channel_layout option ignored for non default layouts

[pkv]

When selecting an output channel layout through -channel_output, if the channel layout is not a default one (ex: quad instead of 4.0, or octagonal instead of 7.1) , the option is ignored and the output is encoded with default layout.

This error does not occur if one uses -af aformat to specify the layout.

The defect does not depend on the muxer; I checked it occurs with mkv, wav, mp4. In the example below I use pcm as codec but I checked that the defect occurs also irrespective of the codec (ex: libopus, ...)

I have pinpointed the source of the defect to:

ffmpeg_opt.c, l.2526, function open_output_file (l.2091 of current head)

if (ost->enc_ctx->channels) {
                    f->channel_layout = av_get_default_channel_layout(ost->enc_ctx->channels);
                }

These lines were introduced in commit 198e8b8e774659eacaa7058c7f5704029af5bbbf ; however I don't think this is a regression because I could check the same errors occur on earlier builds.

The error is most manifest in the case of identical input and output channel layouts (ex: both quad) because an auto-insertion of an auto_resampler filtergraph occurs to convert input from quad to 4.0 while this should never happen (see below, last line of [auto_resampler] ).

Here's uncut command line (build is latest one from zerano winx64 but defect is also on git-master head):

E:\ffmpeg-20170921-183fd30-win64-static\bin> ./ffmpeg -loglevel debug -channel_layout quad -i .\short4.wav -vn -c:a pcm_s16le -channel_layout quad -ar 48k shorter.wav
ffmpeg version N-87353-g183fd30 Copyright (c) 2000-2017 the FFmpeg developers
  built with gcc 7.2.0 (GCC)
  configuration: --enable-gpl --enable-version3 --enable-cuda --enable-cuvid --enable-d3d11va --enable-dxva2 --enable-libmfx --enable-nvenc --enable-avisynth --enable-bzlib --enable-fontconfig --enable-frei0r --enable-gnutls --enable-iconv --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libfreetype --enable-libgme --enable-libgsm --enable-libilbc --enable-libmodplug --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenh264 --enable-libopenjpeg --enable-libopus --enable-librtmp --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib
  libavutil      55. 76.100 / 55. 76.100
  libavcodec     57.106.101 / 57.106.101
  libavformat    57. 82.101 / 57. 82.101
  libavdevice    57.  8.101 / 57.  8.101
  libavfilter     6.105.100 /  6.105.100
  libswscale      4.  7.103 /  4.  7.103
  libswresample   2.  8.100 /  2.  8.100
  libpostproc    54.  6.100 / 54.  6.100
Splitting the commandline.
Reading option '-loglevel' ... matched as option 'loglevel' (set logging level) with argument 'debug'.
Reading option '-channel_layout' ... matched as option 'channel_layout' (set channel layout) with argument 'quad'.
Reading option '-i' ... matched as input url with argument '.\short4.wav'.
Reading option '-vn' ... matched as option 'vn' (disable video) with argument '1'.
Reading option '-c:a' ... matched as option 'c' (codec name) with argument 'pcm_s16le'.
Reading option '-channel_layout' ... matched as option 'channel_layout' (set channel layout) with argument 'quad'.
Reading option '-ar' ... matched as option 'ar' (set audio sampling rate (in Hz)) with argument '48k'.
Reading option 'shorter.wav' ... matched as output url.
Finished splitting the commandline.
Parsing a group of options: global .
Applying option loglevel (set logging level) with argument debug.
Successfully parsed a group of options.
Parsing a group of options: input url .\short4.wav.
Applying option channel_layout (set channel layout) with argument quad.
Successfully parsed a group of options.
Opening an input file: .\short4.wav.
[NULL @ 0000000000706fe0] Opening '.\short4.wav' for reading
[file @ 00000000011c2ec0] Setting default whitelist 'file,crypto'
[wav @ 0000000000706fe0] Format wav probed with size=2048 and score=99
[wav @ 0000000000706fe0] Before avformat_find_stream_info() pos: 102 bytes read:65638 seeks:1 nb_streams:1
[wav @ 0000000000706fe0] parser not found for codec pcm_s24le, packets or times may be invalid.
    Last message repeated 1 times
[wav @ 0000000000706fe0] All info found
[wav @ 0000000000706fe0] After avformat_find_stream_info() pos: 204702 bytes read:295014 seeks:1 frames:50
