Recherche avancée

Sur d’autres sites (13116)

• How to Convert u-law 8000 Hz Audio Files to OPUS in Python

  21 juillet, par user60108

  I'm working on a Python project where I need to convert audio files from μ-law format with an 8000 Hz sampling rate to OPUS format. The challenge is to keep the original channels separated in the conversion process. Despite several attempts with different methods and libraries, I haven't succeeded yet. I'm looking for advice or a Python library that can efficiently handle this conversion.

  


  What I've tried :

  


  1- Using pydub to Load and Export Files : Attempted to load .au files with pydub and export them as OPUS. This approach resulted in ffmpeg related errors, specifically "Invalid data found when processing input".

  


  from pydub import AudioSegment
audio_1 = AudioSegment.from_file("path/to/file1.au", format="au", frame_rate=8000)
audio_2 = AudioSegment.from_file("path/to/file2.au", format="au", frame_rate=8000)
combined_audio = AudioSegment.from_mono_audiosegments(audio_1, audio_2)
combined_audio.export("output.opus", format="opus", bitrate="32k")


  


  2- Direct Conversion Using ffmpeg : Also tried direct conversion with ffmpeg from both the command line and within Python, specifying codecs, but kept encountering "Invalid data found when processing input".

  


  ffmpeg -acodec pcm_mulaw -i input_file.au -ar 8000 -ac 1 output.wav


  


  3- Playback with aplay : Interestingly, the files play back fine using aplay with specified sampling rate and format, confirming the files are not corrupted.

  


  aplay -r 8000 -f MU_LAW input_file.au


  


  How can I convert audio files from u-law 8000 Hz format to OPUS format in Python, ideally keeping the original audio channels separate ? Is there a specific Python library or tool that works well for this type of audio conversion ?

  


• merge rtmp video. Stream specifier ':v' in filtergraph description

  3 avril 2018, par roomy

  I follow the page Create a mosaic out of several input videos to merge videos.
  When I use http://*.flv video as input, it works fine.

  But when I use rtmp ://** as video input.
  Such as :

      ffmpeg-i rtmp://10.240.209.94:9999/live1 -i rtmp://10.240.209.94:9999/live1 -i rtmp://10.240.209.94:9999/live1 -i rtmp://10.240.209.94:9999/live1
  
  -filter_complex "
  
      nullsrc=size=640x480 [base];
  
      [0:v] setpts=PTS-STARTPTS, scale=320x240 [upperleft];
  
      [1:v] setpts=PTS-STARTPTS, scale=320x240 [upperright];
  
      [2:v] setpts=PTS-STARTPTS, scale=320x240 [lowerleft];
  
      [3:v] setpts=PTS-STARTPTS, scale=320x240 [lowerright];
  
      [base][upperleft] overlay=shortest=1 [tmp1];
  
      [tmp1][upperright] overlay=shortest=1:x=320 [tmp2];
  
      [tmp2][lowerleft] overlay=shortest=1:y=240 [tmp3];
  
      [tmp3][lowerright] overlay=shortest=1:x=320:y=240
  
  "
  
  -f flv rtmp://10.240.209.94:9999/live2

  It tells me :

  Stream specifier ':v' in filtergraph description  nullsrc=size=640x480 [base];[0:v] setpts=PTS-STARTPTS, scale=320x240 [upperleft];[1:v] setpts=PTS-STARTPTS, scale=320x240 [upperright];[2:v] setpts=PTS-STARTPTS, scale=320x240 [lowerleft];[3:v] setpts=PTS-STARTPTS, scale=320x240 [lowerright];[base][upperleft] overlay=shortest=1 [tmp1];[tmp1][upperright] overlay=shortest=1:x=320 [tmp2];[tmp2][lowerleft] overlay=shortest=1:y=240 [tmp3];[tmp3][lowerright] overlay=shortest=1:x=320:y=240 matches no streams.

  Is that a bug ? but I use the newest ffmepg.
  bug

• Can you convert a video to an audio file in python without saving to a file ?

  31 janvier 2017, par user2983738

  For a python web project I receive a video file through a flask interface.
  Now to do analysis, I need to extract the audio file from it.

  Problem is that I can’t save the video to file first, which seemingly eliminates using ffmpeg as it requires a link to a file location.

  Are there any python solutions that would allow me to pass and return audio/video files using objects only ?

  Thank you in advance