Recherche avancée

Sur d’autres sites (4688)

• Compiling an entire CMAF (.cmfv and .cmfa) stream into mp4

  17 janvier 2024, par Logan Price

  I am trying to create a general-purpose media downloader in Python which can download online media streams and compile them into one video file (ideally .mp4).

  


  I've encountered media streams using the CMAF format in which a server breaks up complete video files into pieces (video files “.cmfv” and audio files “.cmfa”) and then streams them to the client as they view the content.

  


  I can download all of the individual files, but I am having trouble putting them back together into one video. That is the problem I’m trying to solve

  


  I've tried looping through each of the video files (.cmfv) and writing them all into a new file. After combining, I tried to use FFMPEG to convert the combined .cmfv to mp4. I get an ffmpeg error that the combined .cmfv file cannot be read.

  


  # python

# create empty cmfv file
oldcmfv = open(“somepath”, “w”)

# iterate through individual cmfv files
for file in folder:
    with open(file) as portion:
        # write the cmfv portion to the combined cmfv file
        oldcmfv.write(portion)

oldcmfv.close()

# mp4 path
newmp4 = “somepath”

# attempt to convert cmfv combined file to mp4
# note that I did not try to include the cmfa (audio) files
ffmpeg.output(newmp4).input(oldcmfv).run()



  


  As a side note, it seems that there is very little discussion/information about the CMAF format. There is a technical document published by Apple and one other informational article that I saw but it seems like most resources are explaining how to encode videos into CMAF, but not how to decode them. I found an amazing GitHub repo that downloads CMAF stream media files pretty much automatically, but piecing them together has been a mystery so far.

  


• avisynth AverageLuma() function equivalent in ffmpeg libraries ?

  3 novembre 2013, par KG6ZVP

  I am working on implementing some software to analyze videos and would like to transition the project from avisynth to libavformat/libavcodec.

  The Problem : I would like to seek through every frame in a given input video, detect black frames and write that list to a file. Is there a function which would allow me to get the light level of the current frame ? I realize that I may have to implement such a function myself, but as of now, I don't even know where I could collect the information on individual pixels in each frame to start that analysis. Help is greatly appreciated !

• ffmpeg concat with different codecs in python

  14 avril 2020, par Hans Lutz van Tastique

  i'm working on a videopipeline project and try to add some stuff for it.
Atm it's running fine until the code reach this point :

  



  os.system('ffmpeg -i ' + self._rendition + ' -i ' + self._outChunk + ' -filter_complex "[0:0][1:0]concat=n=2:v=1:a=0[outv]" ' + self._tempRenditon)


  



  There is no error message or anything else. It is used to link individual chunks with different codec parameters. After this line it's just telling me that it cant find self._tempRendition.

  



  Did i made an mistake with the os.system call ?

  



  thx and greetz