The sample buffering logic does not take into account that the blocksize
could change. Reset the buffer if the channel configuration changes,
since if there are leftover samples, it is most likely a broken or
misconcatenated stream. This could lead to negative numbers for
missing_samples during decoding.
Thanks to Michael Niedermeyer for pointing these out.
Its the width in bytes and not the linesize or stride.
This makes the naming more consistent with the rest of the codebase
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This reverts commit 51dab60c7b91a21ec280c7d3042aa8350aae048f.
this fixed nothing, and was just the result of assuming that the
variable named linesize was the linesize. Its not, its the width
in bytes which was already correct.
Found-by: durandal_1707
Removing the /EXPERIMENTAL suffix for codec id field for Opus. Doing this
as the implementation of embedding Opus in Matroska is complete.
Signed-Off By: Vignesh Venkatasubramanian <vigneshv@google.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Next release of ffmpeg will support embedding Opus in Matroska
container. Adding that to the Changelog.
Signed-Off By: Vignesh Venkatasubramanian <vigneshv@google.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* cehoyos/master:
Set bits_per_raw_sample when reading j2k frames.
libopenjpegdec.c: Correctly scale gray16 output if precision < 16
Use rc_max_rate if no video bit_rate was specified when muxing mxf_d10.
Store the video bit_rate in the context when muxing mxf.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
matroskaenc: Fix stray pointers left over from av_reallocp_array refactoring
Conflicts:
libavformat/matroskaenc.c
Mostly not merged as the buggy code was no longer part of FFmpeg
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'e2cd766ca1b9bb2ba72aca5f6ba5d8be15743984':
configure: Mention that icl does not build both static and shared libs
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'b1e469885362febce3d9a4678624e44a92f77da9':
g2meet: Allocate cursor buffers large enough to fit the aligned width
See: 3c4c5ca1c194464f51de87534d15602f99bf9759
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '17d57848fc14e82f76a65ffb25c90f2f011dc4a0':
mpc8: Make sure the first stream exists before parsing the seek table
Conflicts:
libavformat/mpc8.c
See: 69fb605ad5e0f1384ca4d06d38ce0f1b6c8c286d
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '0d61f260010707f3028b818e8b24598e1a83d696':
zmbvdec: Check the buffer size for uncompressed data
Conflicts:
libavcodec/zmbv.c
See: 44c23aa1b85d195db6647d24e3b1d3de051790db
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '183b9d843a9533774fabd3984a52f3987001acbc':
ape: Don't allow the seektable to be omitted
See: e7cb161515fc9fb6d30d1681d64d9ba7ad737a4e
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'b26742cc308552f242ee2bf93b07a3ff509f4edc':
shorten: Break out of loop looking for fmt chunk if none is found
See: 5e9a56a0350c518cd4b38845aff49d41a9c952ae
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '49568851bf1700e3d9ea9cda29208d0df3c2c38b':
shorten: Use a checked bytestream reader for the wave header
Conflicts:
libavcodec/shorten.c
See: 5e9a56a0350c518cd4b38845aff49d41a9c952ae
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'f3d57dc69145f1b7acb4870da9ce60378190a1fd':
smacker: Free memory properly if the init function fails
Conflicts:
libavcodec/smacker.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '0679cec6e8802643bbe6d5f68ca1110a7d3171da':
smacker: Make sure we don't fill in huffman codes out of range
Conflicts:
libavcodec/smacker.c
See: 1285baaab550e3e761590ef6dfb1d9bd9d1332e4
the added test is redundant but will be used after future refactoring
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '48d6556dd46d4f4fac10d0f4a819e314887cd50e':
dsicin: Add some basic sanity checks for fields read from the file
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '5045f6bdfabfa5d3374756d2c23f762fb042acae':
configure: Fix wrong reference to user support mailing list
Conflicts:
configure
No change as the reference was not wrong in ffmpeg
Merged-by: Michael Niedermayer <michaelni@gmx.at>
