* already changed */
continue;
}
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
if ((*i)->solo_isolate_control()->solo_isolated() || !(*i)->can_solo()) {
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
/* route does not get solo propagated to it */
continue;
}
if ((group_already_accounted_for && (*i)->route_group() && (*i)->route_group() == rg)) {
/* this route is a part of the same solo group as the route
* that was changed. Changing that route did change or will
* change all group members appropriately, so we can ignore it
* here
*/
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
continue;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
}
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
(*i)->solo_control()->set_value (0.0, Controllable::NoGroup);
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
}
make "exclusive solo" apply to listen as well as SiP git-svn-id: svn://localhost/ardour2/branches/3.0@7064 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-05 11:47:51 -04:00
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
_listen_cnt++;
make "exclusive solo" apply to listen as well as SiP git-svn-id: svn://localhost/ardour2/branches/3.0@7064 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-05 11:47:51 -04:00
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
} else if (_listen_cnt > 0) {
make "exclusive solo" apply to listen as well as SiP git-svn-id: svn://localhost/ardour2/branches/3.0@7064 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-05 11:47:51 -04:00
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
_listen_cnt--;
}
}
new API for route solo/mute state mgmt Route now calls back into Session when solo/mute/listen state changes. All other interested parties must use the Route::{solo,mute,...}_control()->Changed() to be notified of changes. The Session requires more information than the Changed signal can provide, in order to propagate solo/mute changes across the entire Session correctly. Note that this uses an experimental use of CRTP to isolate a public API within Session
2016-03-09 13:11:53 -05:00
lots of odds and ends to do with solo isolate and its GUI git-svn-id: svn://localhost/ardour2/branches/3.0@7072 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-06 14:40:37 -04:00
void
change API of Controllable::Changed signal to include (from_self, GroupControlDisposition) This allows the signal to convey more information, which may be required by some handlers of a control's Changed signal
2016-03-11 10:34:51 -05:00
Session::route_solo_isolated_changed (boost::weak_ptr<Route> wpr)
lots of odds and ends to do with solo isolate and its GUI git-svn-id: svn://localhost/ardour2/branches/3.0@7072 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-06 14:40:37 -04:00
{
change API of Controllable::Changed signal to include (from_self, GroupControlDisposition) This allows the signal to convey more information, which may be required by some handlers of a control's Changed signal
2016-03-11 10:34:51 -05:00
boost::shared_ptr<Route> route (wpr.lock());
if (!route) {
return;
}
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
bool send_changed = false;
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
if (route->solo_isolate_control()->solo_isolated()) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
if (_solo_isolated_cnt == 0) {
send_changed = true;
}
_solo_isolated_cnt++;
} else if (_solo_isolated_cnt > 0) {
_solo_isolated_cnt--;
if (_solo_isolated_cnt == 0) {
send_changed = true;
}
}
if (send_changed) {
IsolatedChanged (); /* EMIT SIGNAL */
}
lots of odds and ends to do with solo isolate and its GUI git-svn-id: svn://localhost/ardour2/branches/3.0@7072 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-06 14:40:37 -04:00
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
move ever close to working master/slave logic, this time with audio testing
2016-04-20 23:22:29 -04:00
Session::route_solo_changed (bool self_solo_changed, Controllable::GroupControlDisposition group_override, boost::weak_ptr<Route> wpr)
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
{
provisional changes to speed up solo changes to large numbers of routes. Moves global update of solo state and emission of Session::SoloChanged to a single point after 1 to N solo controls are changed. Also avoid unnecessarily emitted Activated() signal for listen controls, though Process::{activate,deactive}() should probably be redesigned to avoid this in a "deeper" way
2016-11-25 05:29:16 -05:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("route solo change, self = %1, update\n", self_solo_changed));
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
change API of Controllable::Changed signal to include (from_self, GroupControlDisposition) This allows the signal to convey more information, which may be required by some handlers of a control's Changed signal
2016-03-11 10:34:51 -05:00
boost::shared_ptr<Route> route (wpr.lock());
if (!route) {
return;
}
if (Config->get_solo_control_is_listen_control()) {
route_listen_changed (group_override, wpr);
return;
}
move ever close to working master/slave logic, this time with audio testing
2016-04-20 23:22:29 -04:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("%1: self %2 masters %3 transition %4\n", route->name(), route->self_soloed(), route->solo_control()->get_masters_value(), route->solo_control()->transitioned_into_solo()));
if (route->solo_control()->transitioned_into_solo() == 0) {
remove redundant (identical) 2nd conditional, and improve comment
2016-11-25 06:50:08 -05:00
/* route solo changed by upstream/downstream or clear all solo state; not interesting
move ever close to working master/slave logic, this time with audio testing
2016-04-20 23:22:29 -04:00
to Session.
*/
DEBUG_TRACE (DEBUG::Solo, string_compose ("%1 not self-soloed nor soloed by master (%2), ignoring\n", route->name(), route->solo_control()->get_masters_value()));
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
return;
}
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
move ever close to working master/slave logic, this time with audio testing
2016-04-20 23:22:29 -04:00
boost::shared_ptr<RouteList> r = routes.reader ();
int32_t delta = route->solo_control()->transitioned_into_solo ();
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
/* the route may be a member of a group that has shared-solo
* semantics. If so, then all members of that group should follow the
* solo of the changed route. But ... this is optional, controlled by a
* Controllable::GroupControlDisposition.
*
* The first argument to the signal that this method is connected to is the
* GroupControlDisposition value that was used to change solo.
*
rename Controllable::WholeGroup => InverseGroup to reflect real intent. Add logic to RouteGroupMember::use_group() to implement the real intent of InverseGroup
2016-01-21 12:04:20 -05:00
* If the solo change was done with group semantics (either InverseGroup
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
* (force the entire group to change even if the group shared solo is
* disabled) or UseGroup (use the group, which may or may not have the
* shared solo property enabled)) then as we propagate the change to
* the entire session we should IGNORE THE GROUP that the changed route
* belongs to.
*/
in exclusive solo mode, allow full-group solo git-svn-id: svn://localhost/ardour2/branches/3.0@9917 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-25 12:10:11 -04:00
RouteGroup* rg = route->route_group ();
use ForGroup to flag and detect route-group based control changes
2016-03-14 10:42:01 -04:00
const bool group_already_accounted_for = (group_override == Controllable::ForGroup);
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
provisional changes to speed up solo changes to large numbers of routes. Moves global update of solo state and emission of Session::SoloChanged to a single point after 1 to N solo controls are changed. Also avoid unnecessarily emitted Activated() signal for listen controls, though Process::{activate,deactive}() should probably be redesigned to avoid this in a "deeper" way
2016-11-25 05:29:16 -05:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("propagate to session, group accounted for ? %1\n", group_already_accounted_for));
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (delta == 1 && Config->get_exclusive_solo()) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
in exclusive solo mode, allow full-group solo git-svn-id: svn://localhost/ardour2/branches/3.0@9917 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-25 12:10:11 -04:00
/* new solo: disable all other solos, but not the group if its solo-enabled */
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
if ((*i) == route) {
/* already changed */
continue;
}
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
if ((*i)->solo_isolate_control()->solo_isolated() || !(*i)->can_solo()) {
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
/* route does not get solo propagated to it */
continue;
}
if ((group_already_accounted_for && (*i)->route_group() && (*i)->route_group() == rg)) {
/* this route is a part of the same solo group as the route
* that was changed. Changing that route did change or will
* change all group members appropriately, so we can ignore it
* here
*/
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
continue;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
}
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
(*i)->solo_control()->set_value (0.0, group_override);
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("propagate solo change, delta = %1\n", delta));
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
RouteList uninvolved;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("%1\n", route->name()));
the big Route structure refactor. !!!! THIS WILL ***NOT LOAD*** PRIOR 3.0 or 2.X SESSIONS !!!! BREAKAGE IS EXPECTED !!!! IF YOU HAVE AND NEED A WORKING 3.0 DO **NOT** UPDATE. !!!! otherwise, update and enjoy the steadily emerging joys of this major reworking of ardour internals git-svn-id: svn://localhost/ardour2/branches/3.0@5137 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-09 16:21:19 -04:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
disabled changes to make solo propagation ignore sends git-svn-id: svn://localhost/ardour2/branches/3.0@6113 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-17 17:29:43 -05:00
bool via_sends_only;
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
bool in_signal_flow;
fix mute & solo behaviour mostly ; remove some verbose debugging output git-svn-id: svn://localhost/ardour2/branches/3.0@5153 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-10 09:59:06 -04:00
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
if ((*i) == route) {
/* already changed */
continue;
}
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
if ((*i)->solo_isolate_control()->solo_isolated() || !(*i)->can_solo()) {
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
/* route does not get solo propagated to it */
MuteMaster should (a) use a Muteable's own ::muted_by_others_soloing() (b) not try to use its own _solo_ignore to track Muteable::can_solo() or solo isolate state
2016-07-20 16:10:11 -04:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("%1 excluded from solo because iso = %2 can_solo = %3\n", (*i)->name(), (*i)->solo_isolate_control()->solo_isolated(),
(*i)->can_solo()));
first compiling, mostly working version of group controls changes
2016-01-21 11:03:14 -05:00
continue;
}
if ((group_already_accounted_for && (*i)->route_group() && (*i)->route_group() == rg)) {
/* this route is a part of the same solo group as the route
* that was changed. Changing that route did change or will
* change all group members appropriately, so we can ignore it
* here
*/
some provisional support for "real" solo-isolate behaviour, subject to more discussions with mr. oofus and others git-svn-id: svn://localhost/ardour2/branches/3.0@6145 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-21 12:20:57 -05:00
continue;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
}
yet more work on Ye Fabled Solo Architecture. now do forward and backward propagation of solo status from a soloed track. tweak GUI appearance git-svn-id: svn://localhost/ardour2/branches/3.0@7001 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-26 23:10:53 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
in_signal_flow = false;
add muted-by-other concept to solo support infrastructure git-svn-id: svn://localhost/ardour2/branches/3.0@7005 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-27 13:10:04 -04:00
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("check feed from %1\n", (*i)->name()));
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if ((*i)->feeds (route, &via_sends_only)) {
add more DEBUG::Solo output ; fix assignment of remote control ID's from XML (#5079) git-svn-id: svn://localhost/ardour2/branches/3.0@13542 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-21 14:56:55 -05:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("\tthere is a feed from %1\n", (*i)->name()));
enable ignore-sends-for-solo-propagation change git-svn-id: svn://localhost/ardour2/branches/3.0@6114 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-17 18:11:42 -05:00
if (!via_sends_only) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (!route->soloed_by_others_upstream()) {
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
(*i)->solo_control()->mod_solo_by_others_downstream (delta);
more solo propagation fixes. handle cases Track(s) -> Bus1 -> Bus2
2015-10-06 08:14:44 -04:00
} else {
DEBUG_TRACE (DEBUG::Solo, "\talready soloed by others upstream\n");
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
add more DEBUG::Solo output ; fix assignment of remote control ID's from XML (#5079) git-svn-id: svn://localhost/ardour2/branches/3.0@13542 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-21 14:56:55 -05:00
} else {
DEBUG_TRACE (DEBUG::Solo, string_compose ("\tthere is a send-only feed from %1\n", (*i)->name()));
some provisional support for "real" solo-isolate behaviour, subject to more discussions with mr. oofus and others git-svn-id: svn://localhost/ardour2/branches/3.0@6145 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-21 12:20:57 -05:00
}
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
in_signal_flow = true;
} else {
add more DEBUG::Solo output ; fix assignment of remote control ID's from XML (#5079) git-svn-id: svn://localhost/ardour2/branches/3.0@13542 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-21 14:56:55 -05:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("\tno feed from %1\n", (*i)->name()));
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("check feed to %1\n", (*i)->name()));
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (route->feeds (*i, &via_sends_only)) {
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
/* propagate solo upstream only if routing other than
sends is involved, but do consider the other route
(*i) to be part of the signal flow even if only
sends are involved.
*/
DEBUG_TRACE (DEBUG::Solo, string_compose ("%1 feeds %2 via sends only %3 sboD %4 sboU %5\n",
move ever close to working master/slave logic, this time with audio testing
2016-04-20 23:22:29 -04:00
route->name(),
(*i)->name(),
via_sends_only,
route->soloed_by_others_downstream(),
route->soloed_by_others_upstream()));
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
if (!via_sends_only) {
more solo propagation fixes. handle cases Track(s) -> Bus1 -> Bus2
2015-10-06 08:14:44 -04:00
//NB. Triggers Invert Push, which handles soloed by downstream
DEBUG_TRACE (DEBUG::Solo, string_compose ("\tmod %1 by %2\n", (*i)->name(), delta));
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
(*i)->solo_control()->mod_solo_by_others_upstream (delta);
add more DEBUG::Solo output ; fix assignment of remote control ID's from XML (#5079) git-svn-id: svn://localhost/ardour2/branches/3.0@13542 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-21 14:56:55 -05:00
} else {
DEBUG_TRACE (DEBUG::Solo, string_compose ("\tfeed to %1 ignored, sends-only\n", (*i)->name()));
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
in_signal_flow = true;
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
} else {
NO-OP: mixbus codebase unification
2019-04-12 13:22:59 -04:00
DEBUG_TRACE (DEBUG::Solo, string_compose("\tno feed to %1\n", (*i)->name()) );
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
the great solo model simplification (without much code removal) git-svn-id: svn://localhost/ardour2/branches/3.0@7049 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-03 18:07:47 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (!in_signal_flow) {
uninvolved.push_back (*i);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
make handling of sends-only feed-based solo propagation symmetrical upstream and downstream (fixes #4252) and add more DEBUG::Solo tracing for future aid git-svn-id: svn://localhost/ardour2/branches/3.0@9972 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-08-10 11:13:15 -04:00
DEBUG_TRACE (DEBUG::Solo, "propagation complete\n");
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
/* now notify that the mute state of the routes not involved in the signal
pathway of the just-solo-changed route may have altered.
*/
the great solo model simplification (without much code removal) git-svn-id: svn://localhost/ardour2/branches/3.0@7049 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-03 18:07:47 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
for (RouteList::iterator i = uninvolved.begin(); i != uninvolved.end(); ++i) {
add more DEBUG::Solo output ; fix assignment of remote control ID's from XML (#5079) git-svn-id: svn://localhost/ardour2/branches/3.0@13542 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-21 14:56:55 -05:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("mute change for %1, which neither feeds or is fed by %2\n", (*i)->name(), route->name()));
Resolve only active notes when muted/non-soloed. Fixes bug #6206.
2015-03-24 13:47:37 -04:00
(*i)->act_on_mute ();
provisional changes to speed up solo changes to large numbers of routes. Moves global update of solo state and emission of Session::SoloChanged to a single point after 1 to N solo controls are changed. Also avoid unnecessarily emitted Activated() signal for listen controls, though Process::{activate,deactive}() should probably be redesigned to avoid this in a "deeper" way
2016-11-25 05:29:16 -05:00
/* Session will emit SoloChanged() after all solo changes are
* complete, which should be used by UIs to update mute status
*/
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
void
nick m's fix for markers etc ; several tweaks for mute/solo ; rename run_in_place() as run() git-svn-id: svn://localhost/ardour2/branches/3.0@5155 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-10 14:10:07 -04:00
Session::update_route_solo_state (boost::shared_ptr<RouteList> r)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
how about that ... a monitor/main section .. GUI is still unfinished .. several small fixes to processor loading/listen mgmt and a few debug output lines rmeoved. knob images are provisional, and can be found in icons/knob.png and related files git-svn-id: svn://localhost/ardour2/branches/3.0@6744 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-10 12:31:16 -05:00
/* now figure out if anything that matters is soloed (or is "listening")*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
nick m's fix for markers etc ; several tweaks for mute/solo ; rename run_in_place() as run() git-svn-id: svn://localhost/ardour2/branches/3.0@5155 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-10 14:10:07 -04:00
bool something_soloed = false;
Change last commit to use it's own variable rather than borrowing solo's
2016-01-16 17:40:03 -05:00
bool something_listening = false;
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
uint32_t listeners = 0;
uint32_t isolated = 0;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
nick m's fix for markers etc ; several tweaks for mute/solo ; rename run_in_place() as run() git-svn-id: svn://localhost/ardour2/branches/3.0@5155 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-10 14:10:07 -04:00
if (!r) {
r = routes.reader();
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
some tweaks for solo logic to get things working as they were before (correctly)
2016-04-10 18:20:11 -04:00
if ((*i)->can_solo()) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (Config->get_solo_control_is_listen_control()) {
provisional changes to speed up solo changes to large numbers of routes. Moves global update of solo state and emission of Session::SoloChanged to a single point after 1 to N solo controls are changed. Also avoid unnecessarily emitted Activated() signal for listen controls, though Process::{activate,deactive}() should probably be redesigned to avoid this in a "deeper" way
2016-11-25 05:29:16 -05:00
if ((*i)->solo_control()->soloed_by_self_or_masters()) {
some tweaks for solo logic to get things working as they were before (correctly)
2016-04-10 18:20:11 -04:00
listeners++;
something_listening = true;
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
} else {
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
(*i)->set_listen (false);
provisional changes to speed up solo changes to large numbers of routes. Moves global update of solo state and emission of Session::SoloChanged to a single point after 1 to N solo controls are changed. Also avoid unnecessarily emitted Activated() signal for listen controls, though Process::{activate,deactive}() should probably be redesigned to avoid this in a "deeper" way
2016-11-25 05:29:16 -05:00
if ((*i)->can_solo() && (*i)->solo_control()->soloed_by_self_or_masters()) {
some tweaks for solo logic to get things working as they were before (correctly)
2016-04-10 18:20:11 -04:00
something_soloed = true;
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
}
lots of odds and ends to do with solo isolate and its GUI git-svn-id: svn://localhost/ardour2/branches/3.0@7072 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-06 14:40:37 -04:00
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
if ((*i)->solo_isolate_control()->solo_isolated()) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
isolated++;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (something_soloed != _non_soloed_outs_muted) {
_non_soloed_outs_muted = something_soloed;
SoloActive (_non_soloed_outs_muted); /* EMIT SIGNAL */
}
the last (?) piece of the internal send/listen/monitor/control outs track/bus architecture puzzle: a send from the post-fader master bus to the control/listen/monitor bus inputs that is silent when something is soloed, but the rest of the time delivers the main mix to the control/listen/monitor bus. Tweaks to follow, surely ... git-svn-id: svn://localhost/ardour2/branches/3.0@6740 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-06 14:47:34 -05:00
Change last commit to use it's own variable rather than borrowing solo's
2016-01-16 17:40:03 -05:00
if (something_listening != _listening) {
_listening = something_listening;
SoloActive (_listening);
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
_listen_cnt = listeners;
lots of odds and ends to do with solo isolate and its GUI git-svn-id: svn://localhost/ardour2/branches/3.0@7072 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-06 14:40:37 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (isolated != _solo_isolated_cnt) {
_solo_isolated_cnt = isolated;
IsolatedChanged (); /* EMIT SIGNAL */
}
add more DEBUG::Solo output ; fix assignment of remote control ID's from XML (#5079) git-svn-id: svn://localhost/ardour2/branches/3.0@13542 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-21 14:56:55 -05:00
DEBUG_TRACE (DEBUG::Solo, string_compose ("solo state updated by session, soloed? %1 listeners %2 isolated %3\n",
something_soloed, listeners, isolated));
provisional changes to speed up solo changes to large numbers of routes. Moves global update of solo state and emission of Session::SoloChanged to a single point after 1 to N solo controls are changed. Also avoid unnecessarily emitted Activated() signal for listen controls, though Process::{activate,deactive}() should probably be redesigned to avoid this in a "deeper" way
2016-11-25 05:29:16 -05:00
SoloChanged (); /* EMIT SIGNAL */
set_dirty();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Move special-cased FP8 mute-state into libardour
2017-05-05 10:47:25 -04:00
bool
Session::muted () const
{
// TODO consider caching the value on every MuteChanged signal,
// Note that API users may also subscribe to MuteChanged and hence
// this method needs to be called first.
bool muted = false;
StripableList all;
get_stripables (all);
for (StripableList::const_iterator i = all.begin(); i != all.end(); ++i) {
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
assert (!(*i)->is_auditioner()); // XXX remove me
if ((*i)->is_monitor()) {
Move special-cased FP8 mute-state into libardour
2017-05-05 10:47:25 -04:00
continue;
}
boost::shared_ptr<Route> r = boost::dynamic_pointer_cast<Route>(*i);
if (r && !r->active()) {
continue;
}
boost::shared_ptr<MuteControl> mc = (*i)->mute_control();
if (mc && mc->muted ()) {
muted = true;
break;
}
}
return muted;
}
std::vector<boost::weak_ptr<AutomationControl> >
Session::cancel_all_mute ()
{
StripableList all;
get_stripables (all);
std::vector<boost::weak_ptr<AutomationControl> > muted;
boost::shared_ptr<ControlList> cl (new ControlList);
for (StripableList::const_iterator i = all.begin(); i != all.end(); ++i) {
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
assert (!(*i)->is_auditioner());
if ((*i)->is_monitor()) {
Move special-cased FP8 mute-state into libardour
2017-05-05 10:47:25 -04:00
continue;
}
boost::shared_ptr<Route> r = boost::dynamic_pointer_cast<Route> (*i);
if (r && !r->active()) {
continue;
}
boost::shared_ptr<AutomationControl> ac = (*i)->mute_control();
if (ac && ac->get_value () > 0) {
cl->push_back (ac);
muted.push_back (boost::weak_ptr<AutomationControl>(ac));
}
}
if (!cl->empty ()) {
set_controls (cl, 0.0, PBD::Controllable::UseGroup);
}
return muted;
}
add Session::get_stripables()
2016-05-16 16:43:00 -04:00
void
get_stripables and get_routelist are now presentationinfo aware
2018-11-07 16:02:49 -05:00
Session::get_stripables (StripableList& sl, PresentationInfo::Flag fl) const
add Session::get_stripables()
2016-05-16 16:43:00 -04:00
{
boost::shared_ptr<RouteList> r = routes.reader ();
get_stripables and get_routelist are now presentationinfo aware
2018-11-07 16:02:49 -05:00
for (RouteList::iterator it = r->begin(); it != r->end(); ++it) {
if ((*it)->presentation_info ().flags () & fl) {
sl.push_back (*it);
}
}
add Session::get_stripables()
2016-05-16 16:43:00 -04:00
get_stripables and get_routelist are now presentationinfo aware
2018-11-07 16:02:49 -05:00
if (fl & PresentationInfo::VCA) {
VCAList v = _vca_manager->vcas ();
sl.insert (sl.end(), v.begin(), v.end());
}
add Session::get_stripables()
2016-05-16 16:43:00 -04:00
}
Add Lua bindings to query all stripables
2017-06-16 22:35:48 -04:00
StripableList
Session::get_stripables () const
{
get_stripables and get_routelist are now presentationinfo aware
2018-11-07 16:02:49 -05:00
PresentationInfo::Flag fl = PresentationInfo::AllStripables;
Add Lua bindings to query all stripables
2017-06-16 22:35:48 -04:00
StripableList rv;
get_stripables and get_routelist are now presentationinfo aware
2018-11-07 16:02:49 -05:00
Session::get_stripables (rv, fl);
Add Lua bindings to query all stripables
2017-06-16 22:35:48 -04:00
rv.sort (Stripable::Sorter ());
return rv;
}
Add method to conveniently retrieve a sorted routelist
2018-08-02 15:02:31 -04:00
RouteList
get_stripables and get_routelist are now presentationinfo aware
2018-11-07 16:02:49 -05:00
Session::get_routelist (bool mixer_order, PresentationInfo::Flag fl) const
Add method to conveniently retrieve a sorted routelist
2018-08-02 15:02:31 -04:00
{
boost::shared_ptr<RouteList> r = routes.reader ();
RouteList rv;
get_stripables and get_routelist are now presentationinfo aware
2018-11-07 16:02:49 -05:00
for (RouteList::iterator it = r->begin(); it != r->end(); ++it) {
if ((*it)->presentation_info ().flags () & fl) {
rv.push_back (*it);
}
}
Add method to conveniently retrieve a sorted routelist
2018-08-02 15:02:31 -04:00
rv.sort (Stripable::Sorter (mixer_order));
return rv;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
boost::shared_ptr<RouteList>
add track/bus now allows specifying aux or direct bus role; processor box menus sort of kindof allow New Aux -> Aux Bus git-svn-id: svn://localhost/ardour2/branches/3.0@6128 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-18 23:25:46 -05:00
Session::get_routes_with_internal_returns() const
{
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
add track/bus now allows specifying aux or direct bus role; processor box menus sort of kindof allow New Aux -> Aux Bus git-svn-id: svn://localhost/ardour2/branches/3.0@6128 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-18 23:25:46 -05:00
boost::shared_ptr<RouteList> rl (new RouteList);
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
if ((*i)->internal_return ()) {
rl->push_back (*i);
}
}
return rl;
}
commits 6001-6525 inclusive from 2.X applied to 3.0 git-svn-id: svn://localhost/ardour2/branches/3.0@6942 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-20 22:24:38 -04:00
bool
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Session::io_name_is_legal (const std::string& name) const
commits 6001-6525 inclusive from 2.X applied to 3.0 git-svn-id: svn://localhost/ardour2/branches/3.0@6942 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-20 22:24:38 -04:00
{
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Check for reserved i/o that are not routes, fixes #7171
2016-12-10 18:03:44 -05:00
for (map<string,bool>::const_iterator reserved = reserved_io_names.begin(); reserved != reserved_io_names.end(); ++reserved) {
if (name == reserved->first) {
if (!route_by_name (reserved->first)) {
Fix another possible case of illegal use of reserved port-names Session::io_name_is_legal() is used for Routes and Processors: Route::set_name(), ProcessorBox::rename_processor() and and Route::ensure_track_or_route_name() -- it is not used for actual I/O objects.
