Opened 12 years ago
Closed 12 years ago
#2605 closed defect (fixed)
Raw audio in MP4 container
| Reported by: | Alex Sukhanov | Owned by: | |
|---|---|---|---|
| Priority: | important | Component: | avformat |
| Version: | git-master | Keywords: | mov regression |
| Cc: | Blocked By: | ||
| Blocking: | Reproduced by developer: | yes | |
| Analyzed by developer: | no |
Description
Summary of the bug:
I found that commit http://ffmpeg.org/pipermail/ffmpeg-cvslog/2012-March/048842.html
breaks MOV demuxing in case if input file has PCM audio inside and wrong STSZ sample_size for this audio trak.
I have a file which I successfully transcoded with ffmpeg before this patch:
Audio trak is PCM signed 16 LE.
Timescale = 44100
num_channels = 2
bits_per_sample = 16
Each sample duration = 1, so Each sample size should be = num_channels * bits_per_sample / 8 * duration = 4 bytes
But STSZ wrongly has sample_size = 1 byte.
I uploaded file on ftp://upload.ffmpeg.org:
audio_silence_after_ffmpeg_upgrade.mov
Before this patch http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=50059bde77674977d9134f3c1151a63cb7a2391c
ffmpeg was able to fix this stream error and decode audio successfully.
How to reproduce:
% ffplay audio_silence_after_ffmpeg_upgrade.mov
ffplay version git-2013-05-15-1a36c75 Copyright (c) 2003-2013 the FFmpeg developers
built on May 14 2013 19:05:26 with gcc 4.6 (Ubuntu/Linaro 4.6.3-1ubuntu5)
configuration: --enable-gpl --enable-nonfree --enable-pthreads --enable-libx264 --enable-libfaac --enable-libmp3lame --extra-cflags=-I/usr/local/include --extra-ldflags=-L/usr/local/lib
libavutil 52. 31.100 / 52. 31.100
libavcodec 55. 9.100 / 55. 9.100
libavformat 55. 7.100 / 55. 7.100
libavdevice 55. 0.100 / 55. 0.100
libavfilter 3. 67.100 / 3. 67.100
libswscale 2. 3.100 / 2. 3.100
libswresample 0. 17.102 / 0. 17.102
libpostproc 52. 3.100 / 52. 3.100
Change History (5)
comment:1 by , 12 years ago
comment:2 by , 12 years ago
| Keywords: | mov regression added; MOV MP4 demuxer removed |
|---|---|
| Priority: | normal → important |
| Version: | unspecified → git-master |
comment:4 by , 12 years ago
| Reproduced by developer: | set |
|---|---|
| Status: | new → open |
I had a chance to do some investigation.
In file I provided to you Audio TRAK has wrong STSZ atom. Sample size must be = 4, but it 1.
Before Patch http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=50059bde77674977d9134f3c1151a63cb7a2391c ffmpeg could find this bug in the stream (as long as it's PCM, we can calculate sample size) and fix it.
Also please take a look on Audio STTS:
EntryCount 2
Entry[0].SampleCount 688128
Entry[0].SampleDelta 0
Entry[1].SampleCount 0
Entry[1].SampleDelta 0
As long as we do have 2 entries inside of audio STTS: condition inside of this 'if' statement = true:
http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavformat/mov.c;h=d5e1f893f573d794a8a81c4e20dd2ae026034809;hb=HEAD#l1967
so we use alt_sample_size = 1 (incorrect) instead of sample size = 4 (correct)