Recherche avancée

Sur d’autres sites (9074)

• FFMPEG Re-muxing .MKV to MP4\MV4 while retaining Dolby Digital Plus [closed]

  30 décembre 2020, par David F

  I would be grateful if anyone could help me out here. I'm remuxing a large number of video files from .MKV to .MP4\M4V so they play on the Apple platform (including the TV App) without any issues. It has been a bit of a learning curve, but I’m so close now. It's important NOT to convert, but to remux.

  


  I have achieved the ability to remux 4K HDR with Dolby Digital Plus in a .MKV to 4K HDR with Dolby Digtial 5.1. MP4. The issue is retaining the Dolby Digital Plus. This contains the ATMOS information for object based sound.

  


  The media players on Apple reports Dolby Digital Plus for the .MKV and Dolby Digital 5.1 for .MP4

  


  I'm using FFMPEG with the following command :-

  


  ffmpeg -i original.mkv -c:v copy -c:a eac3 -tag:v hvc1 output.mp4

  


  Output from MediaInfo (note .MKV contains only 1 video and 1 audio track)

  


  Audio -----Original.mkv

  


  

  • ID : 2

  


  • Format : E-AC-3 JOC

  


  • Format/Info : Enhanced AC-3 with Joint
Object Coding

  


  • Commercial name : Dolby Digital Plus with Dolby
Atmos

  


  • Codec ID : **A_**EAC3

  


  • Duration : 51 min 51 s

  


  • Bit rate mode : Constant

  


  


  Audio ----- Remuxed .mp4

  


  

  • ID : 2

  


  • Format : E-AC-3

  


  • Format/Info : Enhanced AC-3

  


  • Commercial name : Dolby Digital Plus

  


  • Codec ID : ec-3

  


  • Duration : 51 min 51 s

  


  • Bit rate mode : Constant

  


  


  The remux operation finished in about 1 minute for a 8Gb 4k video file, so we I know there is no reencoding. My theory is the stream is still a E-AC-3 with Joint Object Control, but its marked as a E-AC-3.

  


  I can see that .mp4 can take E-AC-3 Joint Object Control (JOC) with the demo files from Dolby labs.

  


  I've also remuxed an MKV to MKV and able to see the same result. The JOC markings have been removed with FFMPEG

  


  My ask is, does anyone have any knowledge of how to remux and keep Enhanced AC-3 with Joint Object Control with FFMPEG. Does FFMPEG support this ?

  


  Or know if just editing the Format, Format/Info, and Code ID is enough and the sound information in my .MP4\M4V is E-AC-3 JOC. If so can I just edit the media info, if so how ?

  


  I know Dolby Labs to allow free remuxing of Dolby Digital Plus/ Atmos with sample code and information they have uploaded onto Github for developers. I just wonder if FFMPEG supports it

  


  Thanks in advanced.

  


  David

  


• Script for video to j2k conversion not working

  8 février 2015, par Subhajit

  I am trying to convert a 2k prores file to j2c image sequence so can I wrap them in mxf file. I am trying to bypass intermediate dpx or tiff conversion.
  But with my limited programming knowledge I am not getting success. Can u plz guide me what is wrong with my code.

  import subprocess as sp
  
  
  
  import subprocess as sp
  
  
  
  import sys
  
  
  
  import numpy
  
  
  
  
  
  FFMPEG = "ffmpeg"
  
  
  
  IMAGEMAGICK = "convert"
  
  
  
  COLOUR_MATRIX = "0.4124564,0.3575761,0.1804375,0.2126729,0.7151522,0.0721750,0.0193339,0.1191920,0.9503041"
  
  
  
  infile = sys.argv[1]
  
  fps = sys.argv[2]
  
  outfile = ('0'.zfill(8))
  
  
  
  
  
  to_dpx = [FFMPEG,'-i',infile,'-f','image2pipe','-pix_fmt','rgb24','-c:v','dpx','-r',fps,'-']
  
  to_j2k = [IMAGEMAGICK,'-','-alpha','off','-depth','12','-gamma','0.454545','-recolor',COLOUR_MATRIX,'-evaluate','multiply','0.9166',str(outfile)+'.j2c']
  
