This commit fixes an issue that the last audio packet is sometimes not
written into mp4 format. Since libavformat internally calculates the
packet duration from the frame_size specified in the codec parameter, it
is necessary to set frame_size.
When releasing a darray mux_packets at error, contents of mux_packets
are not released. It causes memory leaks if error occurs during saving
the replay buffer to a file.
Per FFmpeg commit 337f777f378c [1], FFmpeg removed nvenc_h264_encoder
and nvenc_hevc_encoder after deprecation in FFmpeg commit 888a5c794778
[2]. The names to be used are ff_h264_nvenc_encoder and
ff_hevc_nvenc_encoder. So we must allow alternative search of codec as
h264_nvenc or nvenc_h264 in obs-ffmpeg.c.
[1]: 337f777f37
[2]: 888a5c7947
AVFormatContext::oformat was made const on April 27, 2021 [1]. If we
respect the constness of AVOutputFormat and do not cast results from
FFmpeg functions to non-const, we cannot modify the results after the
fact. Our choices are either to cast them to non-const (and presumably
have them implicitly casted back to const on later function calls), or
only try to modify the results in versions of FFmpeg where these are not
expected to be const.
Instead of relying on casts, we can set the encoder values in the
ffmpeg_cfg struct, which are later passed to new_stream.
Also modifies deps/media-playback. Removes compiler warnings.
Some avformat functions return const AV(In/Out)putFormat per [1], so
ifdef as needed.
[1]: 56450a0ee4
lavf 59.0.100 avformat.h
avformat: Constify the API wrt AV(In|Out)putFormat
Also constify AVProbeData.
avcodec.h stopped including channel_layout.h per FFmpeg commit
1be3d8a0cb77 [1]. avformat.h stopped including avcodec.h per FFmpeg
commit e67e02d15672 [2]. As a result, we need to explicitly include
avutil/channel_layout.h when needed. Fixes compilation error against
FFmpeg later than the two mentioned commits.
[1]: 1be3d8a0cb
[2]: e67e02d156
FFmpeg commit e67e02d15672 [1] drops avcodec.h from avformat.h includes.
Therefore we need to include it explicitly.
[1]: e67e02d156
lavf/avformat.h: drop the avcodec.h include
The commit 4a5ac3974 fixed the bug the first PTS is not zero. This
commit fixed the same issue in replay-buffer.
A variable `video_dts_offset` is renamed to `video_pts_offset` and the
variable is initialized by `pts` instead of `dts`.
This adds RIST to the list of protocols supported by the obs
ffmpeg_mpegts_muxer.
RIST is container and codec agnostic. But this implementation relies
on ffmpeg avformat integration and uses mpeg-ts container.
When video is disabled, no audio was sent to the container so that the
audio was not saved. Prior to this change, all audio was discarded until
the first video packet arrives. This change limits to discard audio only
if video is available.
Unfortunately these encoders tend to crash or freeze after after
multiple bitrate changes. Users with dynamic bitrate enabled
experience a random encoder freeze which manifests as OBS dropping to
0kb/sec and disconnecting and it's not at all obvious that dynamic
bitrate is the cause. Disable it for now until we can figure out if
there is any workaround or bug on our end.
Only use lossless encode if the capability is supported.
Set qpPrimeYZeroTransformBypassFlag to 1 for lossless.
Do not set profileGUID for lossless.
For both NVENC implementations, retry with a heavier reset because both
are unable to recover from failure lightly.
For new NVENC, warn if PVT is requested, but unsupported by the GPU.
Also retry without PVT on failure to try to catch bad cominbations.
For old NVENC, if PVT is enabled when FFmpeg failure occurs, retry
without PVT.
This was changed to default on but is not exposed for ffmpeg nvenc.
Where cards without temporal AQ support will fail to initialize and we
cannot do checks beforehand. This exposes the parameter to allow users
to disable this feature when using the ffmpeg implementation.
If DEBUG_FFMPEG_MUX is enabled, when starting a recording without
configuring global_stream_key, obs-ffmpeg-mux will hang. Apply same
check to global_stream_key as in init_params() to prevent this.
Previously the save callback would be called when the replay buffer
hotkey was pressed and not when the saving of the replay was finished.
When the 'get_last_replay' procedure was called after the saved callback
function, it would return the incorrect path, as the path would still
be the previous path.
