Recherche avancée

Sur d’autres sites (14352)

• ffmpeg poor audio(only) streaming quality ? how to configure

  9 mai 2021, par Tera

  ffmpeg streams audio only with bad quality.. i need better quality, lantency isn't a problem !

  



  im making a discord bot that streams audio from my local input..

  



  i saw the -aq short for -q:audio but changing that doesn't seem to make a diffrence.. im sure 48k samplerate should be plenty.
need stereo channel.

  



  have no idea how to change the bitrate or the compression... maybe that might be the problem ?

  



  using ffmpeg i got it working but i found that the quality isn't perfect..
how do i get loseless audio only ? or is there somthing wrong with my parameters ?

  



    {
    FileName = "ffmpeg.exe",
    Arguments = $"-hide_banner -loglevel panic -f dshow -i audio=\"{Program.AUDIO_DEVICE_NAME}\" -aq 8 -ac 2 -f s16le -ar 48000 pipe:1",
    UseShellExecute = false,
    RedirectStandardOutput = true
  }


  



  expected : close too or lossless quality, high lantency is fine (not more than 10 secs)

  



  reuslt : 8/10 quality and 1.5 sec lantency.. could be better

  



  if ffmpeg isn't the answer to my solution what else can i use ?

  



  edit : found out that the -f s16le is an audio format i tried switching to other pcm with no luck s24le gives me white noise with barely audbile audio
same as s32le ? im using voice meter btw as a virtual input i suspect the low quality is because the format is only 16bit..

  



  edit2 : bitrate other than 48000 seems to dail the songs in diffrent speeds

  


• Concatenate Videos with FFMPEG - Poor Audio

  7 septembre 2018, par swg1cor14

  So here is my workflow....all using ffmpeg in Ubuntu.

  Step 1 : Convert main mp4 video to mpegts

  Step 2 : Convert mp3 recording to m4a.

  Step 3 : Attach m4a to still image

  Step 4 : Convert into to mpegts

  Step 5 : Concat intro and main video and convert to mp4

  /home/admin/ffmpeg/ffmpeg -i Wilfried_Ligthart_Profits.mp4 -c copy -bsf:v h264_mp4toannexb -f mpegts -r 30 Wilfried_Ligthart_Profits.ts
  
  /home/admin/ffmpeg/ffmpeg -i 5b8955ef7c324-intro.mp3 -c:a libfdk_aac -strict -2 intro_audio_90.m4a
  
  /home/admin/ffmpeg/ffmpeg -loop 1 -i slide_9.png -i intro_audio_90.m4a -acodec copy -shortest intro_merged_90.mp4
  
  /home/admin/ffmpeg/ffmpeg -i intro_merged_90.mp4 -c copy -bsf:v h264_mp4toannexb -f mpegts -r 30 introFile90.ts
  
  /home/admin/ffmpeg/ffmpeg -i "concat:introFile90.ts|Wilfried_Ligthart_Profits.ts" -c copy -bsf:a aac_adtstoasc merged90.mp4

  So each part (introFile90.ts and Wilfried_Lighthart_Profits.ts) sound fine. However on the last step, the merged90.mp4, the audio is slowed down.

  I also get a bunch of these errors :

  [mp4 @ 0x31ad080] Non-monotonous DTS in output stream 0:1; previous: 59837710, current: 59837012; changing to 59837711. This may result in incorrect timestamps in the output file.

  Anyways, what is wrong ? I was told to convert them to mpegts would make renders faster.

• vaapi_encode : Destroy output buffer pool before VA context

  17 décembre 2017, par Mark Thompson

  vaapi_encode : Destroy output buffer pool before VA context
  The buffers are created associated with the context, so they should be
  
  destroyed before the context is.  This is enforced by the iHD driver.
  

  • [DBH] libavcodec/vaapi_encode.c