2017-05-10 14:28:47 -04:00
/* first instance of a reserved name is allowed for some */
return reserved->second;
don't force rename of reserved route/IO names if the name isn't taken yet
2015-11-07 22:15:59 -05:00
}
more correct semantics for reserved IO name checking
2015-11-09 08:24:35 -05:00
/* all other instances of a reserved name are not allowed */
return false;
better more reliable checks on renamed, newly created and imported track/bus names
2015-09-28 14:49:49 -04:00
}
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
if ((*i)->name() == name) {
return false;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if ((*i)->has_io_processor_named (name)) {
return false;
}
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
return true;
commits 6001-6525 inclusive from 2.X applied to 3.0 git-svn-id: svn://localhost/ardour2/branches/3.0@6942 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-20 22:24:38 -04:00
}
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
void
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
Session::set_exclusive_input_active (boost::shared_ptr<RouteList> rl, bool onoff, bool flip_others)
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
{
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
RouteList rl2;
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
vector<string> connections;
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
/* if we are passed only a single route and we're not told to turn
* others off, then just do the simple thing.
*/
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
if (flip_others == false && rl->size() == 1) {
boost::shared_ptr<MidiTrack> mt = boost::dynamic_pointer_cast<MidiTrack> (rl->front());
if (mt) {
mt->set_input_active (onoff);
return;
}
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
}
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
for (RouteList::iterator rt = rl->begin(); rt != rl->end(); ++rt) {
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
PortSet& ps ((*rt)->input()->ports());
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
for (PortSet::iterator p = ps.begin(); p != ps.end(); ++p) {
p->get_connections (connections);
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
for (vector<string>::iterator s = connections.begin(); s != connections.end(); ++s) {
routes_using_input_from (*s, rl2);
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
/* scan all relevant routes to see if others are on or off */
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
bool others_are_already_on = false;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
for (RouteList::iterator r = rl2.begin(); r != rl2.end(); ++r) {
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
boost::shared_ptr<MidiTrack> mt = boost::dynamic_pointer_cast<MidiTrack> (*r);
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
if (!mt) {
continue;
}
if ((*r) != (*rt)) {
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
if (mt->input_active()) {
others_are_already_on = true;
}
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
} else {
/* this one needs changing */
mt->set_input_active (onoff);
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
}
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
if (flip_others) {
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
/* globally reverse other routes */
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
for (RouteList::iterator r = rl2.begin(); r != rl2.end(); ++r) {
if ((*r) != (*rt)) {
boost::shared_ptr<MidiTrack> mt = boost::dynamic_pointer_cast<MidiTrack> (*r);
if (mt) {
mt->set_input_active (!others_are_already_on);
}
}
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
}
}
}
}
void
Session::routes_using_input_from (const string& str, RouteList& rl)
{
add alt-i as a binding in both the editor and mixer windows to toggle the state of MIDI input on the selected track(s); in the mixer this will also operate on the strip under the mouse. fixes #4838 git-svn-id: svn://localhost/ardour2/branches/3.0@13475 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-12 21:19:04 -05:00
boost::shared_ptr<RouteList> r = routes.reader();
ctrl-alt-click now toggles input active status of other MIDI tracks using (any of) the same input(s) as the clicked track git-svn-id: svn://localhost/ardour2/branches/3.0@9789 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-07-03 11:01:21 -04:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
if ((*i)->input()->connected_to (str)) {
rl.push_back (*i);
}
}
}
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<Route>
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Session::route_by_name (string name) const
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
if ((*i)->name() == name) {
return *i;
}
}
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
return boost::shared_ptr<Route> ((Route*) 0);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<Route>
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Session::route_by_id (PBD::ID id) const
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
if ((*i)->id() == id) {
return *i;
}
}
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
return boost::shared_ptr<Route> ((Route*) 0);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
libardour now has CoreSelection object to manage selection status of Stripables and AutomationControls
2017-05-05 07:31:21 -04:00
boost::shared_ptr<Stripable>
Session::stripable_by_id (PBD::ID id) const
{
StripableList sl;
get_stripables (sl);
for (StripableList::const_iterator s = sl.begin(); s != sl.end(); ++s) {
if ((*s)->id() == id) {
return *s;
}
}
return boost::shared_ptr<Stripable>();
}
add processor lookup by PBD::ID via session In preparation for Lua bindings this completes the basic set of Session object lookup: route, source, controllable, region, processor
2016-02-20 14:22:40 -05:00
boost::shared_ptr<Processor>
Session::processor_by_id (PBD::ID id) const
{
boost::shared_ptr<RouteList> r = routes.reader ();
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
boost::shared_ptr<Processor> p = (*i)->Route::processor_by_id (id);
if (p) {
return p;
}
}
return boost::shared_ptr<Processor> ();
}
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<Route>
use correct types in arguments to Session::get_remote_nth...()
2016-05-17 15:26:35 -04:00
Session::get_remote_nth_route (PresentationInfo::order_t n) const
redesign Route and VCA objects to inherit from ARDOUR::Stripable
2016-03-07 14:59:40 -05:00
{
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
return boost::dynamic_pointer_cast<Route> (get_remote_nth_stripable (n, PresentationInfo::Route));
redesign Route and VCA objects to inherit from ARDOUR::Stripable
2016-03-07 14:59:40 -05:00
}
boost::shared_ptr<Stripable>
use correct types in arguments to Session::get_remote_nth...()
2016-05-17 15:26:35 -04:00
Session::get_remote_nth_stripable (PresentationInfo::order_t n, PresentationInfo::Flag flags) const
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
fix implementation of Session::get_remote_nth_stripable()
2016-05-17 15:05:36 -04:00
StripableList sl;
use correct types in arguments to Session::get_remote_nth...()
2016-05-17 15:26:35 -04:00
PresentationInfo::order_t match_cnt = 0;
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
fix implementation of Session::get_remote_nth_stripable()
2016-05-17 15:05:36 -04:00
get_stripables (sl);
Use Stripable::Sorter in libs
2017-06-16 17:44:29 -04:00
sl.sort (Stripable::Sorter());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
fix implementation of Session::get_remote_nth_stripable()
2016-05-17 15:05:36 -04:00
for (StripableList::const_iterator s = sl.begin(); s != sl.end(); ++s) {
make Session::get_remote_nth_stripable() ignore hidden stripables unless asked. this more closely matches the semantics of ye olde get_*_by_rid() from ardour 4.x and earlier
2016-08-01 12:02:30 -04:00
if ((*s)->presentation_info().hidden()) {
/* if the caller didn't explicitly ask for hidden
stripables, ignore hidden ones. This matches
the semantics of the pre-PresentationOrder
"get by RID" logic of Ardour 4.x and earlier.
XXX at some point we should likely reverse
the logic of the flags, because asking for "the
hidden stripables" is not going to be common,
whereas asking for visible ones is normal.
*/
if (! (flags & PresentationInfo::Hidden)) {
continue;
}
}
fix implementation of Session::get_remote_nth_stripable()
2016-05-17 15:05:36 -04:00
if ((*s)->presentation_info().flag_match (flags)) {
if (match_cnt++ == n) {
return *s;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
fix implementation of Session::get_remote_nth_stripable()
2016-05-17 15:05:36 -04:00
/* there is no nth stripable that matches the given flags */
return boost::shared_ptr<Stripable>();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
implement Session::route_by_selected_count()
2017-01-22 13:48:10 -05:00
boost::shared_ptr<Route>
Session::route_by_selected_count (uint32_t id) const
{
RouteList r (*(routes.reader ()));
Use Stripable::Sorter in libs
2017-06-16 17:44:29 -04:00
r.sort (Stripable::Sorter());
implement Session::route_by_selected_count()
2017-01-22 13:48:10 -05:00
RouteList::iterator i;
for (i = r.begin(); i != r.end(); ++i) {
libardour now has CoreSelection object to manage selection status of Stripables and AutomationControls
2017-05-05 07:31:21 -04:00
if ((*i)->is_selected()) {
implement Session::route_by_selected_count()
2017-01-22 13:48:10 -05:00
if (id == 0) {
return *i;
}
--id;
}
}
return boost::shared_ptr<Route> ();
}
implement track numbering
2014-06-25 15:16:09 -04:00
void
Session::reassign_track_numbers ()
{
int64_t tn = 0;
int64_t bn = 0;
RouteList r (*(routes.reader ()));
Use Stripable::Sorter in libs
2017-06-16 17:44:29 -04:00
r.sort (Stripable::Sorter());
implement track numbering
2014-06-25 15:16:09 -04:00
use state Session::StateProtector
2014-06-28 15:36:13 -04:00
StateProtector sp (this);
implement track numbering
2014-06-25 15:16:09 -04:00
for (RouteList::iterator i = r.begin(); i != r.end(); ++i) {
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
assert (!(*i)->is_auditioner());
implement track numbering
2014-06-25 15:16:09 -04:00
if (boost::dynamic_pointer_cast<Track> (*i)) {
(*i)->set_track_number(++tn);
}
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
else if (!(*i)->is_master() && !(*i)->is_monitor()) {
implement track numbering
2014-06-25 15:16:09 -04:00
(*i)->set_track_number(--bn);
}
}
const uint32_t decimals = ceilf (log10f (tn + 1));
const bool decimals_changed = _track_number_decimals != decimals;
_track_number_decimals = decimals;
if (decimals_changed && config.get_track_name_number ()) {
for (RouteList::iterator i = r.begin(); i != r.end(); ++i) {
boost::shared_ptr<Track> t = boost::dynamic_pointer_cast<Track> (*i);
if (t) {
t->resync_track_name();
}
}
// trigger GUI re-layout
config.ParameterChanged("track-name-number");
}
clean up debug output for OrderKeys
2016-05-17 09:36:28 -04:00
#ifndef NDEBUG
if (DEBUG_ENABLED(DEBUG::OrderKeys)) {
boost::shared_ptr<RouteList> rl = routes.reader ();
for (RouteList::iterator i = rl->begin(); i != rl->end(); ++i) {
DEBUG_TRACE (DEBUG::OrderKeys, string_compose ("%1 numbered %2\n", (*i)->name(), (*i)->track_number()));
}
}
#endif /* NDEBUG */
implement track numbering
2014-06-25 15:16:09 -04:00
}
Fix session range location by hooking into the undo system. Fixes #3654. git-svn-id: svn://localhost/ardour2/branches/3.0@8539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:56 -05:00
void
Session::playlist_region_added (boost::weak_ptr<Region> w)
{
boost::shared_ptr<Region> r = w.lock ();
if (!r) {
return;
}
/* These are the operations that are currently in progress... */
list<GQuark> curr = _current_trans_quarks;
curr.sort ();
/* ...and these are the operations during which we want to update
the session range location markers.
*/
list<GQuark> ops;
ops.push_back (Operations::capture);
ops.push_back (Operations::paste);
ops.push_back (Operations::duplicate_region);
ops.push_back (Operations::insert_file);
ops.push_back (Operations::insert_region);
ops.push_back (Operations::drag_region_brush);
ops.push_back (Operations::region_drag);
ops.push_back (Operations::selection_grab);
ops.push_back (Operations::region_fill);
ops.push_back (Operations::fill_selection);
ops.push_back (Operations::create_region);
Fix update of session range location on copy-dragging regions (#4038). git-svn-id: svn://localhost/ardour2/branches/3.0@9504 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-05-12 17:29:27 -04:00
ops.push_back (Operations::region_copy);
ops.push_back (Operations::fixed_time_region_copy);
Fix session range location by hooking into the undo system. Fixes #3654. git-svn-id: svn://localhost/ardour2/branches/3.0@8539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:56 -05:00
ops.sort ();
/* See if any of the current operations match the ones that we want */
list<GQuark> in;
set_intersection (_current_trans_quarks.begin(), _current_trans_quarks.end(), ops.begin(), ops.end(), back_inserter (in));
/* If so, update the session range markers */
if (!in.empty ()) {
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
maybe_update_session_range (r->position (), r->last_sample ());
Fix session range location by hooking into the undo system. Fixes #3654. git-svn-id: svn://localhost/ardour2/branches/3.0@8539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:56 -05:00
}
}
/** Update the session range markers if a is before the current start or
* b is after the current end.
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::maybe_update_session_range (samplepos_t a, samplepos_t b)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (loading ()) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
return;
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
samplepos_t session_end_marker_shift_samples = session_end_shift * _nominal_sample_rate;
merge in Tracks-derived session-end-is-N-seconds beyond last region end, but define N as zero for Ardour/Mixbus
2015-05-13 10:39:24 -04:00
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
if (_session_range_location == 0) {
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
merge in Tracks-derived session-end-is-N-seconds beyond last region end, but define N as zero for Ardour/Mixbus
2015-05-13 10:39:24 -04:00
set_session_range_location (a, b + session_end_marker_shift_samples);
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
} else {
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Session-range behavior (libardour part)
2019-02-06 10:54:13 -05:00
if (_session_range_is_free && (a < _session_range_location->start())) {
Fix session range location by hooking into the undo system. Fixes #3654. git-svn-id: svn://localhost/ardour2/branches/3.0@8539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:56 -05:00
_session_range_location->set_start (a);
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Session-range behavior (libardour part)
2019-02-06 10:54:13 -05:00
if (_session_range_is_free && (b > _session_range_location->end())) {
Fix session range location by hooking into the undo system. Fixes #3654. git-svn-id: svn://localhost/ardour2/branches/3.0@8539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:56 -05:00
_session_range_location->set_end (b);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
once the user has explicitly set the session range end, playlist/range changes do not move it. The user may drag the marker, edit in the Location UI, or use nudge, to set the end
2016-07-12 11:19:15 -04:00
void
Session-range behavior (libardour part)
2019-02-06 10:54:13 -05:00
Session::set_session_range_is_free (bool yn)
once the user has explicitly set the session range end, playlist/range changes do not move it. The user may drag the marker, edit in the Location UI, or use nudge, to set the end
2016-07-12 11:19:15 -04:00
{
Session-range behavior (libardour part)
2019-02-06 10:54:13 -05:00
_session_range_is_free = yn;
once the user has explicitly set the session range end, playlist/range changes do not move it. The user may drag the marker, edit in the Location UI, or use nudge, to set the end
2016-07-12 11:19:15 -04:00
}
Fix session range location by hooking into the undo system. Fixes #3654. git-svn-id: svn://localhost/ardour2/branches/3.0@8539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:56 -05:00
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::playlist_ranges_moved (list<Evoral::RangeMove<samplepos_t> > const & ranges)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
for (list<Evoral::RangeMove<samplepos_t> >::const_iterator i = ranges.begin(); i != ranges.end(); ++i) {
Fix session range location by hooking into the undo system. Fixes #3654. git-svn-id: svn://localhost/ardour2/branches/3.0@8539 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-01-19 12:38:56 -05:00
maybe_update_session_range (i->to, i->to + i->length);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
Update session range on trimming regions. git-svn-id: svn://localhost/ardour2/branches/3.0@9004 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-01 11:23:31 -05:00
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::playlist_regions_extended (list<Evoral::Range<samplepos_t> > const & ranges)
Update session range on trimming regions. git-svn-id: svn://localhost/ardour2/branches/3.0@9004 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-01 11:23:31 -05:00
{
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
for (list<Evoral::Range<samplepos_t> >::const_iterator i = ranges.begin(); i != ranges.end(); ++i) {
Update session range on trimming regions. git-svn-id: svn://localhost/ardour2/branches/3.0@9004 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-01 11:23:31 -05:00
maybe_update_session_range (i->from, i->to);
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* Region management */
boost::shared_ptr<Region>
the mega-properties/SequenceProperty patch. split is broken at present (right hand starts has start-in-source of zero) git-svn-id: svn://localhost/ardour2/branches/3.0@6718 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-01 19:00:00 -05:00
Session::find_whole_file_parent (boost::shared_ptr<Region const> child) const
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
const RegionFactory::RegionMap& regions (RegionFactory::regions());
remove the session region list; GUI now represents (a relatively unfiltered view of) the raw region list that always contains every single region ever created git-svn-id: svn://localhost/ardour2/branches/3.0@6739 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-06 10:40:42 -05:00
RegionFactory::RegionMap::const_iterator i;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
boost::shared_ptr<Region> region;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (region_lock);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (i = regions.begin(); i != regions.end(); ++i) {
region = i->second;
if (region->whole_file()) {
if (child->source_equivalent (region)) {
return region;
}
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
return boost::shared_ptr<Region> ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
int
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
Session::destroy_sources (list<boost::shared_ptr<Source> > srcs)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
set<boost::shared_ptr<Region> > relevant_regions;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
for (list<boost::shared_ptr<Source> >::iterator s = srcs.begin(); s != srcs.end(); ++s) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
RegionFactory::get_regions_using_source (*s, relevant_regions);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
for (set<boost::shared_ptr<Region> >::iterator r = relevant_regions.begin(); r != relevant_regions.end(); ) {
set<boost::shared_ptr<Region> >::iterator tmp;
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
tmp = r;
++tmp;
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
Remove global Session::playlists variable, use getter method (1/2) Global variables that can written by anyone are to be avoided. This also simplifies exposing SessionPlaylists as Lua bindings.
2019-03-19 00:14:00 -04:00
_playlists->destroy_region (*r);
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
RegionFactory::map_remove (*r);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
(*r)->drop_sources ();
(*r)->drop_references ();
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
relevant_regions.erase (r);
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
r = tmp;
}
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
for (list<boost::shared_ptr<Source> >::iterator s = srcs.begin(); s != srcs.end(); ) {
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
{
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock ls (source_lock);
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
/* remove from the main source list */
sources.erase ((*s)->id());
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
(*s)->mark_for_remove ();
(*s)->drop_references ();
(Source List) Source property signals (libardour part)
2018-11-15 10:21:31 -05:00
SourceRemoved(*s);
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
s = srcs.erase (s);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
return 0;
}
int
Session::remove_last_capture ()
{
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
list<boost::shared_ptr<Source> > srcs;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
boost::shared_ptr<RouteList> rl = routes.reader ();
for (RouteList::iterator i = rl->begin(); i != rl->end(); ++i) {
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
if (!tr) {
continue;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
list<boost::shared_ptr<Source> >& l = tr->last_capture_sources();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (!l.empty()) {
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
srcs.insert (srcs.end(), l.begin(), l.end());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
l.clear ();
}
}
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
destroy_sources (srcs);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
save_state (_current_snapshot_name);
return 0;
}
(Source List) Region Tags (libardour part) Rough-in: Region-Tags. More correct implementation of tags property (libardour). Region Tags (libardour part)
2018-10-19 21:02:44 -04:00
void
Session::get_last_capture_sources (std::list<boost::shared_ptr<Source> >& srcs)
{
boost::shared_ptr<RouteList> rl = routes.reader ();
for (RouteList::iterator i = rl->begin(); i != rl->end(); ++i) {
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
if (!tr) {
continue;
}
list<boost::shared_ptr<Source> >& l = tr->last_capture_sources();
if (!l.empty()) {
srcs.insert (srcs.end(), l.begin(), l.end());
l.clear ();
}
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* Source Management */
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::add_source (boost::shared_ptr<Source> source)
{
pair<SourceMap::key_type, SourceMap::mapped_type> entry;
pair<SourceMap::iterator,bool> result;
entry.first = source->id();
entry.second = source;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (source_lock);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
result = sources.insert (entry);
}
if (result.second) {
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
/* yay, new source */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
don't alter session search path frm within FileSource constructor, do it when a source is added to the session; related - don't add sources that are created with announce=false to the session, and as a (new) corollary, no longer cause them to potentially alter the session search path (this applies mostly to sources created for auditioning) git-svn-id: svn://localhost/ardour2/branches/3.0@12456 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-27 14:07:30 -04:00
boost::shared_ptr<FileSource> fs = boost::dynamic_pointer_cast<FileSource> (source);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
don't alter session search path frm within FileSource constructor, do it when a source is added to the session; related - don't add sources that are created with announce=false to the session, and as a (new) corollary, no longer cause them to potentially alter the session search path (this applies mostly to sources created for auditioning) git-svn-id: svn://localhost/ardour2/branches/3.0@12456 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-27 14:07:30 -04:00
if (fs) {
if (!fs->within_session()) {
ensure_search_path_includes (Glib::path_get_dirname (fs->path()), fs->type());
}
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
midway snapshot of work done on managing Region & Source lifetimes correctly. may fix missing MIDI file bug ; save empty playlists because they may be referred to by the history file ; undo commands auto-delete when objects they refer to die (currently not commands built from XML deserialization); Sources now know how many regions are using them for something, meaning that we know if we can delete the files holding any data for the source git-svn-id: svn://localhost/ardour2/branches/3.0@7291 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 16:14:07 -04:00
set_dirty();
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
boost::shared_ptr<AudioFileSource> afs;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if ((afs = boost::dynamic_pointer_cast<AudioFileSource>(source)) != 0) {
if (Config->get_auto_analyse_audio()) {
Analyser::queue_source_for_analysis (source, false);
}
}
Make the session watch sources' DropReferences and drop accordingly. Remove a now-redundant call to remove_source in AudioDiskstream. Make the MidiDiskstream drop references on write sources that it wants rid of, so that the sources are actually destroyed. git-svn-id: svn://localhost/ardour2/branches/3.0@8998 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-28 21:04:50 -05:00
source->DropReferences.connect_same_thread (*this, boost::bind (&Session::remove_source, this, boost::weak_ptr<Source> (source)));
(Source List) Source property signals (libardour part)
2018-11-15 10:21:31 -05:00
SourceAdded(source);
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
void
Session::remove_source (boost::weak_ptr<Source> src)
{
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (deletion_in_progress ()) {
Less subtle attempt to prevent crashes on removing sources during session deletion. git-svn-id: svn://localhost/ardour2/branches/3.0@9028 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-02 06:30:37 -05:00
return;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
SourceMap::iterator i;
boost::shared_ptr<Source> source = src.lock();
if (!source) {
return;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
{
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (source_lock);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if ((i = sources.find (source->id())) != sources.end()) {
sources.erase (i);
(Source List) Source property signals (libardour part)
2018-11-15 10:21:31 -05:00
SourceRemoved(source);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (!in_cleanup () && !loading ()) {
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* save state so we don't end up with a session file
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
* referring to non-existent sources.