Input #0, wav, from '.\short4.wav':
  Metadata:
    encoder         : Lavf57.82.101
  Duration: 00:00:05.00, bitrate: 4608 kb/s
    Stream #0:0, 50, 1/48000: Audio: pcm_s24le ([1][0][0][0] / 0x0001), 48000 Hz, quad, s32 (24 bit), 4608 kb/s
Successfully opened the file.
Parsing a group of options: output url shorter.wav.
Applying option vn (disable video) with argument 1.
Applying option c:a (codec name) with argument pcm_s16le.
Applying option channel_layout (set channel layout) with argument quad.
Applying option ar (set audio sampling rate (in Hz)) with argument 48k.
Successfully parsed a group of options.
Opening an output file: shorter.wav.
[file @ 0000000001212820] Setting default whitelist 'file,crypto'
Successfully opened the file.
Stream mapping:
  Stream #0:0 -> #0:0 (pcm_s24le (native) -> pcm_s16le (native))
Press [q] to stop, [?] for help
cur_dts is invalid (this is harmless if it occurs once at the start per stream)
detected 8 logical cores
[graph_0_in_0_0 @ 0000000001258280] Setting 'time_base' to value '1/48000'
[graph_0_in_0_0 @ 0000000001258280] Setting 'sample_rate' to value '48000'
[graph_0_in_0_0 @ 0000000001258280] Setting 'sample_fmt' to value 's32'
[graph_0_in_0_0 @ 0000000001258280] Setting 'channel_layout' to value '0x33'
[graph_0_in_0_0 @ 0000000001258280] tb:1/48000 samplefmt:s32 samplerate:48000 chlayout:0x33
[format_out_0_0 @ 000000000070b560] Setting 'sample_fmts' to value 's16'
[format_out_0_0 @ 000000000070b560] Setting 'sample_rates' to value '48000'
[format_out_0_0 @ 000000000070b560] Setting 'channel_layouts' to value '0x107'
[format_out_0_0 @ 000000000070b560] auto-inserting filter 'auto_resampler_0' between the filter 'Parsed_anull_0' and the filter 'format_out_0_0'
[AVFilterGraph @ 0000000001215b40] query_formats: 4 queried, 6 merged, 3 already done, 0 delayed
[auto_resampler_0 @ 000000000070bac0] [SWR @ 0000000001258fa0] Using fltp internally between filters
[auto_resampler_0 @ 000000000070bac0] [SWR @ 0000000001258fa0] Matrix coefficients:
[auto_resampler_0 @ 000000000070bac0] [SWR @ 0000000001258fa0] FL: FL:0.707107 FR:0.000000 BL:0.000000 BR:0.000000
[auto_resampler_0 @ 000000000070bac0] [SWR @ 0000000001258fa0] FR: FL:0.000000 FR:0.707107 BL:0.000000 BR:0.000000
[auto_resampler_0 @ 000000000070bac0] [SWR @ 0000000001258fa0] FC: FL:0.000000 FR:0.000000 BL:0.000000 BR:0.000000
[auto_resampler_0 @ 000000000070bac0] [SWR @ 0000000001258fa0] BC: FL:0.000000 FR:0.000000 BL:0.500000 BR:0.500000
[auto_resampler_0 @ 000000000070bac0] ch:4 chl:quad fmt:s32 r:48000Hz -> ch:4 chl:4.0 fmt:s16 r:48000Hz
Output #0, wav, to 'shorter.wav':
  Metadata:
    ISFT            : Lavf57.82.101
    Stream #0:0, 0, 1/48000: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 48000 Hz, quad, s16, 3072 kb/s
    Metadata:
      encoder         : Lavc57.106.101 pcm_s16le
[out_0_0 @ 000000000070b360] EOF on sink link out_0_0:default.
No more output streams to write to, finishing.
size=    1875kB time=00:00:05.00 bitrate=3072.2kbits/s speed= 277x
video:0kB audio:1875kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.005312%
Input file #0 (.\short4.wav):
  Input stream #0:0 (audio): 704 packets read (2880480 bytes); 704 frames decoded (240040 samples);
  Total: 704 packets (2880480 bytes) demuxed
Output file #0 (shorter.wav):
  Output stream #0:0 (audio): 704 frames encoded (240040 samples); 704 packets muxed (1920320 bytes);
  Total: 704 packets (1920320 bytes) muxed
704 frames successfully decoded, 0 decoding errors
[AVIOContext @ 00000000007084a0] Statistics: 4 seeks, 11 writeouts
[AVIOContext @ 0000000000707680] Statistics: 2946118 bytes read, 1 seeks

I have started investigating a patch; but I may need a few pointers on how to retrieve channel_layout , for instance from an OptionsContext (there is no channel_layout in the struct; presumably from OptionGroup g->codec_opts then ? not familiar enough with AvDictionary ).

[pkv]

patch submitted on ffmpeg-devel list

[pkv]

don't know if it is a regression, but I tested up to git-718607b (2012-06-08) and the bug is already there; earlier binaries from january 2012 did not have the channel_layout option. Auto-insert of a filter when a non default channel is selected through -channel_layout.

[Elon Musk]

Feel free to reopen if still happens in latest ffmpeg and similar option.