Recherche avancée

Sur d’autres sites (7775)

• Trim .wav file combining sox, ffmpeg in a windows batch script

  20 février 2020, par pbc278

  I need to trim 300 .wav files to be a specific sample-length (x times 1152 samples)
  So if a .wav file is 1727821 samples in length (sox --i -s xx.wav) = 1727821.

  I need to divide the outcome of the sox command by 1152 (1727821/1152 = 1499,844...)

  Then I need to use the outcome (1499 without decimals) and multiply it by 1152 (1499*1152=1726848)

  Then I need to trim the sample with ffmpeg to that outcome

  ffmpeg -i xx.wav -af atrim=end_sample=1726848

  How do I do this in a batch file ?

  in pseudo code something like

  FOR %%i IN (*.*) DO (
  
  
  
    @set sample_length = (sox --i -s %%~ni.wav)
  
    @set multiplier = %sample_length% / 1152
  
    @set new_samplelength = %multiplier% * 1152
  
    ffmpeg -i %%~ni.wav -af atrim=end_sample=%new_samplelength%
  
  
  
  )

  I can’t find anything understandable on how to combine all these different tools in one batch script

• avformat/matroskadec : Fix demuxing RealAudio 28.8

  20 avril 2020, par Andreas Rheinhardt

  avformat/matroskadec : Fix demuxing RealAudio 28.8
  RealAudio 28.8 does not need or use sub_packet_size for its demuxing
  
  and this field is therefore commonly set to zero. But since 18ca491b
  
  the Real Audio specific demuxing is no longer applied if sub_packet_size
  
  is zero because the codepath for cook and ATRAC3 divide by it ; this made
  
  these files undecodable.
  Furthermore, since 569d18aa (merged in 2c8d876d) sub_packet_size being
  
  zero is used as an indicator for invalid data, so that a file containing
  
  such a track was completely skipped.
  This commit fixes this by not checking sub_packet_size for RealAudio
  
  28.8 at all.
  Signed-off-by : Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
  

  • [DH] libavformat/matroskadec.c

• ffmpeg frame perfect video splitting

  5 mai 2020, par aosborne

  I'm working on a project that needs to calculate some computer vision stuff (openCV with Python) on every frame of the video. I'm trying to speed this up by splitting the video into chunks and giving each chunk to a different processor to handle.

  



  This works- except I can't seem to break up the video right. I've been looking at FFMPEG and have read about -ss and -t, -segment, even re-encoding the video to add more key frames.

  



  No matter what I've tried, I cannot accomplish my goal : cut a video into segments, where the exact placement of the segment divide doesn't matter, but if I were to line up the segments back together there would be no lost frames and no added frames (duplicated).

  



  If I can't figure this out, I expect I will use FFMPEG to loosely segment the video and I will find where the video overlaps in my program code by comparing frames.