*/
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
save_state (_current_snapshot_name);
}
}
boost::shared_ptr<Source>
Session::source_by_id (const PBD::ID& id)
{
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (source_lock);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
SourceMap::iterator i;
boost::shared_ptr<Source> source;
if ((i = sources.find (id)) != sources.end()) {
source = i->second;
}
return source;
}
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
boost::shared_ptr<AudioFileSource>
fix data loss/file deletion caused by mistaken used of Session::source_by_path() to check if an AUDIO filesource with a given path already exists. ::source_by_path() was written for MIDI files only. I fixed the call and renamed the two similar functions (one for audio and one for MIDI) to make it more clear.
2014-05-29 16:27:33 -04:00
Session::audio_source_by_path_and_channel (const string& path, uint16_t chn) const
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
/* Restricted to audio files because only audio sources have channel
as a property.
*/
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (source_lock);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
for (SourceMap::const_iterator i = sources.begin(); i != sources.end(); ++i) {
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
boost::shared_ptr<AudioFileSource> afs
= boost::dynamic_pointer_cast<AudioFileSource>(i->second);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (afs && afs->path() == path && chn == afs->channel()) {
return afs;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
return boost::shared_ptr<AudioFileSource>();
}
boost::shared_ptr<MidiSource>
Extend API to allow calling new_midi_source_path() with source_lock held This is in preparation for cloning MIDI-sources during snapshot save.
2017-06-24 19:23:03 -04:00
Session::midi_source_by_path (const std::string& path, bool need_source_lock) const
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
{
/* Restricted to MIDI files because audio sources require a channel
for unique identification, in addition to a path.
*/
Extend API to allow calling new_midi_source_path() with source_lock held This is in preparation for cloning MIDI-sources during snapshot save.
2017-06-24 19:23:03 -04:00
Glib::Threads::Mutex::Lock lm (source_lock, Glib::Threads::NOT_LOCK);
if (need_source_lock) {
lm.acquire ();
}
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
for (SourceMap::const_iterator s = sources.begin(); s != sources.end(); ++s) {
boost::shared_ptr<MidiSource> ms
= boost::dynamic_pointer_cast<MidiSource>(s->second);
boost::shared_ptr<FileSource> fs
= boost::dynamic_pointer_cast<FileSource>(s->second);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
if (ms && fs && fs->path() == path) {
return ms;
}
}
return boost::shared_ptr<MidiSource>();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
handle multiple imports of the same file better (via better source naming); make session properties editor pretty much work for search paths git-svn-id: svn://localhost/ardour2/branches/3.0@7989 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 17:18:52 -05:00
uint32_t
Session::count_sources_by_origin (const string& path)
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
uint32_t cnt = 0;
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (source_lock);
handle multiple imports of the same file better (via better source naming); make session properties editor pretty much work for search paths git-svn-id: svn://localhost/ardour2/branches/3.0@7989 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 17:18:52 -05:00
for (SourceMap::iterator i = sources.begin(); i != sources.end(); ++i) {
boost::shared_ptr<FileSource> fs
= boost::dynamic_pointer_cast<FileSource>(i->second);
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (fs && fs->origin() == path) {
++cnt;
}
handle multiple imports of the same file better (via better source naming); make session properties editor pretty much work for search paths git-svn-id: svn://localhost/ardour2/branches/3.0@7989 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 17:18:52 -05:00
}
return cnt;
}
rework peakfile handling: - copy old peak-files to new (do not require re-calc) - keep old peak-files (for now, backwards compat) - fix cleanup-sources to remove *new* peak-file - include channel-number in hash (like it was done before) see also 624f76b TODO: add Session > Cleanup > remove/re-create peaks
2015-09-09 18:48:10 -04:00
static string
peak_file_helper (const string& peak_path, const string& file_path, const string& file_base, bool hash) {
if (hash) {
std::string checksum = Glib::Checksum::compute_checksum(Glib::Checksum::CHECKSUM_SHA1, file_path + G_DIR_SEPARATOR + file_base);
return Glib::build_filename (peak_path, checksum + peakfile_suffix);
} else {
return Glib::build_filename (peak_path, file_base + peakfile_suffix);
}
}
remove Glib::ustring from libardour; allow any characters except '/' and '\' in paths (may cause issues when loading creatively named 2.X sessions; fix a couple of details of name collection and usage from the startup dialog git-svn-id: svn://localhost/ardour2/branches/3.0@7772 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-09-14 11:45:21 -04:00
string
use FileSource::within_session() for peak-files
2015-09-10 10:39:35 -04:00
Session::construct_peak_filepath (const string& filepath, const bool in_session, const bool old_peak_name) const
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
{
Fixes case where audiofiles used wrong peakfiles
2015-08-29 07:48:05 -04:00
string interchange_dir_string = string (interchange_dir_name) + G_DIR_SEPARATOR;
if (Glib::path_is_absolute (filepath)) {
change the semantics of AudioFileSource::peak_path() and Session::peak_path() a little. If given a full path that points outside the current session but is within another ardour session, attempt to use the peakfiles present in that other session.
2015-05-28 19:08:47 -04:00
/* rip the session dir from the audiofile source */
string session_path;
bool in_another_session = true;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Fixes case where audiofiles used wrong peakfiles
2015-08-29 07:48:05 -04:00
if (filepath.find (interchange_dir_string) != string::npos) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Fixes case where audiofiles used wrong peakfiles
2015-08-29 07:48:05 -04:00
session_path = Glib::path_get_dirname (filepath); /* now ends in audiofiles */
change the semantics of AudioFileSource::peak_path() and Session::peak_path() a little. If given a full path that points outside the current session but is within another ardour session, attempt to use the peakfiles present in that other session.
2015-05-28 19:08:47 -04:00
session_path = Glib::path_get_dirname (session_path); /* now ends in session name */
session_path = Glib::path_get_dirname (session_path); /* now ends in interchange */
session_path = Glib::path_get_dirname (session_path); /* now has session path */
/* see if it is within our session */
for (vector<space_and_path>::const_iterator i = session_dirs.begin(); i != session_dirs.end(); ++i) {
if (i->path == session_path) {
in_another_session = false;
break;
}
}
} else {
in_another_session = false;
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
change the semantics of AudioFileSource::peak_path() and Session::peak_path() a little. If given a full path that points outside the current session but is within another ardour session, attempt to use the peakfiles present in that other session.
2015-05-28 19:08:47 -04:00
if (in_another_session) {
SessionDirectory sd (session_path);
use FileSource::within_session() for peak-files
2015-09-10 10:39:35 -04:00
return peak_file_helper (sd.peak_path(), "", Glib::path_get_basename (filepath), !old_peak_name);
change the semantics of AudioFileSource::peak_path() and Session::peak_path() a little. If given a full path that points outside the current session but is within another ardour session, attempt to use the peakfiles present in that other session.
2015-05-28 19:08:47 -04:00
}
}
rework peakfile handling: - copy old peak-files to new (do not require re-calc) - keep old peak-files (for now, backwards compat) - fix cleanup-sources to remove *new* peak-file - include channel-number in hash (like it was done before) see also 624f76b TODO: add Session > Cleanup > remove/re-create peaks
2015-09-09 18:48:10 -04:00
/* 1) if file belongs to this session
* it may be a relative path (interchange/...)
* or just basename (session_state, remove source)
* -> just use the basename
*/
Fixes case where audiofiles used wrong peakfiles
2015-08-29 07:48:05 -04:00
std::string filename = Glib::path_get_basename (filepath);
std::string path;
rework peakfile handling: - copy old peak-files to new (do not require re-calc) - keep old peak-files (for now, backwards compat) - fix cleanup-sources to remove *new* peak-file - include channel-number in hash (like it was done before) see also 624f76b TODO: add Session > Cleanup > remove/re-create peaks
2015-09-09 18:48:10 -04:00
/* 2) if the file is outside our session dir:
* (imported but not copied) add the path for check-summming */
use FileSource::within_session() for peak-files
2015-09-10 10:39:35 -04:00
if (!in_session) {
Fixes case where audiofiles used wrong peakfiles
2015-08-29 07:48:05 -04:00
path = Glib::path_get_dirname (filepath);
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
use FileSource::within_session() for peak-files
2015-09-10 10:39:35 -04:00
return peak_file_helper (_session_dir->peak_path(), path, Glib::path_get_basename (filepath), !old_peak_name);
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
}
string
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
Session::new_audio_source_path_for_embedded (const std::string& path)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
/* embedded source:
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
*
* we know that the filename is already unique because it exists
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
* out in the filesystem.
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
*
* However, when we bring it into the session, we could get a
* collision.
*
* Eg. two embedded files:
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
*
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
* /foo/bar/baz.wav
* /frob/nic/baz.wav
*
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
* When merged into session, these collide.
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
*
* There will not be a conflict with in-memory sources
* because when the source was created we already picked
* a unique name for it.
*
* This collision is not likely to be common, but we have to guard
* against it. So, if there is a collision, take the md5 hash of the
* the path, and use that as the filename instead.
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
SessionDirectory sdir (get_best_session_directory_for_new_audio());
string base = Glib::path_get_basename (path);
string newpath = Glib::build_filename (sdir.sound_path(), base);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
if (Glib::file_test (newpath, Glib::FILE_TEST_EXISTS)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
MD5 md5;
unique filenames across all searched folders
2014-07-02 17:35:48 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
md5.digestString (path.c_str());
md5.writeToString ();
base = md5.digestChars;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
when merging external source files into session folder, if a collision occurs, use the original file suffix in conjunction with md5 hash
2014-07-08 12:25:36 -04:00
string ext = get_suffix (path);
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
when merging external source files into session folder, if a collision occurs, use the original file suffix in conjunction with md5 hash
2014-07-08 12:25:36 -04:00
if (!ext.empty()) {
base += '.';
base += ext;
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
newpath = Glib::build_filename (sdir.sound_path(), base);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
/* if this collides, we're screwed */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
if (Glib::file_test (newpath, Glib::FILE_TEST_EXISTS)) {
error << string_compose (_("Merging embedded file %1: name collision AND md5 hash collision!"), path) << endmsg;
return string();
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
return newpath;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
/** Return true if there are no audio file sources that use @param name as
* the filename component of their path.
fix logic of Session::audio_source_name_is_unique()
2015-02-20 14:12:12 -05:00
*
* Return false otherwise.
*
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
* This method MUST ONLY be used to check in-session, mono files since it
fix logic of Session::audio_source_name_is_unique()
2015-02-20 14:12:12 -05:00
* hard-codes the channel of the audio file source we are looking for as zero.
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
*
fix logic of Session::audio_source_name_is_unique()
2015-02-20 14:12:12 -05:00
* If/when Ardour supports native files in non-mono formats, the logic here
* will need to be revisited.
*/
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
bool
fix logic of Session::audio_source_name_is_unique()
2015-02-20 14:12:12 -05:00
Session::audio_source_name_is_unique (const string& name)
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
{
std::vector<string> sdirs = source_search_path (DataType::AUDIO);
vector<space_and_path>::iterator i;
uint32_t existing = 0;
second (and hopefully) final part of changes to respond to header format changes sensibly: lookup existing files correctly, and don't end up with gapped "take" numbers for successive files since we now remove ::removable() sources when switching to new ones (for audio) git-svn-id: svn://localhost/ardour2/branches/3.0@7470 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-22 10:52:05 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
for (vector<string>::const_iterator i = sdirs.begin(); i != sdirs.end(); ++i) {
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
/* note that we search *without* the extension so that
we don't end up both "Audio 1-1.wav" and "Audio 1-1.caf"
in the event that this new name is required for
a file format change.
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
const string spath = *i;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
do not call basename_nosuffix() before using matching_unsuffixed_filename_exists_in() because the latter does that call itself. As the code used to be, if we are testing for a candidate filename of aaa.bbb.ccc, the call strips .ccc and matching_... strips .bbb resulting in a match with aaa, which is not intended at all.
2014-07-10 08:16:27 -04:00
if (matching_unsuffixed_filename_exists_in (spath, name)) {
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
existing++;
break;
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
/* it is possible that we have the path already
* assigned to a source that has not yet been written
* (ie. the write source for a diskstream). we have to
* check this in order to make sure that our candidate
* path isn't used again, because that can lead to
* two Sources point to the same file with different
* notions of their removability.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
string possible_path = Glib::build_filename (spath, name);
second (and hopefully) final part of changes to respond to header format changes sensibly: lookup existing files correctly, and don't end up with gapped "take" numbers for successive files since we now remove ::removable() sources when switching to new ones (for audio) git-svn-id: svn://localhost/ardour2/branches/3.0@7470 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-22 10:52:05 -04:00
fix logic of Session::audio_source_name_is_unique()
2015-02-20 14:12:12 -05:00
if (audio_source_by_path_and_channel (possible_path, 0)) {
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
existing++;
break;
}
}
do not call basename_nosuffix() before using matching_unsuffixed_filename_exists_in() because the latter does that call itself. As the code used to be, if we are testing for a candidate filename of aaa.bbb.ccc, the call strips .ccc and matching_... strips .bbb resulting in a match with aaa, which is not intended at all.
2014-07-10 08:16:27 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
return (existing == 0);
}
check whether a source with the same path exists in session list not just on disk. This is intended to catch a potential wierd condition where a source exists as a stub, but has not yet been written to disk, and then a new source is created with the same path. Currently it is not understood how/when this might happen
2014-04-13 11:35:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
string
Session::format_audio_source_name (const string& legalized_base, uint32_t nchan, uint32_t chan, bool destructive, bool take_required, uint32_t cnt, bool related_exists)
{
ostringstream sstr;
const string ext = native_header_format_extension (config.get_native_file_header_format(), DataType::AUDIO);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
sstr << legalized_base;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
goodbye Profile->...trx
2019-09-25 15:02:31 -04:00
if (take_required || related_exists) {
sstr << '-';
sstr << cnt;
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
if (nchan == 2) {
if (chan == 0) {
sstr << "%L";
} else {
sstr << "%R";
}
} else if (nchan > 2) {
if (nchan < 26) {
sstr << '%';
sstr << 'a' + chan;
} else {
/* XXX what? more than 26 channels! */
sstr << '%';
sstr << chan+1;
}
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
sstr << ext;
more work on linking file existence and removability
2014-04-14 03:03:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
return sstr.str();
}
check whether a source with the same path exists in session list not just on disk. This is intended to catch a potential wierd condition where a source exists as a stub, but has not yet been written to disk, and then a new source is created with the same path. Currently it is not understood how/when this might happen
2014-04-13 11:35:49 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
/** Return a unique name based on \a base for a new internal audio source */
string
Session::new_audio_source_path (const string& base, uint32_t nchan, uint32_t chan, bool destructive, bool take_required)
{
uint32_t cnt;
string possible_name;
const uint32_t limit = 9999; // arbitrary limit on number of files with the same basic name
string legalized;
bool some_related_source_name_exists = false;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
legalized = legalize_for_path (base);
// Find a "version" of the base name that doesn't exist in any of the possible directories.
for (cnt = (destructive ? ++destructive_index : 1); cnt <= limit; ++cnt) {
possible_name = format_audio_source_name (legalized, nchan, chan, destructive, take_required, cnt, some_related_source_name_exists);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
fix logic of Session::audio_source_name_is_unique()
2015-02-20 14:12:12 -05:00
if (audio_source_name_is_unique (possible_name)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
break;
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
some_related_source_name_exists = true;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (cnt > limit) {
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
error << string_compose(
_("There are already %1 recordings for %2, which I consider too many."),
limit, base) << endmsg;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
destroy ();
throw failed_constructor();
}
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
/* We've established that the new name does not exist in any session
* directory, so now find out which one we should use for this new
* audio source.
*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
SessionDirectory sdir (get_best_session_directory_for_new_audio());
std::string s = Glib::build_filename (sdir.sound_path(), possible_name);
carl's wondrous DnD VBox patch - processor boxes are now vboxes and not listviews - which included a couple of minor marker/time axis view patches for opaqueness etc. git-svn-id: svn://localhost/ardour2/branches/3.0@6174 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-25 09:37:20 -05:00
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
return s;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Fix crash when copying MIDI regions with midi-copy-is-fork is on (#6002).
2014-11-14 23:53:59 -05:00
/** Return a unique name based on `base` for a new internal MIDI source */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
string
Extend API to allow calling new_midi_source_path() with source_lock held This is in preparation for cloning MIDI-sources during snapshot save.
2017-06-24 19:23:03 -04:00
Session::new_midi_source_path (const string& base, bool need_lock)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
string possible_path;
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
string possible_name;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Fix bumping .mid file name (snapshots & playlist copy) When forking regions, copying playlists or saving snapshots we do not have a reference to the track and cannot use the track's name as basis for the new filename like Editor::fork_region() does. A cloned midi region's name is based on the original region name. This prevents endless addition "name-1-1-1-1-1-1-1-1.mid", adding to the region's basename.
2017-09-07 00:05:26 -04:00
possible_name = legalize_for_path (base);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
// Find a "version" of the file name that doesn't exist in any of the possible directories.
unique filenames across all searched folders
2014-07-02 17:35:48 -04:00
std::vector<string> sdirs = source_search_path(DataType::MIDI);
/* - the main session folder is the first in the vector.
* - after checking all locations for file-name uniqueness,
* we keep the one from the last iteration as new file name
* - midi files are small and should just be kept in the main session-folder
*
* -> reverse the array, check main session folder last and use that as location
* for MIDI files.
*/
std::reverse(sdirs.begin(), sdirs.end());
redesign technique for naming/creating regions for MIDI clone (or other non-capture driven MIDI region creation operations). See comments in Session::new_midi_source_name() for details.
2014-04-10 08:58:04 -04:00
Fix bumping .mid file name (snapshots & playlist copy) When forking regions, copying playlists or saving snapshots we do not have a reference to the track and cannot use the track's name as basis for the new filename like Editor::fork_region() does. A cloned midi region's name is based on the original region name. This prevents endless addition "name-1-1-1-1-1-1-1-1.mid", adding to the region's basename.
2017-09-07 00:05:26 -04:00
while (true) {
possible_name = bump_name_once (possible_name, '-');
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
vector<space_and_path>::iterator i;
uint32_t existing = 0;
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
unique filenames across all searched folders
2014-07-02 17:35:48 -04:00
for (vector<string>::const_iterator i = sdirs.begin(); i != sdirs.end(); ++i) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Fix bumping .mid file name (snapshots & playlist copy) When forking regions, copying playlists or saving snapshots we do not have a reference to the track and cannot use the track's name as basis for the new filename like Editor::fork_region() does. A cloned midi region's name is based on the original region name. This prevents endless addition "name-1-1-1-1-1-1-1-1.mid", adding to the region's basename.
2017-09-07 00:05:26 -04:00
possible_path = Glib::build_filename (*i, possible_name + ".mid");
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
if (Glib::file_test (possible_path, Glib::FILE_TEST_EXISTS)) {
existing++;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Extend API to allow calling new_midi_source_path() with source_lock held This is in preparation for cloning MIDI-sources during snapshot save.
2017-06-24 19:23:03 -04:00
if (midi_source_by_path (possible_path, need_lock)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
existing++;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Fix bumping .mid file name (snapshots & playlist copy) When forking regions, copying playlists or saving snapshots we do not have a reference to the track and cannot use the track's name as basis for the new filename like Editor::fork_region() does. A cloned midi region's name is based on the original region name. This prevents endless addition "name-1-1-1-1-1-1-1-1.mid", adding to the region's basename.
2017-09-07 00:05:26 -04:00
if (possible_path.size () >= PATH_MAX) {
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
error << string_compose(
Fix bumping .mid file name (snapshots & playlist copy) When forking regions, copying playlists or saving snapshots we do not have a reference to the track and cannot use the track's name as basis for the new filename like Editor::fork_region() does. A cloned midi region's name is based on the original region name. This prevents endless addition "name-1-1-1-1-1-1-1-1.mid", adding to the region's basename.
2017-09-07 00:05:26 -04:00
_("There are already many recordings for %1, resulting in a too long file-path %2."),
base, possible_path) << endmsg;
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
destroy ();
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
return 0;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Fix bumping .mid file name (snapshots & playlist copy) When forking regions, copying playlists or saving snapshots we do not have a reference to the track and cannot use the track's name as basis for the new filename like Editor::fork_region() does. A cloned midi region's name is based on the original region name. This prevents endless addition "name-1-1-1-1-1-1-1-1.mid", adding to the region's basename.
2017-09-07 00:05:26 -04:00
if (existing == 0) {
break;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
/* No need to "find best location" for software/app-based RAID, because
MIDI is so small that we always put it in the same place.
*/
return possible_path;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
/** Create a new within-session audio source */
boost::shared_ptr<AudioFileSource>
Session::create_audio_source_for_session (size_t n_chans, string const & base, uint32_t chan, bool destructive)
{
const string path = new_audio_source_path (base, n_chans, chan, destructive, true);
if (!path.empty()) {
return boost::dynamic_pointer_cast<AudioFileSource> (
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
SourceFactory::createWritable (DataType::AUDIO, *this, path, destructive, sample_rate(), true, true));
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
} else {
throw failed_constructor ();
}
}
attempt to remove confusion and errors caused by unclear semantics of _is_embedded for FileSources; member renamed _within_session, and is now ALWAYS determined by the _path of the FileSource, never by the creator git-svn-id: svn://localhost/ardour2/branches/3.0@6213 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-30 08:16:38 -05:00
/** Create a new within-session MIDI source */
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
boost::shared_ptr<MidiSource>
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
Session::create_midi_source_for_session (string const & basic_name)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
const string path = new_midi_source_path (basic_name);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
if (!path.empty()) {
return boost::dynamic_pointer_cast<SMFSource> (
SourceFactory::createWritable (
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
DataType::MIDI, *this, path, false, sample_rate()));
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
} else {
throw failed_constructor ();
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
}
}
GUI-created MIDI regions now steal the pending MIDISource from the track's diskstream, to keep numbering sane; don't create any new MIDI regions if capture collected no data (fixes a crash in my previous commit, and is just logically much more sensible git-svn-id: svn://localhost/ardour2/branches/3.0@7295 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 21:37:24 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
/** Create a new within-session MIDI source */
boost::shared_ptr<MidiSource>
Session::create_midi_source_by_stealing_name (boost::shared_ptr<Track> track)
{
/* the caller passes in the track the source will be used in,
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
so that we can keep the numbering sane.
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
Rationale: a track with the name "Foo" that has had N
captures carried out so far will ALREADY have a write source
named "Foo-N+1.mid" waiting to be used for the next capture.
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
If we call new_midi_source_name() we will get "Foo-N+2". But
there is no region corresponding to "Foo-N+1", so when
"Foo-N+2" appears in the track, the gap presents the user
with odd behaviour - why did it skip past Foo-N+1?
