Recherche avancée

Sur d’autres sites (4977)

• FFmpeg black screen using mpeg2video

  16 juin 2020, par Konata

  I have a proprietary piece of hardware for broadcasting, that accepts extremely specific video format.
The software that we use for it is extremely slow and outdated.

  



  I was looking at adapting FFmpeg to convert .avi/.mp4/etc to the specific format which in our case is

  



  


  .mpg ( 720*576 16:9, 25FPS, MPEG Video PAL v.2 (Main@High) BVOP)
 acceptable bitrate is 9000k - 15000k.

  


  



  So I have constructed this FFmpeg command :

  



  -i input.avi -codec:a mp2 -b:a 384k -c:v mpeg2video -aspect 16:9 -bf 2 -b:v 10000k -maxrate 10000k -minrate 10000k -bufsize 5000k  -profile:v main -level:v 4.0 output.mpg

  



  The audio is working fine, but instead of video I get a black screen, this is the conversion log

  



  "E:\auto_video\ffmpeg\bin\ffmpeg.exe" -i input.avi -codec:a mp2 -b:a 384k -c:v mpeg2video -aspect 16:9 -bf 2 -b:v 10000k -maxrate 10000k -minrate 10000k -bufsize 5000k -r 25  -profile:v main -level:v 4.0 output.mpg
ffmpeg version git-2020-06-12-38737b3 Copyright (c) 2000-2020 the FFmpeg developers
  built with gcc 9.3.1 (GCC) 20200523
  configuration: --enable-gpl --enable-version3 --enable-sdl2 --enable-fontconfig --enable-gnutls --enable-iconv --enable-libass --enable-libdav1d --enable-libbluray --enable-libfreetype --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libsrt --enable-libtheora --enable-libtwolame --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib --enable-gmp --enable-libvidstab --enable-libvmaf --enable-libvorbis --enable-libvo-amrwbenc --enable-libmysofa --enable-libspeex --enable-libxvid --enable-libaom --disable-w32threads --enable-libmfx --enable-ffnvcodec --enable-cuda-llvm --enable-cuvid --enable-d3d11va --enable-nvenc --enable-nvdec --enable-dxva2 --enable-avisynth --enable-libopenmpt --enable-amf
  libavutil      56. 54.100 / 56. 54.100
  libavcodec     58. 92.100 / 58. 92.100
  libavformat    58. 46.101 / 58. 46.101
  libavdevice    58. 11.100 / 58. 11.100
  libavfilter     7. 86.100 /  7. 86.100
  libswscale      5.  8.100 /  5.  8.100
  libswresample   3.  8.100 /  3.  8.100
  libpostproc    55.  8.100 / 55.  8.100
Guessed Channel Layout for Input Stream #0.1 : stereo
Input #0, avi, from 'input.avi':
  Metadata:
    date            : 2020-06-04T14:11:03+06:00
  Duration: 00:00:10.00, start: 0.000000, bitrate: 30465 kb/s
    Stream #0:0: Video: dvvideo (dvsd / 0x64737664), yuv420p, 720x576 [SAR 16:15 DAR 4:3], 28915 kb/s, 25 fps, 25 tbr, 25 tbn, 25 tbc
    Stream #0:1: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 48000 Hz, stereo, s16, 1536 kb/s
File 'output.mpg' already exists. Overwrite? [y/N] y
Stream mapping:
  Stream #0:0 -> #0:0 (dvvideo (native) -> mpeg2video (native))
  Stream #0:1 -> #0:1 (pcm_s16le (native) -> mp2 (native))
Press [q] to stop, [?] for help
Output #0, mpeg, to 'output.mpg':
  Metadata:
    date            : 2020-06-04T14:11:03+06:00
    encoder         : Lavf58.46.101
    Stream #0:0: Video: mpeg2video (Main), yuv420p, 720x576 [SAR 64:45 DAR 16:9], q=2-31, 10000 kb/s, 25 fps, 90k tbn, 25 tbc
    Metadata:
      encoder         : Lavc58.92.100 mpeg2video
    Side data:
      cpb: bitrate max/min/avg: 10000000/10000000/10000000 buffer size: 5000000 vbv_delay: N/A
    Stream #0:1: Audio: mp2, 48000 Hz, stereo, s16, 384 kb/s
    Metadata:
      encoder         : Lavc58.92.100 mp2
frame=  250 fps=0.0 q=2.5 Lsize=   12574kB time=00:00:09.99 bitrate=10302.7kbits/s speed=16.8x
