Repository navigation
Automated release: version and notes generated from pull requests merged since 4.6.5.
Added
audio.py --track FILE: several audio tracks in one file, each with a name and a language (#420).
The input's own sound (--audio-stream) and the first audio stream of each--trackbecome one audio
stream apiece after the picture (stream-copied; an audio input writes none), in.mov,.mkvor.mp4.
Each is written as it is: its own channel layout, rate and length (nothing mixed, resampled, padded or
trimmed), as 24-bit PCM in.movand.mkvand AAC in.mp4.--track-name NAME(also--track-title)
and--track-lang CODE(also--track-language) are repeated in track order, the input's sound first;
a language is an ISO 639-2 code of three lowercase letters. The result'strackslists
{index, file, title, language, channels, duration}read back from the file, atrack_tagsstep
checks the names and languages against it (.movkeeps only the languages QuickTime has a code for,
and says so by failing the step), and a track longer than the picture is kept whole with a
track_longer_than_picturewarning. Refused, witherror.reason: an output that is not
.mov,.mkvor.mp4(track_container; a WAV holds one stream, and FFmpeg's MXF muxer stores no
names or languages and cuts every track to the shortest), a flag that changes the sound
(track_rebuilds_sound), more titles or languages than tracks (track_labels) and a language that is
not three letters (track_language). A BWF or multitrack WAV with a start timecode is not written.- feat: audio.py --track adds named, language-tagged audio tracks (#420) (#714)
What's Changed
Full Changelog: v4.6.5...v4.7.0