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
We could explain this to the user in some odd way, but
instead we rename "Foo-N+1.mid" as "Foo-N+2.mid", and then
use "Foo-N+1" here.
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
If that attempted rename fails, we get "Foo-N+2.mid" anyway.
*/
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
boost::shared_ptr<MidiTrack> mt = boost::dynamic_pointer_cast<MidiTrack> (track);
assert (mt);
std::string name = track->steal_write_source_name ();
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
redesign technique for naming/creating regions for MIDI clone (or other non-capture driven MIDI region creation operations). See comments in Session::new_midi_source_name() for details.
2014-04-10 08:58:04 -04:00
if (name.empty()) {
dramatic change in logic and naming for operations related to adding a MIDI region on demand and cloning/unlinking Existing code would cause data loss due to creation of two Source objects referring the same path, one with removable flags and one without. Careful code review suggested a variety of thinkos, function naming problems and other confusion that caused this. I have tried ot extensively comment what is going on with these operations, because it is one key area in which MIDI differs from audio: with audio, capture is the only way to add a new audio region, but for MIDI there are GUI input events that can add a new region.
2014-04-13 10:29:07 -04:00
return boost::shared_ptr<MidiSource>();
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
}
GUI-created MIDI regions now steal the pending MIDISource from the track's diskstream, to keep numbering sane; don't create any new MIDI regions if capture collected no data (fixes a crash in my previous commit, and is just logically much more sensible git-svn-id: svn://localhost/ardour2/branches/3.0@7295 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-06-23 21:37:24 -04:00
tentative fix for losing (empty) MIDI files. Incomplete because testing shows issues with some workflows
2014-08-28 14:01:52 -04:00
/* MIDI files are small, just put them in the first location of the
session source search path.
*/
const string path = Glib::build_filename (source_search_path (DataType::MIDI).front(), name);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Fix the horrible mess that was anything related to sources and paths. Most significant changes: - Factor out FileSource from AudioFileSource, use for SMFSource too - Explicitly pass embedded rather than mysterious name mangling or whatever - Destroy a ton of duplicated or very-nearly-duplicated code - Clean up and document all that weird source stuff in session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@4609 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-02-16 21:11:49 -05:00
return boost::dynamic_pointer_cast<SMFSource> (
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
SourceFactory::createWritable (
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
DataType::MIDI, *this, path, false, sample_rate()));
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
(Source List) Region List rewrite (libardour part)
2018-10-16 18:44:11 -04:00
bool
Session::playlist_is_active (boost::shared_ptr<Playlist> playlist)
{
(Source List) merge fixes
2019-05-29 17:18:23 -04:00
Glib::Threads::Mutex::Lock lm (_playlists->lock);
for (SessionPlaylists::List::iterator i = _playlists->playlists.begin(); i != _playlists->playlists.end(); i++) {
(Source List) Region List rewrite (libardour part)
2018-10-16 18:44:11 -04:00
if ( (*i) == playlist ) {
return true;
}
}
return false;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Add Import from session -functionality git-svn-id: svn://localhost/ardour2/branches/3.0@3805 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-09-26 04:29:30 -04:00
void
Session::add_playlist (boost::shared_ptr<Playlist> playlist, bool unused)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
if (playlist->hidden()) {
return;
}
Remove global Session::playlists variable, use getter method (1/2) Global variables that can written by anyone are to be avoided. This also simplifies exposing SessionPlaylists as Lua bindings.
2019-03-19 00:14:00 -04:00
_playlists->add (playlist);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Add Import from session -functionality git-svn-id: svn://localhost/ardour2/branches/3.0@3805 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-09-26 04:29:30 -04:00
if (unused) {
playlist->release();
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
set_dirty();
}
void
Session::remove_playlist (boost::weak_ptr<Playlist> weak_playlist)
{
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (deletion_in_progress ()) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
return;
}
boost::shared_ptr<Playlist> playlist (weak_playlist.lock());
if (!playlist) {
return;
}
Remove global Session::playlists variable, use getter method (1/2) Global variables that can written by anyone are to be avoided. This also simplifies exposing SessionPlaylists as Lua bindings.
2019-03-19 00:14:00 -04:00
_playlists->remove (playlist);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
set_dirty();
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
void
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Session::set_audition (boost::shared_ptr<Region> r)
{
pending_audition_region = r;
monster commit: transport mgmt changes from 2.X (omnibus edition); make slave use nframes64_t ; avoid crashes in Drags when commiting reversible transactions that do not exist git-svn-id: svn://localhost/ardour2/branches/3.0@6034 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-08 11:28:21 -05:00
add_post_transport_work (PostTransportAudition);
Move butler methods from Session to Butler. Slay the dragon. A lil' bit. git-svn-id: svn://localhost/ardour2/branches/3.0@5901 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-23 20:39:28 -04:00
_butler->schedule_transport_work ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
void
Session::audition_playlist ()
{
move Session::Event into SessionEvent class; add SessionEventManager (Session IS-A SessionEventManager); make session ops to toggle all track rec-enable be atomic with respect to process() git-svn-id: svn://localhost/ardour2/branches/3.0@6273 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-03 21:15:12 -05:00
SessionEvent* ev = new SessionEvent (SessionEvent::Audition, SessionEvent::Add, SessionEvent::Immediate, 0, 0.0);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
ev->region.reset ();
queue_event (ev);
}
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
void
Session::register_lua_function (
const std::string& name,
const std::string& script,
const LuaScriptParamList& args
)
{
Glib::Threads::Mutex::Lock lm (lua_lock);
lua_State* L = lua.getState();
const std::string& bytecode = LuaScripting::get_factory_bytecode (script);
luabridge::LuaRef tbl_arg (luabridge::newTable(L));
for (LuaScriptParamList::const_iterator i = args.begin(); i != args.end(); ++i) {
if ((*i)->optional && !(*i)->is_set) { continue; }
tbl_arg[(*i)->name] = (*i)->value;
}
(*_lua_add)(name, bytecode, tbl_arg); // throws luabridge::LuaException
Emit a signal when Lua session script change (prepare UI update to un/set)
2017-02-18 18:06:01 -05:00
lm.release();
LuaScriptsChanged (); /* EMIT SIGNAL */
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
set_dirty();
}
void
Session::unregister_lua_function (const std::string& name)
{
Glib::Threads::Mutex::Lock lm (lua_lock);
(*_lua_del)(name); // throws luabridge::LuaException
lua.collect_garbage ();
Emit a signal when Lua session script change (prepare UI update to un/set)
2017-02-18 18:06:01 -05:00
lm.release();
LuaScriptsChanged (); /* EMIT SIGNAL */
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
set_dirty();
}
std::vector<std::string>
Session::registered_lua_functions ()
{
Glib::Threads::Mutex::Lock lm (lua_lock);
std::vector<std::string> rv;
try {
luabridge::LuaRef list ((*_lua_list)());
for (luabridge::Iterator i (list); !i.isNil (); ++i) {
if (!i.key ().isString ()) { assert(0); continue; }
rv.push_back (i.key ().cast<std::string> ());
}
Lua may call C++ functions with throw. Catch them
2017-08-18 19:05:18 -04:00
} catch (...) { }
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
return rv;
}
#ifndef NDEBUG
static void _lua_print (std::string s) {
std::cout << "SessionLua: " << s << "\n";
}
#endif
void
Session::try_run_lua (pframes_t nframes)
{
if (_n_lua_scripts == 0) return;
Glib::Threads::Mutex::Lock tm (lua_lock, Glib::Threads::TRY_LOCK);
if (tm.locked ()) {
Lua may call C++ functions with throw. Catch them
2017-08-18 19:05:18 -04:00
try { (*_lua_run)(nframes); } catch (...) { }
Lua: Lock bindings into memory for rt-scripts Empirically this decreases gc-spike duration (worst-case) by a factor of two and speeds up the average gc-run by a factor of over 4 (depending on the amount of memory used by the plugin).
2018-03-18 21:26:24 -04:00
lua.collect_garbage_step ();
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
}
}
void
Session::setup_lua ()
{
#ifndef NDEBUG
lua.Print.connect (&_lua_print);
#endif
Centralize Lua sandboxing
2017-08-08 22:07:23 -04:00
lua.sandbox (true);
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
lua.do_command (
"function ArdourSession ()"
" local self = { scripts = {}, instances = {} }"
""
" local remove = function (n)"
" self.scripts[n] = nil"
" self.instances[n] = nil"
" Session:scripts_changed()" // call back
" end"
""
" local addinternal = function (n, f, a)"
" assert(type(n) == 'string', 'function-name must be string')"
" assert(type(f) == 'function', 'Given script is a not a function')"
" assert(type(a) == 'table' or type(a) == 'nil', 'Given argument is invalid')"
" assert(self.scripts[n] == nil, 'Callback \"'.. n ..'\" already exists.')"
" self.scripts[n] = { ['f'] = f, ['a'] = a }"
Whitelist bit32 Lua library for session rt-scripts
2017-01-01 18:46:01 -05:00
" local env = { print = print, tostring = tostring, assert = assert, ipairs = ipairs, error = error, select = select, string = string, type = type, tonumber = tonumber, collectgarbage = collectgarbage, pairs = pairs, math = math, table = table, pcall = pcall, bit32=bit32, Session = Session, PBD = PBD, Timecode = Timecode, Evoral = Evoral, C = C, ARDOUR = ARDOUR }"
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
" self.instances[n] = load (string.dump(f, true), nil, nil, env)(a)"
" Session:scripts_changed()" // call back
" end"
""
" local add = function (n, b, a)"
" assert(type(b) == 'string', 'ByteCode must be string')"
" load (b)()" // assigns f
" assert(type(f) == 'string', 'Assigned ByteCode must be string')"
" addinternal (n, load(f), a)"
" end"
""
" local run = function (...)"
" for n, s in pairs (self.instances) do"
" local status, err = pcall (s, ...)"
" if not status then"
" print ('fn \"'.. n .. '\": ', err)"
" remove (n)"
" end"
" end"
" collectgarbage()"
" end"
""
" local cleanup = function ()"
" self.scripts = nil"
" self.instances = nil"
" end"
""
" local list = function ()"
" local rv = {}"
" for n, _ in pairs (self.scripts) do"
" rv[n] = true"
" end"
" return rv"
" end"
""
" local function basic_serialize (o)"
" if type(o) == \"number\" then"
" return tostring(o)"
" else"
" return string.format(\"%q\", o)"
" end"
" end"
""
" local function serialize (name, value)"
" local rv = name .. ' = '"
" collectgarbage()"
" if type(value) == \"number\" or type(value) == \"string\" or type(value) == \"nil\" then"
" return rv .. basic_serialize(value) .. ' '"
" elseif type(value) == \"table\" then"
" rv = rv .. '{} '"
" for k,v in pairs(value) do"
" local fieldname = string.format(\"%s[%s]\", name, basic_serialize(k))"
" rv = rv .. serialize(fieldname, v) .. ' '"
" collectgarbage()" // string concatenation allocates a new string :(
" end"
" return rv;"
" elseif type(value) == \"function\" then"
" return rv .. string.format(\"%q\", string.dump(value, true))"
" else"
" error('cannot save a ' .. type(value))"
" end"
" end"
""
""
" local save = function ()"
" return (serialize('scripts', self.scripts))"
" end"
""
" local restore = function (state)"
" self.scripts = {}"
" load (state)()"
" for n, s in pairs (scripts) do"
" addinternal (n, load(s['f']), s['a'])"
" end"
" end"
""
" return { run = run, add = add, remove = remove,"
" list = list, restore = restore, save = save, cleanup = cleanup}"
" end"
" "
" sess = ArdourSession ()"
" ArdourSession = nil"
" "
"function ardour () end"
);
lua_State* L = lua.getState();
try {
luabridge::LuaRef lua_sess = luabridge::getGlobal (L, "sess");
lua.do_command ("sess = nil"); // hide it.
lua.do_command ("collectgarbage()");
_lua_run = new luabridge::LuaRef(lua_sess["run"]);
_lua_add = new luabridge::LuaRef(lua_sess["add"]);
_lua_del = new luabridge::LuaRef(lua_sess["remove"]);
_lua_list = new luabridge::LuaRef(lua_sess["list"]);
_lua_save = new luabridge::LuaRef(lua_sess["save"]);
_lua_load = new luabridge::LuaRef(lua_sess["restore"]);
_lua_cleanup = new luabridge::LuaRef(lua_sess["cleanup"]);
} catch (luabridge::LuaException const& e) {
fatal << string_compose (_("programming error: %1"),
Lua may call C++ functions with throw. Catch them
2017-08-18 19:05:18 -04:00
std::string ("Failed to setup session Lua interpreter") + e.what ())
<< endmsg;
abort(); /*NOTREACHED*/
} catch (...) {
fatal << string_compose (_("programming error: %1"),
X_("Failed to setup session Lua interpreter"))
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
<< endmsg;
abort(); /*NOTREACHED*/
}
Lua: Lock bindings into memory for rt-scripts Empirically this decreases gc-spike duration (worst-case) by a factor of two and speeds up the average gc-run by a factor of over 4 (depending on the amount of memory used by the plugin).
2018-03-18 21:26:24 -04:00
lua_mlock (L, 1);
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
LuaBindings::stddef (L);
LuaBindings::common (L);
LuaBindings::dsp (L);
Lua: Lock bindings into memory for rt-scripts Empirically this decreases gc-spike duration (worst-case) by a factor of two and speeds up the average gc-run by a factor of over 4 (depending on the amount of memory used by the plugin).
2018-03-18 21:26:24 -04:00
lua_mlock (L, 0);
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
luabridge::push <Session *> (L, this);
lua_setglobal (L, "Session");
}
void
Session::scripts_changed ()
{
assert (!lua_lock.trylock()); // must hold lua_lock
try {
luabridge::LuaRef list ((*_lua_list)());
int cnt = 0;
for (luabridge::Iterator i (list); !i.isNil (); ++i) {
if (!i.key ().isString ()) { assert(0); continue; }
++cnt;
}
_n_lua_scripts = cnt;
} catch (luabridge::LuaException const& e) {
Lua may call C++ functions with throw. Catch them
2017-08-18 19:05:18 -04:00
fatal << string_compose (_("programming error: %1"),
std::string ("Indexing Lua Session Scripts failed.") + e.what ())
<< endmsg;
abort(); /*NOTREACHED*/
} catch (...) {
Implement Lua session-scripts
2016-02-23 09:41:21 -05:00
fatal << string_compose (_("programming error: %1"),
X_("Indexing Lua Session Scripts failed."))
<< endmsg;
abort(); /*NOTREACHED*/
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::non_realtime_set_audition ()
{
Remove unused playlist audition code. git-svn-id: svn://localhost/ardour2/branches/3.0@11897 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-04-11 11:10:12 -04:00
assert (pending_audition_region);
auditioner->audition_region (pending_audition_region);
pending_audition_region.reset ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
AuditionActive (true); /* EMIT SIGNAL */
}
void
Session::audition_region (boost::shared_ptr<Region> r)
{
move Session::Event into SessionEvent class; add SessionEventManager (Session IS-A SessionEventManager); make session ops to toggle all track rec-enable be atomic with respect to process() git-svn-id: svn://localhost/ardour2/branches/3.0@6273 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-03 21:15:12 -05:00
SessionEvent* ev = new SessionEvent (SessionEvent::Audition, SessionEvent::Add, SessionEvent::Immediate, 0, 0.0);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
ev->region = r;
queue_event (ev);
}
void
Session::cancel_audition ()
{
clean up after monitoring-section on session close
2014-06-29 22:04:35 -04:00
if (!auditioner) {
return;
}
would you look at that! all those changes just to make the auditioner work again .... git-svn-id: svn://localhost/ardour2/branches/3.0@6794 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-24 23:40:07 -04:00
if (auditioner->auditioning()) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
auditioner->cancel_audition ();
AuditionActive (false); /* EMIT SIGNAL */
}
}
bool
Session::is_auditioning () const
{
/* can be called before we have an auditioner object */
if (auditioner) {
would you look at that! all those changes just to make the auditioner work again .... git-svn-id: svn://localhost/ardour2/branches/3.0@6794 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-24 23:40:07 -04:00
return auditioner->auditioning();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
} else {
return false;
}
}
void
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
Session::graph_reordered (bool called_from_backend)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
/* don't do this stuff if we are setting up connections
tracing and small fixes to improve object destruction pathways git-svn-id: svn://localhost/ardour2/branches/3.0@6195 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-27 19:49:04 -05:00
from a set_state() call or creating new tracks. Ditto for deletion.
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
*/
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (inital_connect_or_deletion_in_progress () || _adding_routes_in_progress || _reconnecting_routes_in_progress || _route_deletion_in_progress) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
return;
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
resort_routes ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
/* force all diskstreams to update their capture offset values to
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
* reflect any changes in latencies within the graph.
*/
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
update_latency_compensation (true, called_from_backend);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
/** @return Number of samples that there is disk space available to write,
Handle failure of statfs to return available disc space on Samba shares mounted via GVFS, reporting available space as unknown in this case (#4657). git-svn-id: svn://localhost/ardour2/branches/3.0@12674 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-12 12:41:29 -04:00
* if known.
*/
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
boost::optional<samplecnt_t>
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Session::available_capture_duration ()
{
convert from Glib:: to Glib::Threads for all thread-related API git-svn-id: svn://localhost/ardour2/branches/3.0@13084 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-07-25 13:48:55 -04:00
Glib::Threads::Mutex::Lock lm (space_lock);
Fix missing lock causing erroneous disk space reports when dragging audio clocks. git-svn-id: svn://localhost/ardour2/branches/3.0@12961 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-28 12:08:48 -04:00
Handle failure of statfs to return available disc space on Samba shares mounted via GVFS, reporting available space as unknown in this case (#4657). git-svn-id: svn://localhost/ardour2/branches/3.0@12674 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-12 12:41:29 -04:00
if (_total_free_4k_blocks_uncertain) {
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
return boost::optional<samplecnt_t> ();
Handle failure of statfs to return available disc space on Samba shares mounted via GVFS, reporting available space as unknown in this case (#4657). git-svn-id: svn://localhost/ardour2/branches/3.0@12674 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-12 12:41:29 -04:00
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
float sample_bytes_on_disk = 4.0; // keep gcc happy
First stage of options rework. - Split Configuration into RCConfiguration and SessionConfiguration; the first for options which are saved to .rc files and the second for options which are saved in a session file. - Move some options from the old `master' Configuration object into SessionConfiguration; this needs more refinement. - Reflect many RCConfiguration options in an expanded Edit->Preferences dialog; my intention is to remove the corresponding menu items eventually. git-svn-id: svn://localhost/ardour2/branches/3.0@5075 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-13 20:13:27 -04:00
switch (config.get_native_file_data_format()) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
case FormatFloat:
sample_bytes_on_disk = 4.0;
break;
case FormatInt24:
sample_bytes_on_disk = 3.0;
break;
case FormatInt16:
sample_bytes_on_disk = 2.0;
break;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
default:
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* impossible, but keep some gcc versions happy */
fatal << string_compose (_("programming error: %1"),
X_("illegal native file data format"))
<< endmsg;
add abort() to non-reached code This cleans up a lot of false-positives in static analysis and also helps compilers to optimize code paths in general. (tagging the fatal stingstream operator as ‘noreturn’ is far less trivial)
2014-11-14 04:47:43 -05:00
abort(); /*NOTREACHED*/
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
double scale = 4096.0 / sample_bytes_on_disk;
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
if (_total_free_4k_blocks * scale > (double) max_samplecnt) {
return max_samplecnt;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
return (samplecnt_t) floor (_total_free_4k_blocks * scale);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
void
change PropertyChange from a bitfield into a real object, with all the many widespread changes that causes git-svn-id: svn://localhost/ardour2/branches/3.0@6701 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-19 13:09:08 -05:00
Session::tempo_map_changed (const PropertyChange&)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
clear_clicks ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Remove global Session::playlists variable, use getter method (1/2) Global variables that can written by anyone are to be avoided. This also simplifies exposing SessionPlaylists as Lua bindings.
2019-03-19 00:14:00 -04:00
_playlists->update_after_tempo_map_change ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Allow markers to be glued to bar/beat time. Fixes #1815. git-svn-id: svn://localhost/ardour2/branches/3.0@7573 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-09 12:40:31 -04:00
_locations->apply (*this, &Session::update_locations_after_tempo_map_change);
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
set_dirty ();
}
Allow markers to be glued to bar/beat time. Fixes #1815. git-svn-id: svn://localhost/ardour2/branches/3.0@7573 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-09 12:40:31 -04:00
void
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
Session::update_locations_after_tempo_map_change (const Locations::LocationList& loc)
Allow markers to be glued to bar/beat time. Fixes #1815. git-svn-id: svn://localhost/ardour2/branches/3.0@7573 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-09 12:40:31 -04:00
{
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour. Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and object-level change-related signals.
2014-10-24 12:18:40 -04:00
for (Locations::LocationList::const_iterator i = loc.begin(); i != loc.end(); ++i) {
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
(*i)->recompute_samples_from_beat ();
Allow markers to be glued to bar/beat time. Fixes #1815. git-svn-id: svn://localhost/ardour2/branches/3.0@7573 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-09 12:40:31 -04:00
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/** Ensures that all buffers (scratch, send, silent, etc) are allocated for
* the given count with the current block size.