video:12054kB audio:469kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.402694%"E:\auto_video\ffmpeg\bin\ffmpeg.exe" -i input.avi -codec:a mp2 -b:a 384k -c:v mpeg2video -aspect 16:9 -bf 2 -b:v 10000k -maxrate 10000k -minrate 10000k -bufsize 5000k -r 25  -profile:v main -level:v 4.0 output.mpg
ffmpeg version git-2020-06-12-38737b3 Copyright (c) 2000-2020 the FFmpeg developers
  built with gcc 9.3.1 (GCC) 20200523
  configuration: --enable-gpl --enable-version3 --enable-sdl2 --enable-fontconfig --enable-gnutls --enable-iconv --enable-libass --enable-libdav1d --enable-libbluray --enable-libfreetype --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libsrt --enable-libtheora --enable-libtwolame --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib --enable-gmp --enable-libvidstab --enable-libvmaf --enable-libvorbis --enable-libvo-amrwbenc --enable-libmysofa --enable-libspeex --enable-libxvid --enable-libaom --disable-w32threads --enable-libmfx --enable-ffnvcodec --enable-cuda-llvm --enable-cuvid --enable-d3d11va --enable-nvenc --enable-nvdec --enable-dxva2 --enable-avisynth --enable-libopenmpt --enable-amf
  libavutil      56. 54.100 / 56. 54.100
  libavcodec     58. 92.100 / 58. 92.100
  libavformat    58. 46.101 / 58. 46.101
  libavdevice    58. 11.100 / 58. 11.100
  libavfilter     7. 86.100 /  7. 86.100
  libswscale      5.  8.100 /  5.  8.100
  libswresample   3.  8.100 /  3.  8.100
  libpostproc    55.  8.100 / 55.  8.100
Guessed Channel Layout for Input Stream #0.1 : stereo
Input #0, avi, from 'input.avi':
  Metadata:
    date            : 2020-06-04T14:11:03+06:00
  Duration: 00:00:10.00, start: 0.000000, bitrate: 30465 kb/s
    Stream #0:0: Video: dvvideo (dvsd / 0x64737664), yuv420p, 720x576 [SAR 16:15 DAR 4:3], 28915 kb/s, 25 fps, 25 tbr, 25 tbn, 25 tbc
    Stream #0:1: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 48000 Hz, stereo, s16, 1536 kb/s
File 'output.mpg' already exists. Overwrite? [y/N] y
Stream mapping:
  Stream #0:0 -> #0:0 (dvvideo (native) -> mpeg2video (native))
  Stream #0:1 -> #0:1 (pcm_s16le (native) -> mp2 (native))
Press [q] to stop, [?] for help
Output #0, mpeg, to 'output.mpg':
  Metadata:
    date            : 2020-06-04T14:11:03+06:00
    encoder         : Lavf58.46.101
    Stream #0:0: Video: mpeg2video (Main), yuv420p, 720x576 [SAR 64:45 DAR 16:9], q=2-31, 10000 kb/s, 25 fps, 90k tbn, 25 tbc
    Metadata:
      encoder         : Lavc58.92.100 mpeg2video
    Side data:
      cpb: bitrate max/min/avg: 10000000/10000000/10000000 buffer size: 5000000 vbv_delay: N/A
    Stream #0:1: Audio: mp2, 48000 Hz, stereo, s16, 384 kb/s
    Metadata:
      encoder         : Lavc58.92.100 mp2
frame=  250 fps=0.0 q=2.5 Lsize=   12574kB time=00:00:09.99 bitrate=10302.7kbits/s speed=16.8x
video:12054kB audio:469kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.402694%


  



  Additional info that might help, I get black screen in MPC-HC, Windows Media Player, and the actual hardware that I converted it for. But it plays fine in VLC

  


• doc/encoders : fix the misleading usage of profile

  20 juin 2020, par Limin Wang

  doc/encoders : fix the misleading usage of profile
  users are getting mislead by the integer, although profile
  
  can support both const string and integer.
  
  http://ffmpeg.org/pipermail/ffmpeg-user/2020-June/049025.html
  Also fix the order of high and main, it's not my intention.
  Signed-off-by : Limin Wang <lance.lmwang@gmail.com>
  

  • [DH] doc/encoders.texi

• Why are there vaapi filters that exist in libavfilter but do not exist in my compiled version of ffmpeg ?

