This uses boost::multiprecision::int512_t when multiplying and dividing by the numerator
and denominator of a ratio_t. 128 bits would be sufficient but for some reason, the boost
docs show the 512 bit variant being very slightly faster.
This is a better solution than using a double, which although it will prevent overflow
has fairly limited resolution.
Region fades would sometimes get in a mode with weird behaviour. They
would be drawn in 2d with crossing lines, mainly moving back and forth
horizontally - not as a function of time. It would sound as it looked.
The fade would sometimes jump around when resizing. It could be worked
around by resetting the fade shape. It turned out the problem could be
reproduced by making minute long fades.
This change fixes or works around the problem.
Back story:
timepos_t (in temporal/timeline.h) uses 62 bit for integer value and the
max value is thus 4611686018427387904 ~ 5e18. timepos_t counts
superclocks, where superclock_ticks_per_second is 56448000 ~ 6e7. It can
thus store up to 8e10 seconds - thousands of years.
ratio_t (in temporal/types.h) can represent fractions as 64 bit (signed)
numerator and denominator. timepos_t avoids floating point operations,
but has operator* with ratio_t. To avoid crazy loss of precision it will
multiply the superclock count with the numerator before dividing with
the denominator.
Audio region fade in and out uses a number of increasing timepos_t
values (in a ControlList) up to the length of fade. When dragging to
resize, these values are (in_x_scale) multiplied with the ratio_t of the
new and old fade length. The problem is that the 62 bits will overflow
if using fades more than sqrt(5e18) ~ 2e9 superclock ticks ~ 38 seconds.
It will overflow into the "beat" flag and (at 58 seconds) into the sign
bit. The timepos_t values in the fade will thus jump and can be negative
or change to count beats.
To work around that problem, this changeset just use floating point
values for scaling the timepos_t values. All scaled values are stored as
integer anyway, so it should not make any actual difference for this use
case. There might however be other uses of ControlList where it matters.
As an implementation detail of this "workaround" of using double, it
could perhaps also be nice to implement timepos_t operator* (or
operator*=) for double. But I'm not sure we want floating point support
in timepos_t.
An alternative (and better) solution would be to convince the fraction
multiplication to use 128 bits. It is essential to avoid overflow -
mainly in static analysis, alternatively as runtime checks or asserts.
This also allows to disable the resampler, effectively disabling
varispeed support, for the benefit of adding no additional latency.
By default 2 * 16 samples latency are added, due to port-resampler,
this is not desirable if Ardour is used as mixer only.
Session::route_processors_changed accumulates signals emitted
in realtime and processing is delegated to a dedicated rt-safe
thread. Previously this resulted in any changed to be converted
to a `GeneralChange`, which unconditionally triggered a route-reorder.
Record-arming a track causes a MeterPointChange (meters change to "in"),
and this caused routes to be resorted and a latency-update.
While the former is reasonable (Ardour prefers to process
rec-armed routes first), the latter certainly is not.
Even if we start at time T, if the tempo point before T is ramped, we need to
adjust it's omega value based on the beat time of the following tempo point. If
we don't do this, then using that tempo to compute ::superclock_at() for
subsequent points will be incorrect.
Note: there may be an ordering issue here with Tempo/Meter points. We might
need to pass over all tempo points first, then process meters etc. Something
like that.
libs/ardour/session_state.cc:1500:13: note: in instantiation of
function template specialization 'XMLNode::set_property<unsigned long>'
requested here
libs/pbd/pbd/string_convert.h:151:11: error:
type 'unsigned long' cannot be used prior to '::' because it
has no members
The master needs to be restored first, otherwise setting
the master value with be propagated. Furthermore the
actual value is saved at the time of VCA assignment.
Restoring the vaule needs to scale by the master-reduced value.
changes in tempo can be negative (decelerando/slowing down), and thus so can _omega. We cannot call log() or its
cousins on negative values, so in this scenario use an alternate expression for "t from b" in the tempo.pdf paper
We discovered in the past that the C++ API for GSource/Glib::Source has some fatal and unfixable flaws. Copy similar
code and just use the C API for GSource instead
This is based on code from earlier commits that were later reversed, but we need some mechanism
to ensure that threads have a thread local tempo map ptr set. The big difference is that this
time we do not implement this for all instances of an AbstractUI - implementation is left to
each thread/event loop