*/
void
Session::ensure_buffers (ChanCount howmany)
{
apply gain automation on bounce - fixes 5887
2014-05-26 00:58:44 -04:00
BufferManager::ensure_buffers (howmany, bounce_processing() ? bounce_chunk_size : 0);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
Preliminary MIDI plugin support. git-svn-id: svn://localhost/ardour2/branches/3.0@5036 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-04 11:50:51 -04:00
void
Session::ensure_buffer_set(BufferSet& buffers, const ChanCount& count)
{
for (DataType::iterator t = DataType::begin(); t != DataType::end(); ++t) {
buffers.ensure_buffers(*t, count.get(*t), _engine.raw_buffer_size(*t));
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
uint32_t
Session::next_insert_id ()
{
/* this doesn't really loop forever. just think about it */
while (true) {
fix send-id "0" when loading session Send::set_state() unsets the current send-id before setting the new one, but by default _bitset defaults to zero. This may lead to unset an existing send "0"
2016-04-22 17:36:41 -04:00
for (boost::dynamic_bitset<uint32_t>::size_type n = 1; n < insert_bitset.size(); ++n) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (!insert_bitset[n]) {
insert_bitset[n] = true;
return n;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* none available, so resize and try again */
insert_bitset.resize (insert_bitset.size() + 16, false);
}
}
uint32_t
Session::next_send_id ()
{
/* this doesn't really loop forever. just think about it */
while (true) {
fix send-id "0" when loading session Send::set_state() unsets the current send-id before setting the new one, but by default _bitset defaults to zero. This may lead to unset an existing send "0"
2016-04-22 17:36:41 -04:00
for (boost::dynamic_bitset<uint32_t>::size_type n = 1; n < send_bitset.size(); ++n) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (!send_bitset[n]) {
send_bitset[n] = true;
return n;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* none available, so resize and try again */
send_bitset.resize (send_bitset.size() + 16, false);
}
}
fix two major assert failures arising from the optional monitor section commit; separate numbering of aux sends, sends and listens to fix #3671 (still testing, but the assert failures are critical) git-svn-id: svn://localhost/ardour2/branches/3.0@11263 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-18 12:51:57 -05:00
uint32_t
Session::next_aux_send_id ()
{
/* this doesn't really loop forever. just think about it */
while (true) {
fix send-id "0" when loading session Send::set_state() unsets the current send-id before setting the new one, but by default _bitset defaults to zero. This may lead to unset an existing send "0"
2016-04-22 17:36:41 -04:00
for (boost::dynamic_bitset<uint32_t>::size_type n = 1; n < aux_send_bitset.size(); ++n) {
fix two major assert failures arising from the optional monitor section commit; separate numbering of aux sends, sends and listens to fix #3671 (still testing, but the assert failures are critical) git-svn-id: svn://localhost/ardour2/branches/3.0@11263 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-18 12:51:57 -05:00
if (!aux_send_bitset[n]) {
aux_send_bitset[n] = true;
return n;
}
}
/* none available, so resize and try again */
aux_send_bitset.resize (aux_send_bitset.size() + 16, false);
}
}
Returns (i.e. sidechains). And lo, upon the revision of our hoarde 5061, was the last Big Feature committed to Three Poino, who, now more than ever, lurks imposingly on the sidelines, heir to the throne, and eventual ruler of the realm. His eventual succession all but guaranteed, only time and the number of heads that must roll remain mysteries. git-svn-id: svn://localhost/ardour2/branches/3.0@5061 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-07 13:31:18 -04:00
uint32_t
Session::next_return_id ()
{
/* this doesn't really loop forever. just think about it */
while (true) {
fix send-id "0" when loading session Send::set_state() unsets the current send-id before setting the new one, but by default _bitset defaults to zero. This may lead to unset an existing send "0"
2016-04-22 17:36:41 -04:00
for (boost::dynamic_bitset<uint32_t>::size_type n = 1; n < return_bitset.size(); ++n) {
Returns (i.e. sidechains). And lo, upon the revision of our hoarde 5061, was the last Big Feature committed to Three Poino, who, now more than ever, lurks imposingly on the sidelines, heir to the throne, and eventual ruler of the realm. His eventual succession all but guaranteed, only time and the number of heads that must roll remain mysteries. git-svn-id: svn://localhost/ardour2/branches/3.0@5061 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-07 13:31:18 -04:00
if (!return_bitset[n]) {
return_bitset[n] = true;
return n;
}
}
/* none available, so resize and try again */
return_bitset.resize (return_bitset.size() + 16, false);
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::mark_send_id (uint32_t id)
{
if (id >= send_bitset.size()) {
send_bitset.resize (id+16, false);
}
if (send_bitset[id]) {
warning << string_compose (_("send ID %1 appears to be in use already"), id) << endmsg;
}
send_bitset[id] = true;
}
fix two major assert failures arising from the optional monitor section commit; separate numbering of aux sends, sends and listens to fix #3671 (still testing, but the assert failures are critical) git-svn-id: svn://localhost/ardour2/branches/3.0@11263 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-18 12:51:57 -05:00
void
Session::mark_aux_send_id (uint32_t id)
{
if (id >= aux_send_bitset.size()) {
aux_send_bitset.resize (id+16, false);
}
if (aux_send_bitset[id]) {
warning << string_compose (_("aux send ID %1 appears to be in use already"), id) << endmsg;
}
aux_send_bitset[id] = true;
}
Returns (i.e. sidechains). And lo, upon the revision of our hoarde 5061, was the last Big Feature committed to Three Poino, who, now more than ever, lurks imposingly on the sidelines, heir to the throne, and eventual ruler of the realm. His eventual succession all but guaranteed, only time and the number of heads that must roll remain mysteries. git-svn-id: svn://localhost/ardour2/branches/3.0@5061 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-05-07 13:31:18 -04:00
void
Session::mark_return_id (uint32_t id)
{
if (id >= return_bitset.size()) {
return_bitset.resize (id+16, false);
}
if (return_bitset[id]) {
warning << string_compose (_("return ID %1 appears to be in use already"), id) << endmsg;
}
return_bitset[id] = true;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::mark_insert_id (uint32_t id)
{
if (id >= insert_bitset.size()) {
insert_bitset.resize (id+16, false);
}
if (insert_bitset[id]) {
warning << string_compose (_("insert ID %1 appears to be in use already"), id) << endmsg;
}
insert_bitset[id] = true;
}
revisit setting up processors during route construction; remove several more XML-based constructors; don't put () parens around inactive plugin names (we have a checkbox); improve management of send, return and insert bitslot IDs; clean up Diskstream construction a bit more git-svn-id: svn://localhost/ardour2/branches/3.0@6819 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-31 21:24:13 -04:00
void
Session::unmark_send_id (uint32_t id)
{
if (id < send_bitset.size()) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
send_bitset[id] = false;
}
revisit setting up processors during route construction; remove several more XML-based constructors; don't put () parens around inactive plugin names (we have a checkbox); improve management of send, return and insert bitslot IDs; clean up Diskstream construction a bit more git-svn-id: svn://localhost/ardour2/branches/3.0@6819 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-31 21:24:13 -04:00
}
fix two major assert failures arising from the optional monitor section commit; separate numbering of aux sends, sends and listens to fix #3671 (still testing, but the assert failures are critical) git-svn-id: svn://localhost/ardour2/branches/3.0@11263 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-18 12:51:57 -05:00
void
Session::unmark_aux_send_id (uint32_t id)
{
if (id < aux_send_bitset.size()) {
aux_send_bitset[id] = false;
}
}
revisit setting up processors during route construction; remove several more XML-based constructors; don't put () parens around inactive plugin names (we have a checkbox); improve management of send, return and insert bitslot IDs; clean up Diskstream construction a bit more git-svn-id: svn://localhost/ardour2/branches/3.0@6819 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-31 21:24:13 -04:00
void
Session::unmark_return_id (uint32_t id)
{
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (deletion_in_progress ()) {
return;
}
revisit setting up processors during route construction; remove several more XML-based constructors; don't put () parens around inactive plugin names (we have a checkbox); improve management of send, return and insert bitslot IDs; clean up Diskstream construction a bit more git-svn-id: svn://localhost/ardour2/branches/3.0@6819 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-31 21:24:13 -04:00
if (id < return_bitset.size()) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
return_bitset[id] = false;
}
revisit setting up processors during route construction; remove several more XML-based constructors; don't put () parens around inactive plugin names (we have a checkbox); improve management of send, return and insert bitslot IDs; clean up Diskstream construction a bit more git-svn-id: svn://localhost/ardour2/branches/3.0@6819 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-31 21:24:13 -04:00
}
void
Session::unmark_insert_id (uint32_t id)
{
if (id < insert_bitset.size()) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
insert_bitset[id] = false;
}
revisit setting up processors during route construction; remove several more XML-based constructors; don't put () parens around inactive plugin names (we have a checkbox); improve management of send, return and insert bitslot IDs; clean up Diskstream construction a bit more git-svn-id: svn://localhost/ardour2/branches/3.0@6819 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-31 21:24:13 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::reset_native_file_format ()
{
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
boost::shared_ptr<RouteList> rl = routes.reader ();
after changing session SR, reset all write sources to use new rate
2015-05-12 10:16:26 -04:00
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
for (RouteList::iterator i = rl->begin(); i != rl->end(); ++i) {
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
if (tr) {
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
/* don't save state as we do this, there's no point */
_state_of_the_state = StateOfTheState (_state_of_the_state | InCleanup);
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
tr->reset_write_sources (false);
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
_state_of_the_state = StateOfTheState (_state_of_the_state & ~InCleanup);
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
bool
Session::route_name_unique (string n) const
{
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (RouteList::const_iterator i = r->begin(); i != r->end(); ++i) {
if ((*i)->name() == n) {
return false;
}
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
return true;
}
Disallow click and auditioner as track / buss names to avoid clashing with internal IOs. git-svn-id: svn://localhost/ardour2/branches/3.0@5009 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 12:15:33 -04:00
bool
Session::route_name_internal (string n) const
{
if (auditioner && auditioner->name() == n) {
return true;
}
if (_click_io && _click_io->name() == n) {
return true;
}
return false;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
int
(1) remove most uses of MementoCommand for Playlist and Region (2) move frozen state from Region into Stateful, renamed "suspend property changes" (3) successive changes to a Property (scalar) after clear_history() do not keep resetting the old value (fixes region trim) git-svn-id: svn://localhost/ardour2/branches/3.0@6720 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-02 13:05:26 -05:00
Session::freeze_all (InterThreadInfo& itt)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
first half of the mega-by-hand-commit-by-commit merge from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4980 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-15 14:04:23 -04:00
boost::shared_ptr<Track> t;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
first half of the mega-by-hand-commit-by-commit merge from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4980 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-15 14:04:23 -04:00
if ((t = boost::dynamic_pointer_cast<Track>(*i)) != 0) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* XXX this is wrong because itt.progress will keep returning to zero at the start
of every track.
*/
(1) remove most uses of MementoCommand for Playlist and Region (2) move frozen state from Region into Stateful, renamed "suspend property changes" (3) successive changes to a Property (scalar) after clear_history() do not keep resetting the old value (fixes region trim) git-svn-id: svn://localhost/ardour2/branches/3.0@6720 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-03-02 13:05:26 -05:00
t->freeze_me (itt);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
return 0;
}
merge from 2.0-ongoing @ 3581 git-svn-id: svn://localhost/ardour2/branches/3.0@3711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-09-10 11:03:30 -04:00
boost::shared_ptr<Region>
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::write_one_track (Track& track, samplepos_t start, samplepos_t end,
Strip trailing whitespace and fix other whitespace errors (e.g. space/tab mixing). Whitespace changes only. Vimmers, try let c_space_errors = 1 in your .vimrc to highlight this kind of stuff in red. I don't know the emacs equivalent... git-svn-id: svn://localhost/ardour2/branches/3.0@5773 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-10-14 12:10:01 -04:00
bool /*overwrite*/, vector<boost::shared_ptr<Source> >& srcs,
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
InterThreadInfo& itt,
radically rethink export/bounce/freeze code design. probably not 100% done by freeze+unfreeze now work and behave sensibly w.r.t. processors that do routing git-svn-id: svn://localhost/ardour2/branches/3.0@11701 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-03-15 17:40:17 -04:00
boost::shared_ptr<Processor> endpoint, bool include_endpoint,
re-work bounce/freeze. Freezer stops at first active delivery. amend to 8f52bf7d9f
2014-06-03 15:08:45 -04:00
bool for_export, bool for_freeze)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
merge from 2.0-ongoing @ 3581 git-svn-id: svn://localhost/ardour2/branches/3.0@3711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-09-10 11:03:30 -04:00
boost::shared_ptr<Region> result;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
boost::shared_ptr<Playlist> playlist;
MIDI bounce.
2014-12-17 21:48:09 -05:00
boost::shared_ptr<Source> source;
Stop bounce / freeze on tracks that have more outputs than inputs and so cannot record all the outputs in their diskstreams. Fix buffer shortage when bouncing tracks whose processing chains temporarily need more buffers than there are inputs. Fixes #3573. git-svn-id: svn://localhost/ardour2/branches/3.0@8239 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-10 17:28:29 -05:00
ChanCount diskstream_channels (track.n_channels());
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
samplepos_t position;
samplecnt_t this_chunk;
samplepos_t to_do;
samplepos_t latency_skip;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
BufferSet buffers;
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
samplepos_t len = end - start;
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
bool need_block_size_reset = false;
Stop bounce / freeze on tracks that have more outputs than inputs and so cannot record all the outputs in their diskstreams. Fix buffer shortage when bouncing tracks whose processing chains temporarily need more buffers than there are inputs. Fixes #3573. git-svn-id: svn://localhost/ardour2/branches/3.0@8239 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-10 17:28:29 -05:00
ChanCount const max_proc = track.max_processor_streams ();
when exporting regions, use ::legalize_for_path() on the playlist name so that filenames are legal. Also, use Glib::build_filename() rather than hardcoding / as the separator. How did this escape the search for this error?
2014-06-22 21:58:07 -04:00
string legal_playlist_name;
string possible_path;
merge from 2.0-ongoing @ 3581 git-svn-id: svn://localhost/ardour2/branches/3.0@3711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-09-10 11:03:30 -04:00
NO-OP/Optimize Bounce: Lookup data-type only once This is to prepare for midi-to-audio bounce: allow to change the data-type to be used as output.
2019-04-07 13:34:37 -04:00
DataType data_type = track.data_type();
merge from 2.0-ongoing @ 3581 git-svn-id: svn://localhost/ardour2/branches/3.0@3711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-09-10 11:03:30 -04:00
if (end <= start) {
error << string_compose (_("Cannot write a range where end <= start (e.g. %1 <= %2)"),
end, start) << endmsg;
return result;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
re-work bounce/freeze. Freezer stops at first active delivery. amend to 8f52bf7d9f
2014-06-03 15:08:45 -04:00
diskstream_channels = track.bounce_get_output_streams (diskstream_channels, endpoint,
include_endpoint, for_export, for_freeze);
bounce to target number of channels
2014-05-25 12:10:01 -04:00
Enable midi-to-audio bounce when processing w/synth
2019-04-07 13:51:08 -04:00
if (data_type == DataType::MIDI && endpoint && !for_export && !for_freeze && diskstream_channels.n(DataType::AUDIO) > 0) {
data_type = DataType::AUDIO;
}
NO-OP/Optimize Bounce: Lookup data-type only once This is to prepare for midi-to-audio bounce: allow to change the data-type to be used as output.
2019-04-07 13:34:37 -04:00
if (diskstream_channels.n(data_type) < 1) {
MIDI bounce.
2014-12-17 21:48:09 -05:00
error << _("Cannot write a range with no data.") << endmsg;
bounce to target number of channels
2014-05-25 12:10:01 -04:00
return result;
}
part 2 of 3 of the 2.8 -> 3.0 merge git-svn-id: svn://localhost/ardour2/branches/3.0@4987 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-16 12:02:25 -04:00
// block all process callback handling
block_processing ();
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
fix bounce with processing (thread buffers) fixes #5920
2014-05-25 02:23:36 -04:00
{
// synchronize with AudioEngine::process_callback()
// make sure processing is not currently running
// and processing_blocked() is honored before
// acquiring thread buffers
Glib::Threads::Mutex::Lock lm (_engine.process_lock());
}
process effect-automation when bouncing w/processing
2014-05-26 00:17:49 -04:00
_bounce_processing_active = true;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* call tree *MUST* hold route_lock */
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
if ((playlist = track.playlist()) == 0) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
goto out;
}
prefer lowercase and no spaces, since this will appear in the filename.
2019-07-31 12:14:38 -04:00
legal_playlist_name = "(bounce)" + legalize_for_path (playlist->name());
when exporting regions, use ::legalize_for_path() on the playlist name so that filenames are legal. Also, use Glib::build_filename() rather than hardcoding / as the separator. How did this escape the search for this error?
2014-06-22 21:58:07 -04:00
NO-OP/Optimize Bounce: Lookup data-type only once This is to prepare for midi-to-audio bounce: allow to change the data-type to be used as output.
2019-04-07 13:34:37 -04:00
for (uint32_t chan_n = 0; chan_n < diskstream_channels.n(data_type); ++chan_n) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
string base_name = string_compose ("%1-%2-bounce", playlist->name(), chan_n);
NO-OP/Optimize Bounce: Lookup data-type only once This is to prepare for midi-to-audio bounce: allow to change the data-type to be used as output.
2019-04-07 13:34:37 -04:00
string path = ((data_type == DataType::AUDIO)
MIDI bounce.
2014-12-17 21:48:09 -05:00
? new_audio_source_path (legal_playlist_name, diskstream_channels.n_audio(), chan_n, false, true)
: new_midi_source_path (legal_playlist_name));
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
substantive changes to the logic and safety for naming of (audio/MIDI) sources, especially when created via import
2014-06-02 11:20:37 -04:00
if (path.empty()) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
goto out;
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
try {
NO-OP/Optimize Bounce: Lookup data-type only once This is to prepare for midi-to-audio bounce: allow to change the data-type to be used as output.
2019-04-07 13:34:37 -04:00
source = SourceFactory::createWritable (data_type, *this, path, false, sample_rate());
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
catch (failed_constructor& err) {
MIDI bounce.
2014-12-17 21:48:09 -05:00
error << string_compose (_("cannot create new file \"%1\" for %2"), path, track.name()) << endmsg;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
goto out;
}
MIDI bounce.
2014-12-17 21:48:09 -05:00
srcs.push_back (source);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
put back revised version of Route::has_external_redirects() and use it to give a warning when attempting to freeze a track with sends/inserts/returns git-svn-id: svn://localhost/ardour2/branches/3.0@11707 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-03-16 12:51:54 -04:00
/* tell redirects that care that we are about to use a much larger
* blocksize. this will flush all plugins too, so that they are ready
* to be used for this process.
*/
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
need_block_size_reset = true;
apply gain automation on bounce - fixes 5887
2014-05-26 00:58:44 -04:00
track.set_block_size (bounce_chunk_size);
_engine.main_thread()->get_buffers ();
forward port 2.X changes up to and including rev 6714 git-svn-id: svn://localhost/ardour2/branches/3.0@7635 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-16 15:58:34 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
position = start;
to_do = len;
re-work bounce/freeze. Freezer stops at first active delivery. amend to 8f52bf7d9f
2014-06-03 15:08:45 -04:00
latency_skip = track.bounce_get_latency (endpoint, include_endpoint, for_export, for_freeze);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* create a set of reasonably-sized buffers */
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
for (DataType::iterator t = DataType::begin(); t != DataType::end(); ++t) {
apply gain automation on bounce - fixes 5887
2014-05-26 00:58:44 -04:00
buffers.ensure_buffers(*t, max_proc.get(*t), bounce_chunk_size);
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
}
Stop bounce / freeze on tracks that have more outputs than inputs and so cannot record all the outputs in their diskstreams. Fix buffer shortage when bouncing tracks whose processing chains temporarily need more buffers than there are inputs. Fixes #3573. git-svn-id: svn://localhost/ardour2/branches/3.0@8239 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-10 17:28:29 -05:00
buffers.set_count (max_proc);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Stop bounce / freeze on tracks that have more outputs than inputs and so cannot record all the outputs in their diskstreams. Fix buffer shortage when bouncing tracks whose processing chains temporarily need more buffers than there are inputs. Fixes #3573. git-svn-id: svn://localhost/ardour2/branches/3.0@8239 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-10 17:28:29 -05:00
for (vector<boost::shared_ptr<Source> >::iterator src = srcs.begin(); src != srcs.end(); ++src) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
boost::shared_ptr<AudioFileSource> afs = boost::dynamic_pointer_cast<AudioFileSource>(*src);
Fix MIDI bounce/consolidate note resolution.
2015-01-17 22:19:57 -05:00
boost::shared_ptr<MidiSource> ms;
if (afs) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
afs->prepare_for_peakfile_writes ();
Fix MIDI bounce/consolidate note resolution.
2015-01-17 22:19:57 -05:00
} else if ((ms = boost::dynamic_pointer_cast<MidiSource>(*src))) {
Source::Lock lock(ms->mutex());
ms->mark_streaming_write_started(lock);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
while (to_do && !itt.cancel) {
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
apply gain automation on bounce - fixes 5887
2014-05-26 00:58:44 -04:00
this_chunk = min (to_do, bounce_chunk_size);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
re-work bounce/freeze. Freezer stops at first active delivery. amend to 8f52bf7d9f
2014-06-03 15:08:45 -04:00
if (track.export_stuff (buffers, start, this_chunk, endpoint, include_endpoint, for_export, for_freeze)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
goto out;
}
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
start += this_chunk;
to_do -= this_chunk;
itt.progress = (float) (1.0 - ((double) to_do / len));
apply gain automation on bounce - fixes 5887
2014-05-26 00:58:44 -04:00
if (latency_skip >= bounce_chunk_size) {
latency_skip -= bounce_chunk_size;
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
continue;
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
const samplecnt_t current_chunk = this_chunk - latency_skip;
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
uint32_t n = 0;
for (vector<boost::shared_ptr<Source> >::iterator src=srcs.begin(); src != srcs.end(); ++src, ++n) {
boost::shared_ptr<AudioFileSource> afs = boost::dynamic_pointer_cast<AudioFileSource>(*src);
MIDI bounce.
2014-12-17 21:48:09 -05:00
boost::shared_ptr<MidiSource> ms;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (afs) {
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
if (afs->write (buffers.get_audio(n).data(latency_skip), current_chunk) != current_chunk) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
goto out;
}
MIDI bounce.
2014-12-17 21:48:09 -05:00
} else if ((ms = boost::dynamic_pointer_cast<MidiSource>(*src))) {
Source::Lock lock(ms->mutex());
const MidiBuffer& buf = buffers.get_midi(0);
for (MidiBuffer::const_iterator i = buf.begin(); i != buf.end(); ++i) {
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Evoral::Event<samplepos_t> ev = *i;
Prepare midi-export for midi-to-audio bouncing. * shift event time into process cycle before calling processors * reset note-trackers when exporting, don't allow any pending events into the queue
2019-04-07 13:18:28 -04:00
if (!endpoint || for_export) {
ev.set_time(ev.time() - position);
}
(Source List) Clean up the natural_position implementation (libardour part).
2018-11-15 10:33:54 -05:00
ms->append_event_samples(lock, ev, ms->natural_position());
MIDI bounce.
2014-12-17 21:48:09 -05:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
latency_skip = 0;
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
/* post-roll, pick up delayed processor output */
re-work bounce/freeze. Freezer stops at first active delivery. amend to 8f52bf7d9f
2014-06-03 15:08:45 -04:00
latency_skip = track.bounce_get_latency (endpoint, include_endpoint, for_export, for_freeze);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
while (latency_skip && !itt.cancel) {
apply gain automation on bounce - fixes 5887
2014-05-26 00:58:44 -04:00
this_chunk = min (latency_skip, bounce_chunk_size);
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
latency_skip -= this_chunk;
buffers.silence (this_chunk, 0);
re-work bounce/freeze. Freezer stops at first active delivery. amend to 8f52bf7d9f
2014-06-03 15:08:45 -04:00
track.bounce_process (buffers, start, this_chunk, endpoint, include_endpoint, for_export, for_freeze);
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
uint32_t n = 0;
for (vector<boost::shared_ptr<Source> >::iterator src=srcs.begin(); src != srcs.end(); ++src, ++n) {
boost::shared_ptr<AudioFileSource> afs = boost::dynamic_pointer_cast<AudioFileSource>(*src);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
compensate for processor latency during bounce
2014-05-25 13:43:37 -04:00
if (afs) {
if (afs->write (buffers.get_audio(n).data(), this_chunk) != this_chunk) {
goto out;
}
}
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
if (!itt.cancel) {
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
time_t now;
struct tm* xnow;
time (&now);
xnow = localtime (&now);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (vector<boost::shared_ptr<Source> >::iterator src=srcs.begin(); src != srcs.end(); ++src) {
boost::shared_ptr<AudioFileSource> afs = boost::dynamic_pointer_cast<AudioFileSource>(*src);
MIDI bounce.
2014-12-17 21:48:09 -05:00
boost::shared_ptr<MidiSource> ms;
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (afs) {
afs->update_header (position, *xnow, now);
afs->flush_header ();
MIDI bounce.
2014-12-17 21:48:09 -05:00
} else if ((ms = boost::dynamic_pointer_cast<MidiSource>(*src))) {
Source::Lock lock(ms->mutex());
ms->mark_streaming_write_completed(lock);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
/* construct a region to represent the bounced material */
the Properties & 64bit region commit git-svn-id: svn://localhost/ardour2/branches/3.0@6695 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-18 08:59:49 -05:00
PropertyList plist;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
the Properties & 64bit region commit git-svn-id: svn://localhost/ardour2/branches/3.0@6695 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-18 08:59:49 -05:00
plist.add (Properties::start, 0);
(Source List) Catch new sources created by bounce-range-to-source-list
2019-07-30 16:43:22 -04:00
plist.add (Properties::whole_file, true);
(Source List) Clean up the natural_position implementation (libardour part).