  13 juin 2020, par John Allard

  I've compiled the most recent snapshot of ffmpeg with vaapi enabled

  



  $ ffmpeg -hwaccesls
ffmpeg version N-98129-g0b182ff Copyright (c) 2000-2020 the FFmpeg developers
  built with gcc 5.4.0 (Ubuntu 5.4.0-6ubuntu1~16.04.12) 20160609
  configuration: --disable-debug --disable-doc --disable-ffplay --enable-shared --enable-avresample --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-gpl --enable-libass --enable-fontconfig --enable-libfreetype --enable-libvidstab --enable-libmp3lame --enable-libopus --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx265 --enable-libxvid --enable-libx264 --enable-nonfree --enable-openssl --enable-libfdk_aac --enable-postproc --enable-small --enable-version3 --enable-libzmq --extra-libs=-ldl --prefix=/opt/ffmpeg --enable-libopenjpeg --enable-libkvazaar --enable-libaom --extra-libs=-lpthread --enable-libsrt --enable-vaapi --extra-cflags=-I/opt/ffmpeg/include --extra-ldflags=-L/opt/ffmpeg/lib
  libavutil      56. 54.100 / 56. 54.100
  libavcodec     58. 92.100 / 58. 92.100
  libavformat    58. 46.101 / 58. 46.101
  libavdevice    58. 11.100 / 58. 11.100
  libavfilter     7. 86.100 /  7. 86.100
  libavresample   4.  0.  0 /  4.  0.  0
  libswscale      5.  8.100 /  5.  8.100
  libswresample   3.  8.100 /  3.  8.100
  libpostproc    55.  8.100 / 55.  8.100
Hardware acceleration methods:
vaapi


  



  I know that vaapi is working because I can use it for hardware decoding and encoding of h264 videos. I can see some vaapi filters as well

  



  $ ffmpeg -filters | grep vaapi
 ... deinterlace_vaapi V->V       (null)
 ... denoise_vaapi     V->V       (null)
 ... procamp_vaapi     V->V       (null)
 ... scale_vaapi       V->V       (null)
 ... sharpness_vaapi   V->V       (null)


  



  However, I notice that this list is missing the filter that I'm specifically looking for, namely , transpose_vaapi. If you look in the libavfilter source code you'll see the following

  



  This shows the transpose_vaapi filter defined in the allfilters.c file
https://github.com/FFmpeg/FFmpeg/blob/master/libavfilter/allfilters.c#L414

  



  This shows the transpose_vaapi filter source code
https://github.com/FFmpeg/FFmpeg/blob/master/libavfilter/vf_transpose_vaapi.c

  



  If the filter is defined in source code, it's defined in allfilters.c, and I've compiled ffmpeg from this source with vaapi enabled, why can I not use this filter with ffmpeg ?

  



  $ ffmpeg -y -hide_banner -nostats -loglevel error \
    -hwaccel vaapi -hwaccel_device /dev/dri/renderD128 -hwaccel_output_format vaapi \
    -i ./test_video.mp4 \
    -vf 'format=nv12,transpose_vaapi=2' \
    -c:v h264_vaapi \
    /tmp/rotated_video.mp4
[AVFilterGraph @ 0xf14000] No such filter: 'transpose_vaapi'
Error reinitializing filters!
Failed to inject frame into filter network: Invalid argument


  



  EDIT -

  



  Looking through the source code for vf_transpose_vaapi.c I see the following logic

  



      if (!pipeline_caps.rotation_flags) {
        av_log(avctx, AV_LOG_ERROR, "VAAPI driver doesn't support transpose\n");
        return AVERROR(EINVAL);
    }


  



  that's inside of the transpose_vaapi_build_filter_params function which is part of the filter initialization process. I guess it's possible that, if that call failed, the filter would fail to be built and it would not be registered as a valid filter ? This seems like something that would happen at runtime when I attempt to run the filter rather than something that would happen at compile time when setting which filters are defined.