[FFmpeg-cvslog] lavc/videotoolbox: deprecate write-only output_callback

ffmpeg | branch: master | Anton Khirnov | Thu Sep 15 15:53:12 2022 +0200| [8576c3c5d82188c1313f20666b8760fe4a29444c] | committer: Anton Khirnov lavc/videotoolbox: deprecate write-only output_callback This field has never been used for anything, so stop setting it and deprecate it. > http://gi

[FFmpeg-cvslog] lavc/videotoolbox: do not pass AVCodecContext to decoder output callback

ffmpeg | branch: master | Anton Khirnov | Thu Sep 15 14:53:36 2022 +0200| [d7f4ad88a0df3c1339e142957bf2c40cd056b8ce] | committer: Anton Khirnov lavc/videotoolbox: do not pass AVCodecContext to decoder output callback The opaque parameter for the callback is set in videotoolbox_start(), called

[FFmpeg-cvslog] avcodec/avcodec: Use the new API fields to validate the layout returned by decoders

ffmpeg | branch: master | James Almer | Fri Sep 16 11:49:17 2022 -0300| [33cdf51a065c568bfe3811846095f8d8ac73fe6d] | committer: James Almer avcodec/avcodec: Use the new API fields to validate the layout returned by decoders This block was scheduled for removal, which means that no validation w

[FFmpeg-cvslog] fate/audio: Add tests for APTX (HD)

ffmpeg | branch: master | Andreas Rheinhardt | Sun Aug 29 18:06:25 2021 +0200| [9884d14711e7a36743861afedd949b5f32ac5e3c] | committer: Andreas Rheinhardt fate/audio: Add tests for APTX (HD) We have de- and encoders for APTX and APTX HD, yet not FATE tests. This commit therefore adds a transcod

[FFmpeg-cvslog] avcodec/encode: Redo checks for small last audio frame

ffmpeg | branch: master | Andreas Rheinhardt | Thu Sep 15 16:57:18 2022 +0200| [dc7b6645575aa431d1d49744a17f60ee415ebc8b] | committer: Andreas Rheinhardt avcodec/encode: Redo checks for small last audio frame In particular, check that there is only one small last frame in case the encoder has

[FFmpeg-cvslog] avcodec/utils: Support APTX (HD) in av_get_audio_frame_duration()

ffmpeg | branch: master | Andreas Rheinhardt | Sun Aug 29 18:29:34 2021 +0200| [6f22d1965eeab80544ecbb190bc4c379063fa481] | committer: Andreas Rheinhardt avcodec/utils: Support APTX (HD) in av_get_audio_frame_duration() APTX decodes four bytes of input to four stereo samples; APTX HD does the

[FFmpeg-cvslog] avcodec/aptxdec: Process data in complete blocks only

ffmpeg | branch: master | Andreas Rheinhardt | Sun Aug 29 22:18:52 2021 +0200| [e6bfb14223e44b4110c1d171d1ecffe9c80302a8] | committer: Andreas Rheinhardt avcodec/aptxdec: Process data in complete blocks only The APTX (HD) decoder decodes blocks of four (six) bytes to four output samples. It ma

[FFmpeg-cvslog] avcodec/aptxenc: Process data in complete blocks of four samples only

ffmpeg | branch: master | Andreas Rheinhardt | Sun Aug 29 22:44:36 2021 +0200| [129cbbd7be16fc0c396cfbb291794174735dbaf7] | committer: Andreas Rheinhardt avcodec/aptxenc: Process data in complete blocks of four samples only Do this by setting AVCodecInternal.pad_samples. This prevents reading

[FFmpeg-cvslog] avformat/aptxdec: Don't set AVCodecParameters.frame_size

ffmpeg | branch: master | Andreas Rheinhardt | Sun Aug 29 21:55:30 2021 +0200| [9d10d3a4ee7ccf3742c4d9dd26a781aaf2ab8ae9] | committer: Andreas Rheinhardt avformat/aptxdec: Don't set AVCodecParameters.frame_size This field was misunderstood: It gives the number of samples in a packet, not the n

[FFmpeg-cvslog] avformat/aptxdec: Don't set AV_PKT_FLAG_CORRUPT mistakenly

ffmpeg | branch: master | Andreas Rheinhardt | Sun Aug 29 22:01:16 2021 +0200| [c8707c105fc7cf3cd2ca59a5f2e645d129369d98] | committer: Andreas Rheinhardt avformat/aptxdec: Don't set AV_PKT_FLAG_CORRUPT mistakenly Just because we try to put multiple units of block_align bytes (the atomic units