2018-11-15 10:33:54 -05:00
plist.add (Properties::length, srcs.front()->length(srcs.front()->natural_position()));
the Properties & 64bit region commit git-svn-id: svn://localhost/ardour2/branches/3.0@6695 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-02-18 08:59:49 -05:00
plist.add (Properties::name, region_name_from_path (srcs.front()->name(), true));
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
(Source List) Catch new sources created by bounce-range-to-source-list
2019-07-30 16:43:22 -04:00
result = RegionFactory::create (srcs, plist, true);
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
out:
merge from 2.0-ongoing @ 3581 git-svn-id: svn://localhost/ardour2/branches/3.0@3711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-09-10 11:03:30 -04:00
if (!result) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (vector<boost::shared_ptr<Source> >::iterator src = srcs.begin(); src != srcs.end(); ++src) {
MIDI bounce.
2014-12-17 21:48:09 -05:00
(*src)->mark_for_remove ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
(*src)->drop_references ();
}
} else {
for (vector<boost::shared_ptr<Source> >::iterator src = srcs.begin(); src != srcs.end(); ++src) {
boost::shared_ptr<AudioFileSource> afs = boost::dynamic_pointer_cast<AudioFileSource>(*src);
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
if (afs)
afs->done_with_peakfile_writes ();
}
}
process effect-automation when bouncing w/processing
2014-05-26 00:17:49 -04:00
_bounce_processing_active = false;
forward port 2.X changes up to and including rev 6714 git-svn-id: svn://localhost/ardour2/branches/3.0@7635 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-16 15:58:34 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (need_block_size_reset) {
apply gain automation on bounce - fixes 5887
2014-05-26 00:58:44 -04:00
_engine.main_thread()->drop_buffers ();
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
track.set_block_size (get_block_size());
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
part 2 of 3 of the 2.8 -> 3.0 merge git-svn-id: svn://localhost/ardour2/branches/3.0@4987 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-16 12:02:25 -04:00
unblock_processing ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
merge from 2.0-ongoing @ 3581 git-svn-id: svn://localhost/ardour2/branches/3.0@3711 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-09-10 11:03:30 -04:00
return result;
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
substantive change: use the JACK wait API and provide "thread buffers" separately from session in preparation for parallelization. lots of debug output at present. If using JACK1, requires a very current version of JACK1 SVN (0.119.0) git-svn-id: svn://localhost/ardour2/branches/3.0@6888 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-13 16:48:33 -04:00
gain_t*
Session::gain_automation_buffer() const
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
return ProcessThread::gain_automation_buffer ();
substantive change: use the JACK wait API and provide "thread buffers" separately from session in preparation for parallelization. lots of debug output at present. If using JACK1, requires a very current version of JACK1 SVN (0.119.0) git-svn-id: svn://localhost/ardour2/branches/3.0@6888 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-13 16:48:33 -04:00
}
prepare trim automation
2015-04-25 15:24:58 -04:00
gain_t*
Session::trim_automation_buffer() const
{
return ProcessThread::trim_automation_buffer ();
}
Make send automation work (#4734). git-svn-id: svn://localhost/ardour2/branches/3.0@12650 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-11 06:42:30 -04:00
gain_t*
Session::send_gain_automation_buffer() const
{
return ProcessThread::send_gain_automation_buffer ();
}
Add a scratch buffer for automation. Useful as temporary buffer: This allows a controllable to get a master's automation-curve and combine it with its own (gain, trim, send) automation buffer.
2017-06-03 06:26:33 -04:00
gain_t*
Session::scratch_automation_buffer() const
{
return ProcessThread::scratch_automation_buffer ();
}
substantive change: use the JACK wait API and provide "thread buffers" separately from session in preparation for parallelization. lots of debug output at present. If using JACK1, requires a very current version of JACK1 SVN (0.119.0) git-svn-id: svn://localhost/ardour2/branches/3.0@6888 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-13 16:48:33 -04:00
pan_t**
Session::pan_automation_buffer() const
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
return ProcessThread::pan_automation_buffer ();
substantive change: use the JACK wait API and provide "thread buffers" separately from session in preparation for parallelization. lots of debug output at present. If using JACK1, requires a very current version of JACK1 SVN (0.119.0) git-svn-id: svn://localhost/ardour2/branches/3.0@6888 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-13 16:48:33 -04:00
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
BufferSet&
Session::get_silent_buffers (ChanCount count)
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
return ProcessThread::get_silent_buffers (count);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
BufferSet&
use zeroed scratch buffers for "silent" plugin runs Plugins rewrite the buffer data in-place and some plugins can produce output even when fed with silence. Hence, during a PluginInsert::silence() run a plugin can inject data into the "silent" buffers which causes side-effects. Kudos to Chris 'oofus' Goddard for finding this issue.
2013-07-31 18:35:24 -04:00
Session::get_scratch_buffers (ChanCount count, bool silence)
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
{
use zeroed scratch buffers for "silent" plugin runs Plugins rewrite the buffer data in-place and some plugins can produce output even when fed with silence. Hence, during a PluginInsert::silence() run a plugin can inject data into the "silent" buffers which causes side-effects. Kudos to Chris 'oofus' Goddard for finding this issue.
2013-07-31 18:35:24 -04:00
return ProcessThread::get_scratch_buffers (count, silence);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
add "no-inplace" buffers. When allowing to cross-connect plugin-ports, inplace processing can no longer be used. We need a complete set of independent input and output buffers. Since scratch and silent buffers are used by the various plugin implementations we cannot re-use them in the PluginInsert. Besides we need a complete BufferSet which can hold both: ins + outs.
2016-03-25 19:40:51 -04:00
BufferSet&
Session::get_noinplace_buffers (ChanCount count)
{
return ProcessThread::get_noinplace_buffers (count);
}
use dedicated buffers for route (and track) "scratch buffers are by definition scratch and their contents are undefined at all times" "silent buffers are by definition all-zero and should not be used for real data" But track & route were using those for actual data; plugins (which may run in the same thread and may get the same buffers) use them for scratch thereby overwriting real data. In particular get_silent_buffers() (used by LadspaPlugin::connect_and_run) clears the buffer which can holds real data: e.g. via Route::passthru_silence() -> plugin1 -> plugin2 (clears output of plugin1)
2013-07-30 10:55:33 -04:00
BufferSet&
Session::get_route_buffers (ChanCount count, bool silence)
{
return ProcessThread::get_route_buffers (count, silence);
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
BufferSet&
Session::get_mix_buffers (ChanCount count)
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
return ProcessThread::get_mix_buffers (count);
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
uint32_t
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Session::ntracks () const
{
uint32_t n = 0;
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (RouteList::const_iterator i = r->begin(); i != r->end(); ++i) {
first half of the mega-by-hand-commit-by-commit merge from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4980 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-15 14:04:23 -04:00
if (boost::dynamic_pointer_cast<Track> (*i)) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
++n;
}
}
return n;
}
* first implementation of MIDI Clock Slave support git-svn-id: svn://localhost/ardour2/branches/3.0@3652 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-08-04 18:37:24 -04:00
uint32_t
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
Session::nbusses () const
{
uint32_t n = 0;
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
for (RouteList::const_iterator i = r->begin(); i != r->end(); ++i) {
first half of the mega-by-hand-commit-by-commit merge from 2.X git-svn-id: svn://localhost/ardour2/branches/3.0@4980 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-15 14:04:23 -04:00
if (boost::dynamic_pointer_cast<Track>(*i) == 0) {
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
++n;
}
}
return n;
}
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
uint32_t
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
Session::nstripables (bool with_monitor) const
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
{
uint32_t rv = routes.reader()->size ();
rv += _vca_manager->vcas ().size ();
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
if (with_monitor) {
Fix stripable order for new strips & master-order
2017-06-16 22:23:55 -04:00
return rv;
}
if (_monitor_out) {
assert (rv > 0);
--rv;
}
return rv;
}
Add method to graphviz plot the process-graph
2019-08-07 11:27:01 -04:00
bool
Session::plot_process_graph (std::string const& file_name) const {
return _process_graph ? _process_graph->plot (file_name) : false;
}
rollback to 3428, before the mysterious removal of libs/* at 3431/3432 git-svn-id: svn://localhost/ardour2/branches/3.0@3435 d708f5d6-7413-0410-9779-e7cbd77b26cf
2008-06-02 17:41:35 -04:00
void
Session::add_automation_list(AutomationList *al)
{
automation_lists[al->id()] = al;
}
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
/** @return true if there is at least one record-enabled track, otherwise false */
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
bool
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
Session::have_rec_enabled_track () const
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
{
tweak for const correctness on apple git-svn-id: svn://localhost/ardour2/branches/3.0@13956 d708f5d6-7413-0410-9779-e7cbd77b26cf
2013-01-21 14:42:34 -05:00
return g_atomic_int_get (const_cast<gint*>(&_have_rec_enabled_track)) == 1;
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
}
provide Session::have_rec_disabled_track()
2015-05-08 13:46:42 -04:00
bool
Session::have_rec_disabled_track () const
{
NO-OP: <tab> after <space> fixes in libs
2019-04-13 13:19:29 -04:00
return g_atomic_int_get (const_cast<gint*>(&_have_rec_disabled_track)) == 1;
provide Session::have_rec_disabled_track()
2015-05-08 13:46:42 -04:00
}
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
/** Update the state of our rec-enabled tracks flag */
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
void
[Summary] Changed function name update_route_active_state() on update_route_record_state() [Reviewed] GZharun Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc libs/ardour/session_state.cc
2014-09-19 09:45:01 -04:00
Session::update_route_record_state ()
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
{
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
boost::shared_ptr<RouteList> rl = routes.reader ();
RouteList::iterator i = rl->begin();
while (i != rl->end ()) {
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
if (tr && tr->rec_enable_control()->get_value()) {
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
break;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
++i;
}
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
int const old = g_atomic_int_get (&_have_rec_enabled_track);
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
g_atomic_int_set (&_have_rec_enabled_track, i != rl->end () ? 1 : 0);
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
if (g_atomic_int_get (&_have_rec_enabled_track) != old) {
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
RecordStateChanged (); /* EMIT SIGNAL */
}
provide Session::have_rec_disabled_track()
2015-05-08 13:46:42 -04:00
another signal related to record state added from Tracks. This doesn't appear to have obviously distinct semantics from RecordStateChanged but Tracks is using it and I can't verify instantly that it could be replaced by RecordStateChanged.
2015-05-08 16:26:52 -04:00
for (i = rl->begin(); i != rl->end (); ++i) {
provide Session::have_rec_disabled_track()
2015-05-08 13:46:42 -04:00
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
universal change in the design of the way Route/Track controls are designed and used. The controls now own their own state, rather than proxy for state in their owners. Massive changes all over the code to accomodate this. Many things are not finished. Consider this a backup safety commit
2016-04-08 16:49:47 -04:00
if (tr && !tr->rec_enable_control()->get_value()) {
provide Session::have_rec_disabled_track()
2015-05-08 13:46:42 -04:00
break;
}
}
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
another signal related to record state added from Tracks. This doesn't appear to have obviously distinct semantics from RecordStateChanged but Tracks is using it and I can't verify instantly that it could be replaced by RecordStateChanged.
2015-05-08 16:26:52 -04:00
g_atomic_int_set (&_have_rec_disabled_track, i != rl->end () ? 1 : 0);
bool record_arm_state_changed = (old != g_atomic_int_get (&_have_rec_enabled_track) );
globally remove all trailing whitespace from ardour code base. Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
another signal related to record state added from Tracks. This doesn't appear to have obviously distinct semantics from RecordStateChanged but Tracks is using it and I can't verify instantly that it could be replaced by RecordStateChanged.
2015-05-08 16:26:52 -04:00
if (record_status() == Recording && record_arm_state_changed ) {
RecordArmStateChanged ();
}
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
Only make record button solid red (and big clock red) when things are actually being recorded (ie when record is in progress and one or more tracks are armed). As per mantis #2604. git-svn-id: svn://localhost/ardour2/branches/3.0@5012 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-04-29 13:30:35 -04:00
}
remove SoloBus solo model, add PFL and AFL. basic testing indicates it works, but probably some oddities. git-svn-id: svn://localhost/ardour2/branches/3.0@5210 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-17 15:30:54 -04:00
void
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
Session::listen_position_changed ()
remove SoloBus solo model, add PFL and AFL. basic testing indicates it works, but probably some oddities. git-svn-id: svn://localhost/ardour2/branches/3.0@5210 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-17 15:30:54 -04:00
{
NO-OP: prefer to use API to check session-state
2019-03-17 12:04:45 -04:00
if (loading ()) {
Tweak session load Listen positions are already correctly set when the route state is restored. The Route directly uses the Config. This skips duplicate calls to re-configure all processors on all Routes during RCConfiguration::map_parameters() on session-load, twice (pfl-position, listen-position)
2019-03-05 19:01:03 -05:00
/* skip duing session restore (already taken care of) */
return;
}
RAII to postpone processor changes/graph recalculation
2016-11-25 12:07:43 -05:00
ProcessorChangeBlocker pcb (this);
boost::shared_ptr<RouteList> r = routes.reader ();
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
(*i)->listen_position_changed ();
remove SoloBus solo model, add PFL and AFL. basic testing indicates it works, but probably some oddities. git-svn-id: svn://localhost/ardour2/branches/3.0@5210 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-17 15:30:54 -04:00
}
}
Optional tabs down the LHS of the editor window to indicate edit group membership. git-svn-id: svn://localhost/ardour2/branches/3.0@5220 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-20 09:41:55 -04:00
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
void
Session::solo_control_mode_changed ()
{
fix logic for Session::solo_control_mode_changed()
2016-04-11 09:47:13 -04:00
if (soloing() || listening()) {
make the session do the right thing when clearing solo state during non-loading conditions
2016-11-25 06:50:39 -05:00
if (loading()) {
/* We can't use ::clear_all_solo_state() here because during
session loading at program startup, that will queue a call
to rt_clear_all_solo_state() that will not execute until
AFTER solo states have been established (thus throwing away
the session's saved solo state). So just explicitly turn
them all off.
*/
set_controls (route_list_to_control_list (get_routes(), &Stripable::solo_control), 0.0, Controllable::NoGroup);
} else {
clear_all_solo_state (get_routes());
}
some tweaks for solo logic to get things working as they were before (correctly)
2016-04-10 18:20:11 -04:00
}
separate solo & listen. some minor fixes and additional related fixes still to come git-svn-id: svn://localhost/ardour2/branches/3.0@5298 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-07-01 09:36:50 -04:00
}
Update route colours when they are added to / removed from groups which share colour (more of #4064). git-svn-id: svn://localhost/ardour2/branches/3.0@10060 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-09-07 07:56:23 -04:00
/** Called when a property of one of our route groups changes */
Optional tabs down the LHS of the editor window to indicate edit group membership. git-svn-id: svn://localhost/ardour2/branches/3.0@5220 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-20 09:41:55 -04:00
void
Revert "mackie: still respond to route groups visibility changes" This reverts commit a9ec547457bfa65655ee946063426f1ba85b6f91.
2016-01-19 14:16:49 -05:00
Session::route_group_property_changed (RouteGroup* rg)
Mix group tabs in the mixer. git-svn-id: svn://localhost/ardour2/branches/3.0@5228 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-20 13:15:33 -04:00
{
Revert "mackie: still respond to route groups visibility changes" This reverts commit a9ec547457bfa65655ee946063426f1ba85b6f91.
2016-01-19 14:16:49 -05:00
RouteGroupPropertyChanged (rg); /* EMIT SIGNAL */
Update route colours when they are added to / removed from groups which share colour (more of #4064). git-svn-id: svn://localhost/ardour2/branches/3.0@10060 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-09-07 07:56:23 -04:00
}
/** Called when a route is added to one of our route groups */
void
Session::route_added_to_route_group (RouteGroup* rg, boost::weak_ptr<Route> r)
{
RouteAddedToRouteGroup (rg, r);
}
/** Called when a route is removed from one of our route groups */
void
Session::route_removed_from_route_group (RouteGroup* rg, boost::weak_ptr<Route> r)
{
[Summary] Changed function name update_route_active_state() on update_route_record_state() [Reviewed] GZharun Conflicts: libs/ardour/ardour/session.h libs/ardour/session.cc libs/ardour/session_state.cc
2014-09-19 09:45:01 -04:00
update_route_record_state ();
RouteRemovedFromRouteGroup (rg, r); /* EMIT SIGNAL */
Remove a group when the last track in the group is removed
2016-11-10 17:57:47 -05:00
if (!rg->has_control_master () && !rg->has_subgroup () && rg->empty()) {
remove_route_group (*rg);
}
Mix group tabs in the mixer. git-svn-id: svn://localhost/ardour2/branches/3.0@5228 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-06-20 13:15:33 -04:00
}
90% done with external sync design changes (GUI now has toggle switch for ext/int sync; source chosen separately, currently in ardour prefs dialog ; fix libmidi++ and compilation scripts to correctly build JACK+ALSA support; minor contiuing tweaks on transport stuff git-svn-id: svn://localhost/ardour2/branches/3.0@6048 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-09 15:05:18 -05:00
PT import: Minor changes to session
2018-10-10 05:17:57 -04:00
boost::shared_ptr<AudioTrack>
Session::get_nth_audio_track (int nth) const
{
boost::shared_ptr<RouteList> rl = routes.reader ();
rl->sort (Stripable::Sorter ());
for (RouteList::const_iterator r = rl->begin(); r != rl->end(); ++r) {
if (!boost::dynamic_pointer_cast<AudioTrack> (*r)) {
continue;
}
if (--nth > 0) {
continue;
}
return boost::dynamic_pointer_cast<AudioTrack> (*r);
}
return boost::shared_ptr<AudioTrack> ();
}
add Session::get_tracks()
2014-07-09 11:29:26 -04:00
boost::shared_ptr<RouteList>
Session::get_tracks () const
{
boost::shared_ptr<RouteList> rl = routes.reader ();
boost::shared_ptr<RouteList> tl (new RouteList);
for (RouteList::const_iterator r = rl->begin(); r != rl->end(); ++r) {
if (boost::dynamic_pointer_cast<Track> (*r)) {
The Auditioner is not in the Session's RouteList Remove various checks, add assert() for now (perhaps some old sessions?) This fixes an off-by one issue when adding tracks (presentation info order in add_routes_inner)
2017-08-15 17:24:32 -04:00
assert (!(*r)->is_auditioner()); // XXX remove me
tl->push_back (*r);
add Session::get_tracks()
2014-07-09 11:29:26 -04:00
}
}
return tl;
}
Implement 2532: option to show tracks with regions under the playhead. git-svn-id: svn://localhost/ardour2/branches/3.0@6207 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-29 07:47:59 -05:00
boost::shared_ptr<RouteList>
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::get_routes_with_regions_at (samplepos_t const p) const
Implement 2532: option to show tracks with regions under the playhead. git-svn-id: svn://localhost/ardour2/branches/3.0@6207 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-29 07:47:59 -05:00
{
remove using boost::shared_ptr stuff from session.cc git-svn-id: svn://localhost/ardour2/branches/3.0@8127 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-29 17:07:42 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
boost::shared_ptr<RouteList> rl (new RouteList);
Implement 2532: option to show tracks with regions under the playhead. git-svn-id: svn://localhost/ardour2/branches/3.0@6207 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-29 07:47:59 -05:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
boost::shared_ptr<Track> tr = boost::dynamic_pointer_cast<Track> (*i);
if (!tr) {
continue;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Move Diskstream ownership to Track, so that Session no longer holds lists of Diskstreams. Breaks 3.0 file format again. git-svn-id: svn://localhost/ardour2/branches/3.0@6945 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-04-21 16:42:22 -04:00
boost::shared_ptr<Playlist> pl = tr->playlist ();
Implement 2532: option to show tracks with regions under the playhead. git-svn-id: svn://localhost/ardour2/branches/3.0@6207 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-29 07:47:59 -05:00
if (!pl) {
continue;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Implement 2532: option to show tracks with regions under the playhead. git-svn-id: svn://localhost/ardour2/branches/3.0@6207 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-11-29 07:47:59 -05:00
if (pl->has_region_at (p)) {
rl->push_back (*i);
}
}
return rl;
}
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
void
Session::goto_end ()
{
if (_session_range_location) {
request_locate (_session_range_location->end(), false);
} else {
request_locate (0, false);
}
}
void
change Session::goto_start() to take an argument to optionally roll after locate
2016-06-17 15:41:28 -04:00
Session::goto_start (bool and_roll)
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
{
if (_session_range_location) {
change Session::goto_start() to take an argument to optionally roll after locate
2016-06-17 15:41:28 -04:00
request_locate (_session_range_location->start(), and_roll);
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
} else {
change Session::goto_start() to take an argument to optionally roll after locate
2016-06-17 15:41:28 -04:00
request_locate (0, and_roll);
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
}
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
samplepos_t
Session::current_start_sample () const
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
{
return _session_range_location ? _session_range_location->start() : 0;
}
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
samplepos_t
Session::current_end_sample () const
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
{
return _session_range_location ? _session_range_location->end() : 0;
}
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::set_session_range_location (samplepos_t start, samplepos_t end)
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
{
allow all types of range location (loop, start, end etc.) to be glued to bars and beats.
2016-12-28 09:39:31 -05:00
_session_range_location = new Location (*this, start, end, _("session"), Location::IsSessionRange, 0);
Allow markers to be glued to bar/beat time. Fixes #1815. git-svn-id: svn://localhost/ardour2/branches/3.0@7573 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-09 12:40:31 -04:00
_locations->add (_session_range_location);
Create the session range location as and when the session first gets some content. Allows both the beginning and end of the range to expand to contain the actual session contents. git-svn-id: svn://localhost/ardour2/branches/3.0@7087 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-05-09 16:48:21 -04:00
}
Sort tracks in the port matrix according to editor order. Fixes #2966. git-svn-id: svn://localhost/ardour2/branches/3.0@7458 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-20 22:17:57 -04:00
step editing now accessed via a context menu on MIDI track rec-enable buttons. along the way, fixed up confusion between set_rec_enable() and set_rec_enabled() methods in Route/Track/Diskstream git-svn-id: svn://localhost/ardour2/branches/3.0@7482 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-24 12:40:56 -04:00
void
Session::step_edit_status_change (bool yn)
{
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
bool send = false;
step editing now accessed via a context menu on MIDI track rec-enable buttons. along the way, fixed up confusion between set_rec_enable() and set_rec_enabled() methods in Route/Track/Diskstream git-svn-id: svn://localhost/ardour2/branches/3.0@7482 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-24 12:40:56 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
bool val = false;
if (yn) {
send = (_step_editors == 0);
val = true;
step editing now accessed via a context menu on MIDI track rec-enable buttons. along the way, fixed up confusion between set_rec_enable() and set_rec_enabled() methods in Route/Track/Diskstream git-svn-id: svn://localhost/ardour2/branches/3.0@7482 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-24 12:40:56 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
_step_editors++;
} else {
send = (_step_editors == 1);
val = false;
step editing now accessed via a context menu on MIDI track rec-enable buttons. along the way, fixed up confusion between set_rec_enable() and set_rec_enabled() methods in Route/Track/Diskstream git-svn-id: svn://localhost/ardour2/branches/3.0@7482 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-24 12:40:56 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (_step_editors > 0) {
_step_editors--;
}
}
step editing now accessed via a context menu on MIDI track rec-enable buttons. along the way, fixed up confusion between set_rec_enable() and set_rec_enabled() methods in Route/Track/Diskstream git-svn-id: svn://localhost/ardour2/branches/3.0@7482 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-24 12:40:56 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (send) {
StepEditStatusChange (val);
}
step editing now accessed via a context menu on MIDI track rec-enable buttons. along the way, fixed up confusion between set_rec_enable() and set_rec_enabled() methods in Route/Track/Diskstream git-svn-id: svn://localhost/ardour2/branches/3.0@7482 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-07-24 12:40:56 -04:00
}
Suspend transport timecode transmission during playhead drag. Should fix #3324. git-svn-id: svn://localhost/ardour2/branches/3.0@7528 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-08-01 21:59:34 -04:00
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::start_time_changed (samplepos_t old)
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
{
/* Update the auto loop range to match the session range
(unless the auto loop range has been changed by the user)
*/
Fix crash on saving session template (#3634). git-svn-id: svn://localhost/ardour2/branches/3.0@8375 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-29 17:36:03 -05:00
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
Location* s = _locations->session_range_location ();
Fix crash on saving session template (#3634). git-svn-id: svn://localhost/ardour2/branches/3.0@8375 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-29 17:36:03 -05:00
if (s == 0) {
return;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
Location* l = _locations->auto_loop_location ();
Fix possible null pointer dereference (#4813). git-svn-id: svn://localhost/ardour2/branches/3.0@11814 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-04-07 06:05:17 -04:00
if (l && l->start() == old) {
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
l->set_start (s->start(), true);
}
Fix #6774, Moving start/end markers doesn't set Session as modified/dirty If a Session is saved and then the start or end session markers are moved, closing the Session will not result in "Unsaved Session" dialog being presented even though the Session has been modified. This change sets the Session as modified when either the Session start or end markers are modified.
