The approach is quite simple and doesn't take all use cases into account,
it only implements support when we are using the internal timecode we
create ourself.
Also the way we compute the sought frame count is naive, but it works
for simple cases.
We might have some old timecodes that are in the future now and have to
drop those to make sure that our queue is correctly ordered and we don't
have multiple timecodes for the same running time.
Directly read them out of the decoder as soon as we passed audio and
then store them in a queue that we handle internally together with their
timestamps. This cleans up memory management and gives us proper control
over the queue instead of guessing how the queue inside the LTC decoder
actually works and when it overflows.
And also introduce 6 instead of 2 frames of latency compared to the LTC
audio input as that seems to be an upper bound for how much the LTC
library is lagging behind.
If one of the inputs is live, add a latency of 2 frames to the video
stream and wait on the clock for that much time to pass to allow for the
LTC audio to be ahead.
In case of live LTC, don't do any waiting but only ensure that we don't
overflow the LTC queue.
Also in non-live LTC audio mode, flush too old items from the LTC queue
if the video is actually ahead instead of potentially waiting forever.
This could've happened if there was a bigger gap in the video stream.
This allows selecting whether we continue updating our last known
upstream timecode whenever a new one arrives or instead only keep the
last known one and from there on count up.
This uses the last known upstream timecode (counted up per frame), or
otherwise zero if none was known.
The normal last-known timestamp uses the internal timecode as fallback
if no upstream timecode was ever known.
We reject caps with other framerates as it's impossible to generate
timecodes unless we actually know a constant framerate. Reflect this
also in the pad template caps.
There's no point in working with invalid LTC timestamps as all future
calculations will be wrong based on this, and invalid LTC timestamps can
sometimes be read via the audio input.
Based on a patch by
Georg Lippitsch <glippitsch@toolsonair.com>
Vivia Nikolaidou <vivia@toolsonair.com>
Using libltc from https://github.com/x42/libltc
We now have a single property to select the timecode source that should
be applied, and for each timecode source the timecode is updated at
every frame. Then based on a set mode, the timecode is added to the
frame if none exists already or all existing timecodes are removed and
the timecode is added.
In addition the real-time clock is considered a proper timecode source
now instead of only allowing to initialize once in the beginning with
it, and also instead of just taking the current time we now take the
current time at the clock time of the video frame.
If recording is set to FALSE after the last audio or video buffer and
before the EOS event then recording stop is never signalled.
Similarly, we should signal recording stop once both audio and video are
EOS, regardless of the recording property, as there's nothing to be
recorded anymore.
This might be necessary temporarily for changing the previous settings.
Make it an actual error if the settings are like this while processing a
buffer.
If the first audio buffer to be dropped started right between two video
buffers (after the end of the first but before the start of the second,
as is often the case with N/1001 video frame rates), we would miss
sending the dropping=true message.
https://bugzilla.gnome.org/show_bug.cgi?id=797248
Previously it was dispatched before the last video buffer, and audio
buffers would follow afterwards. It's misleading to send the
dropping=true message before both streams have really stopped, it can
lead to races when someone is e.g. waiting for that message to send EOS.
Also added some debug output.
https://bugzilla.gnome.org/show_bug.cgi?id=797145
The case is properly handled a few lines below by dropping the buffer.
We shouldn't perpetually block the audio chain function until the
target-timecode is reached.
https://bugzilla.gnome.org/show_bug.cgi?id=796906
It works like a valve in front of the actual avwait. When recording ==
TRUE, other rules are then examined. When recording == FALSE, nothing is
passing through.
https://bugzilla.gnome.org/show_bug.cgi?id=796836
"avwait-status" is posted when avwait starts or stops passing through
data (e.g. because target-timecode and end-timecode respectively have
been reached). The attached structure includes a "dropping" boolean (set
to TRUE if we are currently dropping data, FALSE otherwise), and a
"running-time" GST_CLOCK_TIME which contains the running time of the
change.
https://bugzilla.gnome.org/show_bug.cgi?id=790170
