#252: matroska/rawvideo decoding gives invalid pixel format string -1 @ buffer ---------------------------------------+--------------------- Reporter: kenney | Type: defect Status: new | Priority: normal Component: avformat | Version: git Keywords: matroska rawvideo pix_fmt | Blocked By: Blocking: | Reproduced: 0 Analyzed: 0 | ---------------------------------------+--------------------- The error: {{{ [buffer @ 00000000003F8E80] Invalid pixel format string '-1' Error opening filters! }}} I am converting an mkv file with a V_UNCOMPRESSED or -vcodec rawvideo video stream to any other stream, let's say .avi. First create the .mkv with rawvideo using any (short) valid video file, such as an avi: $ ffmpeg -i goodfile.avi -vcodec rawvideo test.mkv [See output 1 below] Next, I convert the mkv, which doesn't play ofcourse, back to avi: $ ffmpeg -i test.mkv test.avi [See output 2 below] The first step is just for you guys to be able to test this. I need to be able to pipe uncompressed frames with a timecode to ffmpeg.exe commandline, and mkv seems one of the only formats that allows this, as the rest either doesn't have VFR or uncompressed frames. And no, I cannot use the libs. Here is '''output 1''', no problems here: {{{ ffmpeg version git-N-30155-g01a73d6, Copyright (c) 2000-2011 the FFmpeg developers built on May 23 2011 03:47:41 with gcc 4.5.3 configuration: --enable-gpl --enable-version3 --enable-memalign-hack --enable-runtime-cpudetect --enable-avisynth --enable-bzlib --enable- frei0r --enable-libopencore-amrnb --enable-libopencore-amrwb --enable- libfreetype --enable-libgsm --enable-libmp3lame --enable-libopenjpeg --enable-librtmp --enable-libschroedinger --enable-libspeex --enable- libtheora --enable-libvorbis --enable-libvpx --enable-libx264 --enable- libxavs --enable-libxvid --enable-zlib --pkg-config=pkg-config libavutil 51. 2. 1 / 51. 2. 1 libavcodec 53. 6. 0 / 53. 6. 0 libavformat 53. 2. 0 / 53. 2. 0 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 10. 0 / 2. 10. 0 libswscale 0. 14. 0 / 0. 14. 0 libpostproc 51. 2. 0 / 51. 2. 0 [avi @ 000000000033A0C0] non-interleaved AVI Input #0, avi, from 'goodfile.avi': Duration: 00:00:02.54, start: 0.000000, bitrate: 298619 kb/s Stream #0.0: Video: rawvideo, bgr24, 960x540, 24 tbr, 24 tbn, 24 tbc [buffer @ 000000000033CD60] w:960 h:540 pixfmt:bgr24 tb:1/1000000 sar:0/1 sws_param: Output #0, matroska, to 'test.mkv': Metadata: encoder : Lavf53.2.0 Stream #0.0: Video: rawvideo, bgr24, 960x540, q=2-31, 200 kb/s, 1k tbn, 25 tbc Stream mapping: Stream #0.0 -> #0.0 Press [q] to stop, [?] for help frame= 60 fps= 0 q=0.0 size= 91112kB time=2.40 bitrate=310996.0kbits/s frame= 61 fps= 0 q=0.0 Lsize= 92647kB time=2.44 bitrate=311049.8kbits/s video:92644kB audio:0kB global headers:0kB muxing overhead 0.003153% }}} And here is '''output 2''', the problem: {{{ ffmpeg version git-N-30155-g01a73d6, Copyright (c) 2000-2011 the FFmpeg developers built on May 23 2011 03:47:41 with gcc 4.5.3 configuration: --enable-gpl --enable-version3 --enable-memalign-hack --enable-runtime-cpudetect --enable-avisynth --enable-bzlib --enable- frei0r --enable-libopencore-amrnb --enable-libopencore-amrwb --enable- libfreetype --enable-libgsm --enable-libmp3lame --enable-libopenjpeg --enable-librtmp --enable-libschroedinger --enable-libspeex --enable- libtheora --enable-libvorbis --enable-libvpx --enable-libx264 --enable- libxavs --enable-libxvid --enable-zlib --pkg-config=pkg-config libavutil 51. 2. 1 / 51. 2. 1 libavcodec 53. 6. 0 / 53. 6. 0 libavformat 53. 2. 0 / 53. 2. 0 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 10. 0 / 2. 10. 0 libswscale 0. 14. 0 / 0. 14. 0 libpostproc 51. 2. 0 / 51. 2. 0 [matroska,webm @ 00000000003F9F20] Estimating duration from bitrate, this may be inaccurate Input #0, matroska,webm, from 'test.mkv': Metadata: ENCODER : Lavf53.2.0 Duration: 00:00:02.44, start: 0.000000, bitrate: N/A Stream #0.0: Video: rawvideo, 960x540, PAR 1:1 DAR 16:9, 25 fps, 25 tbr, 1k tbn, 25 tbc (default) [buffer @ 00000000003F8E80] Invalid pixel format string '-1' Error opening filters! }}} Now, I realize that there is nowhere in the file stored what the pixel format is. Even when I specify -pix_fmt rgb24 I get this error. Also, according to the specs ( http://www.matroska.org/technical/specs/index.html ) there is a ColourSpace field that might be used but is not written. Btw if there is a workaround to this by specifying internal filters like -vf buffer or something in the mean time (I tried..), I'd appreciate it;) -- Ticket URL: <https://ffmpeg.org/trac/ffmpeg/ticket/252> FFmpeg <http://ffmpeg.org> FFmpeg issue tracker