2016-09-09 08:04:26 -04:00
set_dirty ();
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
}
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::end_time_changed (samplepos_t old)
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
{
/* Update the auto loop range to match the session range
(unless the auto loop range has been changed by the user)
*/
Location* s = _locations->session_range_location ();
Fix crash on saving session template (#3634). git-svn-id: svn://localhost/ardour2/branches/3.0@8375 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-12-29 17:36:03 -05:00
if (s == 0) {
return;
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
Location* l = _locations->auto_loop_location ();
Fix crash when moving end marker after deleting loop range (#4357). git-svn-id: svn://localhost/ardour2/branches/3.0@10196 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-10-07 15:49:21 -04:00
if (l && l->end() == old) {
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
l->set_end (s->end(), true);
}
Fix #6774, Moving start/end markers doesn't set Session as modified/dirty If a Session is saved and then the start or end session markers are moved, closing the Session will not result in "Unsaved Session" dialog being presented even though the Session has been modified. This change sets the Session as modified when either the Session start or end markers are modified.
2016-09-09 08:04:26 -04:00
set_dirty ();
Update auto loop range to match session range until it is changed by the user. Fixes #3472. git-svn-id: svn://localhost/ardour2/branches/3.0@7883 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-10-07 14:33:20 -04:00
}
initial pass at a missing file dialog and "relocatable" source files. lots more to do here git-svn-id: svn://localhost/ardour2/branches/3.0@7983 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 01:03:51 -05:00
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
std::vector<std::string>
initial pass at a missing file dialog and "relocatable" source files. lots more to do here git-svn-id: svn://localhost/ardour2/branches/3.0@7983 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 01:03:51 -05:00
Session::source_search_path (DataType type) const
{
Rename SearchPath class Searchpath Windows headers define SearchPath which means we have to undefine it where necessary. This is a pain and can be tricksy, so I feel renaming the class slightly is the easiest solution.
2013-08-15 06:04:08 -04:00
Searchpath sp;
improved WORKING fix for search path issues git-svn-id: svn://localhost/ardour2/branches/3.0@10940 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 13:52:14 -05:00
if (session_dirs.size() == 1) {
switch (type) {
case DataType::AUDIO:
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
sp.push_back (_session_dir->sound_path());
improved WORKING fix for search path issues git-svn-id: svn://localhost/ardour2/branches/3.0@10940 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 13:52:14 -05:00
break;
case DataType::MIDI:
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
sp.push_back (_session_dir->midi_path());
improved WORKING fix for search path issues git-svn-id: svn://localhost/ardour2/branches/3.0@10940 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 13:52:14 -05:00
break;
}
} else {
for (vector<space_and_path>::const_iterator i = session_dirs.begin(); i != session_dirs.end(); ++i) {
SessionDirectory sdir (i->path);
switch (type) {
case DataType::AUDIO:
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
sp.push_back (sdir.sound_path());
improved WORKING fix for search path issues git-svn-id: svn://localhost/ardour2/branches/3.0@10940 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 13:52:14 -05:00
break;
case DataType::MIDI:
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
sp.push_back (sdir.midi_path());
improved WORKING fix for search path issues git-svn-id: svn://localhost/ardour2/branches/3.0@10940 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 13:52:14 -05:00
break;
}
}
}
initial pass at a missing file dialog and "relocatable" source files. lots more to do here git-svn-id: svn://localhost/ardour2/branches/3.0@7983 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 01:03:51 -05:00
probable fix for not being able to find audio files in a 2.X session that had "illegal" characters in the session name - adds the 2.X version of the search path to the audio file search path, if it exists git-svn-id: svn://localhost/ardour2/branches/3.0@13321 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-10-23 10:52:26 -04:00
if (type == DataType::AUDIO) {
const string sound_path_2X = _session_dir->sound_path_2X();
if (Glib::file_test (sound_path_2X, Glib::FILE_TEST_EXISTS|Glib::FILE_TEST_IS_DIR)) {
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
if (find (sp.begin(), sp.end(), sound_path_2X) == sp.end()) {
sp.push_back (sound_path_2X);
do not add 2.X search path if its already present in the search path git-svn-id: svn://localhost/ardour2/branches/3.0@13503 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-11-14 16:54:04 -05:00
}
probable fix for not being able to find audio files in a 2.X session that had "illegal" characters in the session name - adds the 2.X version of the search path to the audio file search path, if it exists git-svn-id: svn://localhost/ardour2/branches/3.0@13321 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-10-23 10:52:26 -04:00
}
}
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
// now check the explicit (possibly user-specified) search path
fix up some logic so that we don't replicate the session dir path twice in the search path(s) git-svn-id: svn://localhost/ardour2/branches/3.0@10937 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 12:45:18 -05:00
switch (type) {
case DataType::AUDIO:
Rename SearchPath class Searchpath Windows headers define SearchPath which means we have to undefine it where necessary. This is a pain and can be tricksy, so I feel renaming the class slightly is the easiest solution.
2013-08-15 06:04:08 -04:00
sp += Searchpath(config.get_audio_search_path ());
fix up some logic so that we don't replicate the session dir path twice in the search path(s) git-svn-id: svn://localhost/ardour2/branches/3.0@10937 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 12:45:18 -05:00
break;
case DataType::MIDI:
Rename SearchPath class Searchpath Windows headers define SearchPath which means we have to undefine it where necessary. This is a pain and can be tricksy, so I feel renaming the class slightly is the easiest solution.
2013-08-15 06:04:08 -04:00
sp += Searchpath(config.get_midi_search_path ());
fix up some logic so that we don't replicate the session dir path twice in the search path(s) git-svn-id: svn://localhost/ardour2/branches/3.0@10937 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 12:45:18 -05:00
break;
}
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
return sp;
initial pass at a missing file dialog and "relocatable" source files. lots more to do here git-svn-id: svn://localhost/ardour2/branches/3.0@7983 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 01:03:51 -05:00
}
always store source names as relative paths; always add directories of embedded files to search path; never use hardlinks for embedding anymore git-svn-id: svn://localhost/ardour2/branches/3.0@7984 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 12:24:17 -05:00
void
Session::ensure_search_path_includes (const string& path, DataType type)
{
Rename SearchPath class Searchpath Windows headers define SearchPath which means we have to undefine it where necessary. This is a pain and can be tricksy, so I feel renaming the class slightly is the easiest solution.
2013-08-15 06:04:08 -04:00
Searchpath sp;
always store source names as relative paths; always add directories of embedded files to search path; never use hardlinks for embedding anymore git-svn-id: svn://localhost/ardour2/branches/3.0@7984 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 12:24:17 -05:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
if (path == ".") {
return;
}
missing dialog cleanups and logic improvements git-svn-id: svn://localhost/ardour2/branches/3.0@7990 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 18:50:20 -05:00
improved WORKING fix for search path issues git-svn-id: svn://localhost/ardour2/branches/3.0@10940 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 13:52:14 -05:00
switch (type) {
case DataType::AUDIO:
Rename SearchPath class Searchpath Windows headers define SearchPath which means we have to undefine it where necessary. This is a pain and can be tricksy, so I feel renaming the class slightly is the easiest solution.
2013-08-15 06:04:08 -04:00
sp += Searchpath(config.get_audio_search_path ());
improved WORKING fix for search path issues git-svn-id: svn://localhost/ardour2/branches/3.0@10940 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 13:52:14 -05:00
break;
case DataType::MIDI:
Rename SearchPath class Searchpath Windows headers define SearchPath which means we have to undefine it where necessary. This is a pain and can be tricksy, so I feel renaming the class slightly is the easiest solution.
2013-08-15 06:04:08 -04:00
sp += Searchpath (config.get_midi_search_path ());
improved WORKING fix for search path issues git-svn-id: svn://localhost/ardour2/branches/3.0@10940 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-07 13:52:14 -05:00
break;
}
Add path_is_within to decide if a path is within a given directory, taking symlinks into account, and use it to decide whether a file is within the session folder. Should fix #4552. git-svn-id: svn://localhost/ardour2/branches/3.0@12468 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-28 12:32:41 -04:00
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
for (vector<std::string>::iterator i = sp.begin(); i != sp.end(); ++i) {
Prevent duplication of search path elements when symlinks are involved (#4568). git-svn-id: svn://localhost/ardour2/branches/3.0@11592 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-03-04 19:34:45 -05:00
/* No need to add this new directory if it has the same inode as
an existing one; checking inode rather than name prevents duplicated
directories when we are using symlinks.
On Windows, I think we could just do if (*i == path) here.
*/
Move three file utility functions from pbd/filesystem.h to pbd/file_utils.h git-svn-id: svn://localhost/ardour2/branches/3.0@12863 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-23 01:08:14 -04:00
if (PBD::equivalent_paths (*i, path)) {
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
return;
}
}
always store source names as relative paths; always add directories of embedded files to search path; never use hardlinks for embedding anymore git-svn-id: svn://localhost/ardour2/branches/3.0@7984 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 12:24:17 -05:00
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
sp += path;
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
switch (type) {
case DataType::AUDIO:
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
config.set_audio_search_path (sp.to_string());
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
break;
case DataType::MIDI:
Fix portability of Session::source_search_path
2013-07-15 08:05:37 -04:00
config.set_midi_search_path (sp.to_string());
Fix more broken indentation (whitespace changes only). git-svn-id: svn://localhost/ardour2/branches/3.0@8094 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 14:57:03 -05:00
break;
}
always store source names as relative paths; always add directories of embedded files to search path; never use hardlinks for embedding anymore git-svn-id: svn://localhost/ardour2/branches/3.0@7984 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-09 12:24:17 -05:00
}
non-crashing (but also non-functional) integration of VBAP with panner "architecture" git-svn-id: svn://localhost/ardour2/branches/3.0@8056 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-18 19:58:57 -05:00
initial implementation of "bring all media into session folder". Incomplete but basically functional for audio files
2014-07-08 00:53:06 -04:00
void
Session::remove_dir_from_search_path (const string& dir, DataType type)
{
Searchpath sp;
switch (type) {
case DataType::AUDIO:
sp = Searchpath(config.get_audio_search_path ());
break;
case DataType::MIDI:
sp = Searchpath (config.get_midi_search_path ());
break;
}
sp -= dir;
switch (type) {
case DataType::AUDIO:
config.set_audio_search_path (sp.to_string());
break;
case DataType::MIDI:
config.set_midi_search_path (sp.to_string());
break;
}
}
more progress on speakers/vbap etc. etc (still a work in progress) git-svn-id: svn://localhost/ardour2/branches/3.0@8887 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-17 11:43:55 -05:00
boost::shared_ptr<Speakers>
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Session::get_speakers()
non-crashing (but also non-functional) integration of VBAP with panner "architecture" git-svn-id: svn://localhost/ardour2/branches/3.0@8056 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-18 19:58:57 -05:00
{
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
return _speakers;
non-crashing (but also non-functional) integration of VBAP with panner "architecture" git-svn-id: svn://localhost/ardour2/branches/3.0@8056 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-18 19:58:57 -05:00
}
Handle unknown plugins on loading sessions by hiding them from the user and telling them so. git-svn-id: svn://localhost/ardour2/branches/3.0@8097 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-26 18:30:48 -05:00
list<string>
Session::unknown_processors () const
{
list<string> p;
boost::shared_ptr<RouteList> r = routes.reader ();
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
list<string> t = (*i)->unknown_processors ();
copy (t.begin(), t.end(), back_inserter (p));
}
p.sort ();
p.unique ();
return p;
}
provide solo mute gain Config parameter as a controllable and use it in the monitor section GUI to allow MIDI binding git-svn-id: svn://localhost/ardour2/branches/3.0@8825 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-11 11:14:54 -05:00
reinstate missing latency initialization code Removed (mistakenly) during work on transportFSM
2019-12-16 18:13:27 -05:00
void
Session::initialize_latencies ()
{
{
Glib::Threads::Mutex::Lock lm (_engine.process_lock());
update_latency (false);
update_latency (true);
}
set_worst_io_latencies ();
}
part two of new JACK latency API support - should fix broken compilation git-svn-id: svn://localhost/ardour2/branches/3.0@8864 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-15 14:55:14 -05:00
void
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
Session::set_worst_io_latencies_x (IOChange, void *)
part two of new JACK latency API support - should fix broken compilation git-svn-id: svn://localhost/ardour2/branches/3.0@8864 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-15 14:55:14 -05:00
{
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
set_worst_io_latencies ();
}
another (non critical) concurrency issue: Ignore latency-updates while removing tracks. remove_tracks() calls disconnect_all() which can trigger the backend to perform a latency-recompute in a different thread (process thread). That calls Session::update_latency() which iterates over all routes including the one(s) to be deleted. At the same time disconnect_all() may propagate Port::PostDisconnect() leading to a lock contention. (no crash just a lot of unnecessary work. remove_routes() calls update_latency_compensation() which triggers a recompute if needed)
2016-05-03 07:56:08 -04:00
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
void
Session::send_latency_compensation_change ()
{
/* As a result of Send::set_output_latency()
* or InternalReturn::set_playback_offset ()
* the send's own latency can change (source track
* is aligned with target bus).
*
* This can only happen be triggered by
* Route::update_signal_latency ()
* when updating the processor latency.
*
* We need to walk the graph again to take those changes into account
* (we should probably recurse or process the graph in a 2 step process).
*/
++_send_latency_changes;
}
bool
Session::update_route_latency (bool playback, bool apply_to_delayline)
{
/* Note: RouteList is process-graph sorted */
boost::shared_ptr<RouteList> r = routes.reader ();
if (playback) {
/* reverse the list so that we work backwards from the last route to run to the first,
* this is not needed, but can help to reduce the iterations for aux-sends.
*/
RouteList* rl = routes.reader().get();
r.reset (new RouteList (*rl));
reverse (r->begin(), r->end());
}
bool changed = false;
int bailout = 0;
restart:
_send_latency_changes = 0;
_worst_route_latency = 0;
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
// if (!(*i)->active()) { continue ; } // TODO
samplecnt_t l;
if ((*i)->signal_latency () != (l = (*i)->update_signal_latency (apply_to_delayline))) {
changed = true;
}
_worst_route_latency = std::max (l, _worst_route_latency);
}
if (_send_latency_changes > 0) {
// only 1 extra iteration is needed (we allow only 1 level of aux-sends)
// BUT.. jack'n'sends'n'bugs
if (++bailout < 5) {
cerr << "restarting Session::update_latency. # of send changes: " << _send_latency_changes << " iteration: " << bailout << endl;
goto restart;
}
}
DEBUG_TRACE (DEBUG::Latency, string_compose ("worst signal processing latency: %1 (changed ? %2)\n", _worst_route_latency, (changed ? "yes" : "no")));
return changed;
}
void
Session::update_latency (bool playback)
{
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
/* called only from AudioEngine::latency_callback.
* but may indirectly be triggered from
* Session::update_latency_compensation -> _engine.update_latencies
*/
expand DEBUG_TRACE output
2019-12-16 18:13:02 -05:00
DEBUG_TRACE (DEBUG::LatencyCompensation, string_compose ("Engine latency callback: %1 (initial/deletion? %2)\n", (playback ? "PLAYBACK" : "CAPTURE"), inital_connect_or_deletion_in_progress()));
fixes for latency computation and compilation git-svn-id: svn://localhost/ardour2/branches/3.0@8868 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-15 22:25:23 -05:00
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (inital_connect_or_deletion_in_progress () || _adding_routes_in_progress || _route_deletion_in_progress) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
return;
}
Allow adding tracks w/o running backend
2017-07-20 10:45:14 -04:00
if (!_engine.running()) {
return;
}
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
NO-OP: whitespace
2017-09-21 22:15:56 -04:00
/* Note; RouteList is sorted as process-graph */
fixes for latency computation and compilation git-svn-id: svn://localhost/ardour2/branches/3.0@8868 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-15 22:25:23 -05:00
boost::shared_ptr<RouteList> r = routes.reader ();
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
if (playback) {
/* reverse the list so that we work backwards from the last route to run to the first */
NO-OP: whitespace
2017-09-21 22:15:56 -04:00
RouteList* rl = routes.reader().get();
r.reset (new RouteList (*rl));
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
reverse (r->begin(), r->end());
}
fixes for latency computation and compilation git-svn-id: svn://localhost/ardour2/branches/3.0@8868 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-02-15 22:25:23 -05:00
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
samplecnt_t latency = (*i)->set_private_port_latencies (playback);
(*i)->set_public_port_latencies (latency, playback);
NO-OP: whitespace
2017-09-21 22:15:56 -04:00
}
Fix broken whitespace. I'd apologize for the compile times if it was my fault :D git-svn-id: svn://localhost/ardour2/branches/3.0@9654 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-06-01 12:50:12 -04:00
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
if (playback) {
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
Glib::Threads::Mutex::Lock lx (_update_latency_lock);
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
set_worst_output_latency ();
update_route_latency (true, true);
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
} else {
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
Glib::Threads::Mutex::Lock lx (_update_latency_lock);
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
set_worst_input_latency ();
update_route_latency (false, false);
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
}
properly use and publish private/public port latency values for JACK latency API git-svn-id: svn://localhost/ardour2/branches/3.0@9130 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-12 15:25:09 -05:00
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
DEBUG_TRACE (DEBUG::LatencyCompensation, "Engine latency callback: DONE\n");
Emit signal when session latency was updated
2019-09-28 16:29:50 -04:00
LatencyUpdated (); /* EMIT SIGNAL */
sorta-kinda working latency compensation, latency reporting and capture alignment ... working except that we report the wrong information to JACK and i've noticed a couple of odd circumstances where turning on a latent plugin caused punch recording to fail git-svn-id: svn://localhost/ardour2/branches/3.0@9121 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-10 21:55:52 -05:00
}
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
void
Session::set_worst_io_latencies ()
{
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
DEBUG_TRACE (DEBUG::LatencyCompensation, "Session::set_worst_io_latencies\n");
Glib::Threads::Mutex::Lock lx (_update_latency_lock);
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
set_worst_output_latency ();
set_worst_input_latency ();
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
}
void
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
Session::set_worst_output_latency ()
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
{
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (inital_connect_or_deletion_in_progress ()) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
return;
}
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
_worst_output_latency = 0;
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
Consolidate ambiguous engine API calls available(), connected(), running() were ill-defined and used interchangeably.
2018-11-28 09:24:47 -05:00
if (!_engine.running()) {
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
return;
}
boost::shared_ptr<RouteList> r = routes.reader ();
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
_worst_output_latency = max (_worst_output_latency, (*i)->output()->latency());
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
}
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
_worst_output_latency = max (_worst_output_latency, _click_io->latency());
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
DEBUG_TRACE (DEBUG::LatencyCompensation, string_compose ("Worst output latency: %1\n", _worst_output_latency));
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
}
void
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
Session::set_worst_input_latency ()
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
{
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (inital_connect_or_deletion_in_progress ()) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
return;
}
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
_worst_input_latency = 0;
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
Consolidate ambiguous engine API calls available(), connected(), running() were ill-defined and used interchangeably.
2018-11-28 09:24:47 -05:00
if (!_engine.running()) {
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
return;
}
boost::shared_ptr<RouteList> r = routes.reader ();
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
Fix broken indentation. git-svn-id: svn://localhost/ardour2/branches/3.0@9151 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 17:53:10 -04:00
_worst_input_latency = max (_worst_input_latency, (*i)->input()->latency());
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
}
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
DEBUG_TRACE (DEBUG::LatencyCompensation, string_compose ("Worst input latency: %1\n", _worst_input_latency));
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
}
void
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
Session::update_latency_compensation (bool force_whole_graph, bool called_from_backend)
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
{
better comments
2019-10-28 18:53:00 -04:00
/* Called to update Ardour's internal latency values and compensation
* planning. Typically case is from within ::graph_reordered()
*/
NO-OP: Use API to set/test state-of-the-state
2019-03-18 10:33:05 -04:00
if (inital_connect_or_deletion_in_progress ()) {
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
return;
}
better comments
2019-10-28 18:53:00 -04:00
Fix race-condition/deadlock, plugin-copy while rolling lili93's session (#ardour) triggered this w/jackd 512fpp: Drag/Drop copy a latent plugin from one track to another while rolling. The GUI-thread as well as the auto-connect thread concurrently call jack_recompute_total_latencies(). The auto-connect thread holds a process lock while doing so. The GUI does not use any mutexes. This randomly deadlocks in libjack. backtrace: https://pastebin.com/6m3KGhWS
2018-10-24 20:00:08 -04:00
/* this lock is not usually contended, but under certain conditions,
* update_latency_compensation may be called concurrently.
* e.g. drag/drop copy a latent plugin while rolling.
* GUI thread (via route_processors_changed) and
* auto_connect_thread_run may race.
*/
Glib::Threads::Mutex::Lock lx (_update_latency_lock, Glib::Threads::TRY_LOCK);
if (!lx.locked()) {
/* no need to do this twice */
return;
}
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
DEBUG_TRACE (DEBUG::LatencyCompensation, string_compose ("update_latency_compensation %1\n", (force_whole_graph ? "of whole graph" : "")));
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
bool some_track_latency_changed = update_route_latency (false, false);
NOOP, remove trailing tabs/whitespace.
2015-10-05 10:17:49 -04:00
fixes from 2.X for latency/capture alignment stuff: don't reverse route list, update latency and capture offsets at transport stop; remove some debugging output git-svn-id: svn://localhost/ardour2/branches/3.0@10110 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-09-21 13:28:13 -04:00
if (some_track_latency_changed || force_whole_graph) {
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
do not hold a lock when calling AudioEngine::update_latencies() from Session::update_latency_compensation(). Only when using JACK1 is ::update_latencies() a synchronous call (ending up in Session::update_latency() which tries to take the same lock). But the semantics of ::update_latencies() are sufficiently ill-defined that entering that call with a lock held seems like a bad idea, so we release the lock unconditionally here.
