- for those not in the know, this series provides a way to
remove the temporal distortion introduced when using an
audio frame-based gui for music-locked objects.
In short, the gui uses an audio frame representation to move
objects. It displays the object using frame_at_beat(), quantizing
the time value to audio frames. This is fine until the user selects
that frame but expects it to be interpreted as a beat.
Thus beat_at_frame() would not produce the user-expected beat
(temporal quantization error of up to 0.5 audio samples).
This is one method of mapping audio time to music time accurately.
- use exact beats to determine frame position.
- see comments in tempo.cc for more.
- this hasn't been done for split yet, but dragging and
trimming are supported.
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.
When most internal sends are created, they are given a panner shell
which is then responsible for audio dispatch. Other data types were left
there without handling them at all. Ensure that all available data is
sent provided the internal send has enough outgoing buffers.
Note that checking the number of output ports is not needed because
IO::copy_to_outputs() will stop if there are less ports of the right
type than buffers (or even none).
IO::copy_to_output() crashed if there was no channel to copy from. Since
all callers seem to check before calling, just assert() that it is not
the case.
in particular: lua-lifefime (!) C++ instances.
This allows for dynamic allocation of custom user-data, bound to
the lifetime of the allocating lua-context.
If a route has strict I/O then the main delivery follows the output of
the last processor, but libardour ensures it has at least as many
outputs as inputs in the master strip. A good consequence is that mono
tracks get their expected panner.
An akward side-effect is that MIDI-only routes (e.g. midi tracks or busses
without a synth) get two audio channels that have no use (and indeed no
panner is added because there is nothing to pan).
Skip the completion of audio outs if there was no audio out to begin
with.