[FFmpeg-cvslog] avcodec/aptx: Use AVCodecContext.frame_size according to the API

ffmpeg | branch: master | Andreas Rheinhardt | Sun Aug 29 20:25:18 2021 +0200| [c0d483cecb3ca31574f850c031d66a34a4f2d140] | committer: Andreas Rheinhardt avcodec/aptx: Use AVCodecContext.frame_size according to the API Currently the APTX (HD) codecs set frame_size if unset and check whether it

[FFmpeg-cvslog] avcodec/encode: Enable encoders to control padding of last frame

ffmpeg | branch: master | Andreas Rheinhardt | Thu Sep 15 19:20:11 2022 +0200| [017d976629ea272965ea67201543c8b5538cbf09] | committer: Andreas Rheinhardt avcodec/encode: Enable encoders to control padding of last frame Some audio codecs work with atomic units that decode to a fixed number of a

[FFmpeg-cvslog] lavc/vorbisdec: use ptrdiff_t to iterate over intptr_t

ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 19 19:10:42 2022 +0300| [bbf045aa592f7bec895907c8878fd734a0e8713b] | committer: James Almer lavc/vorbisdec: use ptrdiff_t to iterate over intptr_t While this probably never overflows, we are better safe than sorry. The callback prototype

[FFmpeg-cvslog] lavc/vorbisdsp: use ptrdiff_t rather than intptr_t

ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 19 19:10:43 2022 +0300| [b52034270a82ffc4aa584945c8a18aa8e58e741b] | committer: James Almer lavc/vorbisdsp: use ptrdiff_t rather than intptr_t ... for a difference between pointers. > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=comm

[FFmpeg-cvslog] lavc/vorbisdec: use intermediate variables

ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 19 19:10:44 2022 +0300| [5c15cb138eb1292bc16e64d3cb39953f1c09ce97] | committer: James Almer lavc/vorbisdec: use intermediate variables The compiler cannot infer that the two float vectors do not alias, causing unnecessary extra loads and

[FFmpeg-cvslog] avcodec/vorbisdec: Move ff_vorbis_inverse_coupling() to vorbisdsp.c

ffmpeg | branch: master | Andreas Rheinhardt | Mon Sep 19 19:13:28 2022 +0200| [32129d64953da1502a200e850ac0a559791824b9] | committer: Andreas Rheinhardt avcodec/vorbisdec: Move ff_vorbis_inverse_coupling() to vorbisdsp.c Only used there. Also make it static. Reviewed-by: James Almer Signed-

[FFmpeg-cvslog] avutil/dict: Error out in case of key == NULL

ffmpeg | branch: master | Andreas Rheinhardt | Wed Sep 14 21:03:15 2022 +0200| [187cd278325ae2fc6d533141e7d6c557b669932a] | committer: Andreas Rheinhardt avutil/dict: Error out in case of key == NULL Up until now, using NULL as key in av_dict_get() on a non-empty AVDictionary would crash; usin

[FFmpeg-cvslog] swscale/input: Avoid calls to av_pix_fmt_desc_get()

ffmpeg | branch: master | Andreas Rheinhardt | Thu Sep 8 03:45:12 2022 +0200| [4d7a1a4619c9ed89c5f6dbf8a0ae85b9a8aa056b] | committer: Andreas Rheinhardt swscale/input: Avoid calls to av_pix_fmt_desc_get() Up until now, libswscale/input.c used a macro to read an input pixel which involved a ca

[FFmpeg-cvslog] swscale/output: Don't call av_pix_fmt_desc_get() in a loop

ffmpeg | branch: master | Andreas Rheinhardt | Thu Sep 8 05:10:16 2022 +0200| [888a02a126522ad1da802eb1f4e4571afc09a14d] | committer: Andreas Rheinhardt swscale/output: Don't call av_pix_fmt_desc_get() in a loop Up until now, libswscale/output.c used a macro to write an output pixel which inv

[FFmpeg-cvslog] avcodec/mjpegdec: Avoid copying data when flipping image

ffmpeg | branch: master | Andreas Rheinhardt | Thu Apr 14 16:43:53 2022 +0200| [118b36f418fb050aebecd0c868b04bdc45293012] | committer: Andreas Rheinhardt avcodec/mjpegdec: Avoid copying data when flipping image Basically reverts af15c17daa5d8d2940c0263ba4d3ecec761c11ee. Flipping a picture by m

[FFmpeg-cvslog] tests/checkasm: add a test for VorbisDSPContext

ffmpeg | branch: master | James Almer | Mon Sep 19 13:56:34 2022 -0300| [8f119b501eaa1d813ea57b2dfc545d1e114f19a1] | committer: James Almer tests/checkasm: add a test for VorbisDSPContext Signed-off-by: James Almer > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=8f119b501eaa1d813e