2019-10-28 15:39:59 -04:00
/* cannot hold lock while engine initiates a full latency callback */
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
do not hold a lock when calling AudioEngine::update_latencies() from Session::update_latency_compensation(). Only when using JACK1 is ::update_latencies() a synchronous call (ending up in Session::update_latency() which tries to take the same lock). But the semantics of ::update_latencies() are sufficiently ill-defined that entering that call with a lock held seems like a bad idea, so we release the lock unconditionally here.
2019-10-28 15:39:59 -04:00
lx.release ();
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
better comments
2019-10-28 18:53:00 -04:00
/* next call will ask the backend up update its latencies.
*
* The semantics of how the backend does this are not well
* defined (Oct 2019).
*
* In all cases, eventually AudioEngine::latency_callback() is
* invoked, which will call Session::update_latency().
*
* Some backends will do that asynchronously with respect to
* this call. Others (JACK1) will do so synchronously, and in
* those cases this call will return until the backend latency
* callback is complete.
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
*
* Further, if this is called as part of a backend callback,
* then we have to follow the JACK1 rule that we cannot call
* back into the backend during such a callback (otherwise
* deadlock ensues).
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
*/
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
if (!called_from_backend) {
improve debug output
2019-10-28 19:34:38 -04:00
DEBUG_TRACE (DEBUG::LatencyCompensation, "update_latency_compensation: delegate to engine\n");
to avoid deadlock in JACK1 scenarios, do not invoke AudioEngine::update_latencies() from update_latency_compensation() if called from a process thread
2019-10-28 18:55:29 -04:00
_engine.update_latencies ();
improve debug output
2019-10-28 19:34:38 -04:00
} else {
DEBUG_TRACE (DEBUG::LatencyCompensation, "update_latency_compensation called from engine, don't call back into engine\n");
to avoid deadlock in JACK1 scenarios, do not invoke AudioEngine::update_latencies() from update_latency_compensation() if called from a process thread
2019-10-28 18:55:29 -04:00
}
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
} else {
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
DEBUG_TRACE (DEBUG::LatencyCompensation, "update_latency_compensation: directly apply to routes\n");
Aux-Send Latency compensation, part 2 & code-consolidation
2017-09-28 22:17:16 -04:00
boost::shared_ptr<RouteList> r = routes.reader ();
for (RouteList::iterator i = r->begin(); i != r->end(); ++i) {
(*i)->apply_latency_compensation ();
}
fixes from 2.X for latency/capture alignment stuff: don't reverse route list, update latency and capture offsets at transport stop; remove some debugging output git-svn-id: svn://localhost/ardour2/branches/3.0@10110 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-09-21 13:28:13 -04:00
}
Fix latency compensation race-condition Remove need for explicit `initialize_latencies` call that used to be called from GUI-thread post_engine_init(), as well as Session::engine_running(). Further reduce calls, `graph_reordered` implies a latency-update and fix ordering issue. update_latency_compensation() must be called *after* resort_routes().
2019-10-25 19:06:04 -04:00
DEBUG_TRACE (DEBUG::LatencyCompensation, "update_latency_compensation: DONE\n");
more fun and games with latency compensation, and so forth ... not done yet git-svn-id: svn://localhost/ardour2/branches/3.0@9149 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-03-14 16:33:47 -04:00
}
centralize legal-session-name-checkng and include : and ; in characters that we disallow, because they conflict with search path conventions on *nix and windows git-svn-id: svn://localhost/ardour2/branches/3.0@10943 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-08 22:06:58 -05:00
char
Session::session_name_is_legal (const string& path)
{
char illegal_chars[] = { '/', '\\', ':', ';', '\0' };
for (int i = 0; illegal_chars[i]; ++i) {
if (path.find (illegal_chars[i]) != string::npos) {
return illegal_chars[i];
}
}
return 0;
}
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
first pass at the big rethink of managing sort order keys for editor and mixer. this appears to work, but remote control IDs are not yet correct (frequently off by one because of the presence of the master bus in the editor) git-svn-id: svn://localhost/ardour2/branches/3.0@12953 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-27 18:57:06 -04:00
void
first vaguely working version using PresentationInfo remote control ID and "order keys" have been removed.
2016-05-16 07:30:28 -04:00
Session::notify_presentation_info_change ()
first pass at the big rethink of managing sort order keys for editor and mixer. this appears to work, but remote control IDs are not yet correct (frequently off by one because of the presence of the master bus in the editor) git-svn-id: svn://localhost/ardour2/branches/3.0@12953 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-27 18:57:06 -04:00
{
if (deletion_in_progress()) {
return;
drastic rethink of the relationship between remote control ID and route order keys. unless the user explicitly switches to UserOrdered, Route::_remote_control_id is an unallocated pointer, and Route::remote_control_id() simply returns a value based on the relevant order_key() value. Also, change the key used in the Route::order_keys std::map<> from a string to an enum, since there is no evidence that we are benefitting from the theoretical benefit of using a string. Generally tidy up allocation of order keys so that the master and monitor busses always get a "special" MixerSort key value, based on the MMC ID for master (already defined within Ardour), and all other tracks/busses start at zero. Syncing keys between editor and mixer will leave the MixerSort key for the master and monitor bus alone, reflecting the fact that we display these in their own distinct parts of the GUI and they are not orderable like other tracks or busses within the mixer window git-svn-id: svn://localhost/ardour2/branches/3.0@12923 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-25 08:46:13 -04:00
}
implement track numbering
2014-06-25 15:16:09 -04:00
reassign_track_numbers();
make monitor section an optional feature than can be added/removed as needed. this is a big commit, and breakage is possible. it has been moderately tested. this commit also locks the remote control ID of the master bus to 318 and the monitor section (if any) to 319. the numbers are based on MIDI Machine Control limits git-svn-id: svn://localhost/ardour2/branches/3.0@11256 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-01-17 20:30:44 -05:00
}
Add operation_in_progress() method. git-svn-id: svn://localhost/ardour2/branches/3.0@12704 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-06-13 12:46:59 -04:00
bool
Session::operation_in_progress (GQuark op) const
{
return (find (_current_trans_quarks.begin(), _current_trans_quarks.end(), op) != _current_trans_quarks.end());
}
move ownership of LTC I/O ports to Session, and manage as IO objects git-svn-id: svn://localhost/ardour2/branches/3.0@13341 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-10-25 15:46:23 -04:00
boost::shared_ptr<Port>
Session::ltc_output_port () const
{
Fix some more crashes at session-close PortDrop unregisters backend ports, the backend will return a NULL buffer-pointer, but the I/O object still exists (Metronome, LTC) AudioBuffer::_data == 0 #0 msvcrt!memset () from C:\Windows\System32\msvcrt.dll #1 ARDOUR::AudioBuffer::silence (this=0x2c410710, len=256, offset=0) at ../libs/ardour/audio_buffer.cc:88 #2 ARDOUR::AudioPort::cycle_end (this=0x34918730, nframes=256) at ../libs/ardour/audio_port.cc:66 #3 ARDOUR::PortManager::cycle_end (this=this@entry=0x23342770, nframes=nframes@entry=256) [process callback w/o session]
2017-08-01 17:27:34 -04:00
return _ltc_output ? _ltc_output->nth (0) : boost::shared_ptr<Port> ();
move ownership of LTC I/O ports to Session, and manage as IO objects git-svn-id: svn://localhost/ardour2/branches/3.0@13341 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-10-25 15:46:23 -04:00
}
void
Session::reconnect_ltc_output ()
{
if (_ltc_output) {
(re)implement Yevgeny's LTC branch changes in the context of ardour-merge-from-tracks (just the libs/* section)
2015-05-25 20:56:03 -04:00
string src = Config->get_ltc_output_port();
move ownership of LTC I/O ports to Session, and manage as IO objects git-svn-id: svn://localhost/ardour2/branches/3.0@13341 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-10-25 15:46:23 -04:00
_ltc_output->disconnect (this);
if (src != _("None") && !src.empty()) {
_ltc_output->nth (0)->connect (src);
}
}
}
initial version of playback priority design. No GUI control over options yet Conflicts: libs/ardour/ardour/session.h libs/ardour/ardour/types.h libs/ardour/enums.cc libs/ardour/session_transport.cc system_config
2015-01-16 12:17:09 -05:00
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::set_range_selection (samplepos_t start, samplepos_t end)
initial version of playback priority design. No GUI control over options yet Conflicts: libs/ardour/ardour/session.h libs/ardour/ardour/types.h libs/ardour/enums.cc libs/ardour/session_transport.cc system_config
2015-01-16 12:17:09 -05:00
{
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
_range_selection = Evoral::Range<samplepos_t> (start, end);
initial version of playback priority design. No GUI control over options yet Conflicts: libs/ardour/ardour/session.h libs/ardour/ardour/types.h libs/ardour/enums.cc libs/ardour/session_transport.cc system_config
2015-01-16 12:17:09 -05:00
}
void
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
Session::set_object_selection (samplepos_t start, samplepos_t end)
initial version of playback priority design. No GUI control over options yet Conflicts: libs/ardour/ardour/session.h libs/ardour/ardour/types.h libs/ardour/enums.cc libs/ardour/session_transport.cc system_config
2015-01-16 12:17:09 -05:00
{
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
_object_selection = Evoral::Range<samplepos_t> (start, end);
initial version of playback priority design. No GUI control over options yet Conflicts: libs/ardour/ardour/session.h libs/ardour/ardour/types.h libs/ardour/enums.cc libs/ardour/session_transport.cc system_config
2015-01-16 12:17:09 -05:00
}
void
Session::clear_range_selection ()
{
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
_range_selection = Evoral::Range<samplepos_t> (-1,-1);
initial version of playback priority design. No GUI control over options yet Conflicts: libs/ardour/ardour/session.h libs/ardour/ardour/types.h libs/ardour/enums.cc libs/ardour/session_transport.cc system_config
2015-01-16 12:17:09 -05:00
}
void
Session::clear_object_selection ()
{
globally change all use of "frame" to refer to audio into "sample". Generated by tools/f2s. Some hand-editing will be required in a few places to fix up comments related to timecode and video in order to keep the legible
2017-09-18 12:39:17 -04:00
_object_selection = Evoral::Range<samplepos_t> (-1,-1);
initial version of playback priority design. No GUI control over options yet Conflicts: libs/ardour/ardour/session.h libs/ardour/ardour/types.h libs/ardour/enums.cc libs/ardour/session_transport.cc system_config
2015-01-16 12:17:09 -05:00
}
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
void
Session::auto_connect_route (boost::shared_ptr<Route> route, bool connect_inputs,
const ChanCount& input_start,
const ChanCount& output_start,
const ChanCount& input_offset,
const ChanCount& output_offset)
{
Glib::Threads::Mutex::Lock lx (_auto_connect_queue_lock);
_auto_connect_queue.push (AutoConnectRequest (route, connect_inputs,
input_start, output_start,
input_offset, output_offset));
new scheme for managing port deletion shared_ptr<Port> now uses a deleter functor which pushes Port* to a lock-free FIFO so that the Port is always deleted (and thus unregistered with the PortEngine/backend) in a safe context w.r.t. various callbacks in the host. Currently the auto_connect_thread in Session has been tasked with doing these deletions.
2016-10-13 17:18:42 -04:00
auto_connect_thread_wakeup ();
}
void
Session::auto_connect_thread_wakeup ()
{
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
if (pthread_mutex_trylock (&_auto_connect_mutex) == 0) {
pthread_cond_signal (&_auto_connect_cond);
pthread_mutex_unlock (&_auto_connect_mutex);
}
}
move latency-recompute into dedicated thread. this fixes an issue with jack1 and jack_latency_recompute() since must not send a server request from inside the server callback.
2016-07-09 11:42:58 -04:00
void
Session::queue_latency_recompute ()
{
g_atomic_int_inc (&_latency_recompute_pending);
new scheme for managing port deletion shared_ptr<Port> now uses a deleter functor which pushes Port* to a lock-free FIFO so that the Port is always deleted (and thus unregistered with the PortEngine/backend) in a safe context w.r.t. various callbacks in the host. Currently the auto_connect_thread in Session has been tasked with doing these deletions.
2016-10-13 17:18:42 -04:00
auto_connect_thread_wakeup ();
move latency-recompute into dedicated thread. this fixes an issue with jack1 and jack_latency_recompute() since must not send a server request from inside the server callback.
2016-07-09 11:42:58 -04:00
}
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
void
Session::auto_connect (const AutoConnectRequest& ar)
{
boost::shared_ptr<Route> route = ar.route.lock();
if (!route) { return; }
if (!IO::connecting_legal) {
return;
}
/* If both inputs and outputs are auto-connected to physical ports,
* use the max of input and output offsets to ensure auto-connected
* port numbers always match up (e.g. the first audio input and the
* first audio output of the route will have the same physical
* port number). Otherwise just use the lowest input or output
* offset possible.
*/
const bool in_out_physical =
(Config->get_input_auto_connect() & AutoConnectPhysical)
&& (Config->get_output_auto_connect() & AutoConnectPhysical)
&& ar.connect_inputs;
const ChanCount in_offset = in_out_physical
? ChanCount::max(ar.input_offset, ar.output_offset)
: ar.input_offset;
const ChanCount out_offset = in_out_physical
? ChanCount::max(ar.input_offset, ar.output_offset)
: ar.output_offset;
for (DataType::iterator t = DataType::begin(); t != DataType::end(); ++t) {
vector<string> physinputs;
vector<string> physoutputs;
when fetching MIDI ports for bundles and auto-connection, ignore control-only (and "virtual") MIDI ports "virtual" is a placeholder name for ALSA sequencer MIDI through ports
2016-10-21 16:32:46 -04:00
/* for connecting track inputs we only want MIDI ports marked
* for "music".
*/
get_physical_ports (physinputs, physoutputs, *t, MidiPortMusic);
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
if (!physinputs.empty() && ar.connect_inputs) {
uint32_t nphysical_in = physinputs.size();
for (uint32_t i = ar.input_start.get(*t); i < route->n_inputs().get(*t) && i < nphysical_in; ++i) {
string port;
if (Config->get_input_auto_connect() & AutoConnectPhysical) {
port = physinputs[(in_offset.get(*t) + i) % nphysical_in];
}
if (!port.empty() && route->input()->connect (route->input()->ports().port(*t, i), port, this)) {
break;
}
}
}
if (!physoutputs.empty()) {
uint32_t nphysical_out = physoutputs.size();
for (uint32_t i = ar.output_start.get(*t); i < route->n_outputs().get(*t); ++i) {
string port;
if ((*t) == DataType::MIDI && (Config->get_output_auto_connect() & AutoConnectPhysical)) {
port = physoutputs[(out_offset.get(*t) + i) % nphysical_out];
} else if ((*t) == DataType::AUDIO && (Config->get_output_auto_connect() & AutoConnectMaster)) {
/* master bus is audio only */
if (_master_out && _master_out->n_inputs().get(*t) > 0) {
port = _master_out->input()->ports().port(*t,
i % _master_out->input()->n_ports().get(*t))->name();
}
}
if (!port.empty() && route->output()->connect (route->output()->ports().port(*t, i), port, this)) {
break;
}
}
}
}
}
void
Session::auto_connect_thread_start ()
{
Session::_ac_thread_active should be used with atomics to create memory barriers, even though lock ordering is likely correct now
2016-12-19 10:33:54 -05:00
if (g_atomic_int_get (&_ac_thread_active)) {
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
return;
}
Disconnect Signals before dropping ports significantly speeds up session close
2016-04-23 18:24:43 -04:00
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
while (!_auto_connect_queue.empty ()) {
_auto_connect_queue.pop ();
}
Session::_ac_thread_active should be used with atomics to create memory barriers, even though lock ordering is likely correct now
2016-12-19 10:33:54 -05:00
g_atomic_int_set (&_ac_thread_active, 1);
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
if (pthread_create (&_auto_connect_thread, NULL, auto_connect_thread, this)) {
Session::_ac_thread_active should be used with atomics to create memory barriers, even though lock ordering is likely correct now
2016-12-19 10:33:54 -05:00
g_atomic_int_set (&_ac_thread_active, 0);
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
}
}
void
Session::auto_connect_thread_terminate ()
{
Session::_ac_thread_active should be used with atomics to create memory barriers, even though lock ordering is likely correct now
2016-12-19 10:33:54 -05:00
if (!g_atomic_int_get (&_ac_thread_active)) {
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
return;
}
Disconnect Signals before dropping ports significantly speeds up session close
2016-04-23 18:24:43 -04:00
{
Glib::Threads::Mutex::Lock lx (_auto_connect_queue_lock);
while (!_auto_connect_queue.empty ()) {
_auto_connect_queue.pop ();
}
}
Session::auto_connect_thread_terminate() cannot use ::auto_connect_thread_wakeup() because that method is allowed to "fail". Should fix hangs when closing a session with the ALSA backend (and perhaps others too).
2016-12-19 09:35:29 -05:00
/* cannot use auto_connect_thread_wakeup() because that is allowed to
* fail to wakeup the thread.
*/
pthread_mutex_lock (&_auto_connect_mutex);
Session::_ac_thread_active should be used with atomics to create memory barriers, even though lock ordering is likely correct now
2016-12-19 10:33:54 -05:00
g_atomic_int_set (&_ac_thread_active, 0);
Session::auto_connect_thread_terminate() cannot use ::auto_connect_thread_wakeup() because that method is allowed to "fail". Should fix hangs when closing a session with the ALSA backend (and perhaps others too).
2016-12-19 09:35:29 -05:00
pthread_cond_signal (&_auto_connect_cond);
pthread_mutex_unlock (&_auto_connect_mutex);
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
void *status;
pthread_join (_auto_connect_thread, &status);
}
void *
Session::auto_connect_thread (void *arg)
{
Session *s = static_cast<Session *>(arg);
s->auto_connect_thread_run ();
pthread_exit (0);
return 0;
}
void
Session::auto_connect_thread_run ()
{
pthread_set_name (X_("autoconnect"));
we need the process lock after all..
2016-04-24 10:06:38 -04:00
SessionEvent::create_per_thread_pool (X_("autoconnect"), 1024);
PBD::notify_event_loops_about_thread_creation (pthread_self(), X_("autoconnect"), 1024);
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
pthread_mutex_lock (&_auto_connect_mutex);
Session::_ac_thread_active should be used with atomics to create memory barriers, even though lock ordering is likely correct now
2016-12-19 10:33:54 -05:00
while (g_atomic_int_get (&_ac_thread_active)) {
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
we need the process lock after all..
2016-04-24 10:06:38 -04:00
if (!_auto_connect_queue.empty ()) {
// Why would we need the process lock ??
// A: if ports are added while we're connecting, the backend's iterator may be invalidated:
// graph_order_callback() -> resort_routes() -> direct_feeds_according_to_reality () -> backend::connected_to()
// All ardour-internal backends use a std::vector xxxAudioBackend::find_port()
// We have control over those, but what does jack do?
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
Glib::Threads::Mutex::Lock lx (_auto_connect_queue_lock);
we need the process lock after all..
2016-04-24 10:06:38 -04:00
while (!_auto_connect_queue.empty ()) {
const AutoConnectRequest ar (_auto_connect_queue.front());
_auto_connect_queue.pop ();
lx.release ();
auto_connect (ar);
lx.acquire ();
}
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
}
better safe than sorry
2016-07-09 12:31:40 -04:00
if (!actively_recording ()) { // might not be needed,
/* this is only used for updating plugin latencies, the
* graph does not change. so it's safe in general.
* BUT..
Ongoing work on latency compensation The general goal is to align transport-sample to be the audible frame and use that as "anchor" for all processing. transport_sample cannot become negative (00:00:00:00 is the first audible frame). Internally transport pre-rolls (read-ahead) before the transport starts to move. This allows inputs and disk to prefill the pipeline. When starting to roll, the session counts down a global "remaning preroll" counter, which is the worst-latency from in-to-out. Each route in turn will start processing at its own output-latency. Route::process_output_buffers() - which does the actual processing incl disk i/o - begins by offsetting the "current sample" by the route's process-latency and decrements the offset for each latent processor. At the end of the function the output will be aligned and match transport-sample - downstream-playback-latency (if any). PS. This commit is a first step only: transport looping & vari-speed have not yet been implemented/updated.
2017-09-28 00:31:12 -04:00
* update_latency_compensation ()
* calls DiskWriter::set_capture_offset () which
* modifies the capture-offset, which can be a problem.
better safe than sorry
2016-07-09 12:31:40 -04:00
*/
while (g_atomic_int_and (&_latency_recompute_pending, 0)) {
Optimize Delaylines: block-process (not one sample at a time) This also tweaks fade behavior when the latency changes to prefer a x-fade when possible. This new variant does not support concurrent re-allocation and execution. Hence the auto-connect thread needs to take a lock before updating latencies (actually there's no need for an explicit update with built-in backends, so this case remains to be updated further)
2017-11-04 00:20:14 -04:00
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
provide a mechanism to decide if Session::update_latency_compensation() is being called as part of a callback from the backend. If it is, do not call AudioEngine::update_latencies() to avoid JACK1-style deadlock
2019-10-28 19:23:54 -04:00
update_latency_compensation (false, false);
better safe than sorry
2016-07-09 12:31:40 -04:00
}
move latency-recompute into dedicated thread. this fixes an issue with jack1 and jack_latency_recompute() since must not send a server request from inside the server callback.
2016-07-09 11:42:58 -04:00
}
Fix possible crash when removing ports with jack1
2016-12-12 16:47:16 -05:00
{
// this may call ARDOUR::Port::drop ... jack_port_unregister ()
// jack1 cannot cope with removing ports while processing
Glib::Threads::Mutex::Lock lm (AudioEngine::instance()->process_lock ());
AudioEngine::instance()->clear_pending_port_deletions ();
}
new scheme for managing port deletion shared_ptr<Port> now uses a deleter functor which pushes Port* to a lock-free FIFO so that the Port is always deleted (and thus unregistered with the PortEngine/backend) in a safe context w.r.t. various callbacks in the host. Currently the auto_connect_thread in Session has been tasked with doing these deletions.
2016-10-13 17:18:42 -04:00
delegate port-connections to low priority thread. This prevents a deadlock with (some versions) jack: * add new instrument track with instrument * configure processors (takes processor lock) * add I/Os (delivery) -> create ports * auto-connect ports * jack port-connect -> jack graph re-order * Ardour graph-re-order * needs processor lock (to check sends)
2016-04-23 16:11:48 -04:00
pthread_cond_wait (&_auto_connect_cond, &_auto_connect_mutex);
}
pthread_mutex_unlock (&_auto_connect_mutex);
}
a better, deeper fix for "cancel all solo", as Session::cancel_all_solo()
2016-07-13 14:33:23 -04:00
void
Session::cancel_all_solo ()
{
StripableList sl;
get_stripables (sl);
set_controls (stripable_list_to_control_list (sl, &Stripable::solo_control), 0.0, Controllable::NoGroup);
clear_all_solo_state (routes.reader());
}
some rather simple code to allow MIDI Clock to set the tempo of the session. This is only allowed if the session has only 1 tempo marker
2018-10-08 12:59:51 -04:00
void
Session::maybe_update_tempo_from_midiclock_tempo (float bpm)
{
if (_tempo_map->n_tempos() == 1) {
TempoSection& ts (_tempo_map->tempo_section_at_sample (0));
if (fabs (ts.note_types_per_minute() - bpm) > (0.01 * ts.note_types_per_minute())) {
const Tempo tempo (bpm, 4.0, bpm);
std::cerr << "new tempo " << bpm << " old " << ts.note_types_per_minute() << std::endl;
_tempo_map->replace_tempo (ts, tempo, 0.0, 0.0, AudioTime);
}
}
}