  
  
  pipe = sp.Popen(to_dpx,stdout = sp.PIPE, bufsize = 10**8)
  
  
  
  
  
  while(pipe.pull() is NONE):
  
  raw_image = pipe.stdout.read(2048*872*3)
  
  image = numpy.fromstring(raw_image,dtype = 'uint8')
  
  image = image.reshape((2048,872,3))
  
  pipe = sp.Popen(to_j2k,stdin = sp.PIPE,stderr = sp.pipe)
  
  outfile = int(outfile) + 1

  ERROR REPORTS :
  Reported by python :

  $python AurOchs_dc_compiler.py Sample2Kprores.mov 24
  
  Traceback (most recent call last):
  
  File "AurOchs_dc_compiler.py", line 17, in <module>
  
  pipe = sp.Popen(to_dpx,stdout = sp.PIPE, bufsize = 100**8)
  
  File "/usr/lib/python2.7/subprocess.py", line 739, in __init__
  
  self.stdout = os.fdopen(c2pread, 'rb', bufsize)
  
  OverflowError: signed integer is greater than maximum
  
  </module>

  Error messages by ffmpeg :

  Stream mapping:
  
  Stream #0:0 -> #0:0 (prores (native) -> dpx (native))
  
  Press [q] to stop, [?] for help
  
  av_interleaved_write_frame(): Broken pipe
  
  frame= 1 fps=0.0 q=-1.0 Lsize= 5234kB time=00:00:00.04 bitrate=1028964.3kbits/s

• Getting (Estimated) Time Offset of the Errors Reported by ffmpeg

  22 mai 2024, par Haru

  I'm using ffmpeg to check for errors in video files, but while it reports tons of issues it doesn't tell me where exactly the error happens in the file. It just dumps some crazy long hex numbers (0x7fa…) which doesn't tell me anything. Is there a way to make it report some timestamps, or at least estimated timestamps, or byte-offsets into the stream ? I tried the -debug_ts flag, but nothing changed in the output.

  



  $ ffmpeg -debug_ts -v error -i a_broken_file.mp4 -f null -
[h264 @ 0x7fa513023400] Reference 4 >= 4
[h264 @ 0x7fa513023400] error while decoding MB 15 34, bytestream 1304
[h264 @ 0x7fa513023400] Frame num change from 36 to 72
[h264 @ 0x7fa513023400] decode_slice_header error
[h264 @ 0x7fa513023400] Invalid NAL unit size (-1989965076 > 2668).
[h264 @ 0x7fa513023400] Error splitting the input into NAL units.
[aac @ 0x7fa513007200] Prediction is not allowed in AAC-LC.
Error while decoding stream #0:1: Invalid data found when processing input
[aac @ 0x7fa513007200] channel element 3.8 is not allocated
Error while decoding stream #0:1: Invalid data found when processing input
[aac @ 0x7fa513007200] channel element 3.11 is not allocated
Error while decoding stream #0:1: Invalid data found when processing input
[aac @ 0x7fa513007200] channel element 2.10 is not allocated
Error while decoding stream #0:1: Invalid data found when processing input
[aac @ 0x7fa513007200] Inconsistent channel configuration.
[aac @ 0x7fa513007200] get_buffer() failed
Error while decoding stream #0:1: Invalid argument
[aac @ 0x7fa513007200] Inconsistent channel configuration.
[aac @ 0x7fa513007200] get_buffer() failed
Error while decoding stream #0:1: Invalid argument
[aac @ 0x7fa513007200] channel element 3.1 is not allocated
Error while decoding stream #0:1: Invalid data found when processing input
[aac @ 0x7fa513007200] Reserved bit set.
…


  



  Or perhaps I could extract this information from the hex numbers there (I haven't found a single information about what this number stands for) ?