Recherche avancée

Sur d’autres sites (6397)

• FFmpeg delay and mix audio streams while keeping overall volume constant

  5 octobre 2020, par unstuck

  I have about 100 audio streams, all with the same intro music/sound, and in some of them the intro is delayed by a few seconds. I want to align and mix all the audio streams such that all the intros play at the same time and the output remains pretty much the same volume throughout. I know in advance how much each stream needs to be delayed by.

  


  Like this in Audacity. Each audio stream is aligned to the intro, and the duration before the intro is arbitrary. (This doesn't solve the volume problem though.)

  


  What I have so far uses adelay and amix. It looks something like this but with more audio streams.

  


  ffmpeg -i 00.oga \
       -i 01.oga \
       -i 02.oga \
       -i 03.oga -filter_complex \
"[0]adelay=delays=     123S:all=1[a0]; \
 [1]adelay=delays=    2718S:all=1[a1]; \
 [2]adelay=delays= 6283185S:all=1[a2]; \
 [3]adelay=delays=11235813S:all=1[a3]; \
 [a0][a1][a2][a3]amix=inputs=4" output.oga


  


  In this example the first stream is delayed by 123 samples, the second by 2 718, the third by 6 283 185, and the by fourth 11 235 813.

  


  This works, except at the beginning of the output it's very quiet. When fed n streams, amix makes each stream 1/nth its original volume, which is a good thing in principle. In this case it's not an entirely good thing, because at the beginning of the output 3 of the 4 audio streams are silent (adelay fills delayed streams with silence), meaning the only audible stream is 1/4 = 25% of its original volume. When the second stream becomes audible, the overall volume is 2/4, with three audible streams 3/4, and with all four streams audible it's 4/4 = 100%.

  


  Instead, I want the the first stream to be at 100% volume when it's the only audible one, 50% volume each when there are two audible streams, etc.

  


  Is there a way to make it so when there are n audio streams but m non-silent audio streams, the volume for each of the audio streams is 1/m not 1/n ? amix does this when streams end ; if one stream ends it changes the volume of the others from 1/n to 1/n-1 over a period of time (dropout_transition : https://ffmpeg.org/ffmpeg-filters.html#amix).

  


  I found a similar question where someone wanted to do something like this but only with 2 audio streams. The answer was to split, trim, and change the volume manually. This would be incredibly complicated with 100 audio streams or more, like in my situation.

  


  Is there any easy way to achieve this, even without FFmpeg ?

  


• Getting ffmpeg error ‘Ignoring unsupported var reason rtmp :// : I/O error’

  6 septembre 2022, par Lectos Lacious

  Same setting worked normally for months, suddenly start to receive ffmpeg errors about ‘unsupported var’ related to rtmp and streaming stop working. Is this error about unavailable server or bad url ? Can someone enlighten me about meaning of unsupported var ?

  


  NOTE : replaced actual url with ‘someurl’

  


  LOG :
Input #0, rtsp, from 'rtsp ://192.168.40.44:554/1' :
  
Metadata :
  
title : 10
  
Duration : N/A, start : 0.000000, bitrate : N/A
  
Stream #0:0 : Video : h264, yuvj420p(pc, bt709), 1280x720, 15 tbr, 90k tbn, 180k tbc
  
Stream #0:1 : Audio : aac, 48000 Hz, stereo, fltp
  
[rtmp @ 0x7f9fdd58c5e0] Ignoring unsupported var reason
  
rtmp ://someurl : I/O error
  
Exiting normally, received signal 2.
  
Input #0, rtsp, from 'rtsp ://192.168.40.44:554/1' :
  
Metadata :
  
title : 10
  
Duration : N/A, start : 0.000000, bitrate : N/A
  
Stream #0:0 : Video : h264, yuvj420p(pc, bt709), 1280x720, 15 fps, 15 tbr, 90k tbn, 180k tbc
  
Stream #0:1 : Audio : aac, 48000 Hz, stereo, fltp
  
[rtmp @ 0x7f62a5869b20] Ignoring unsupported var reason
  
rtmp ://someurl : I/O error
  
Exiting normally, received signal 2.
  
[rtsp @ 0x7f0b4add32a0] max delay reached. need to consume packet
  
[rtsp @ 0x7f0b4add32a0] RTP : missed 2 packets
  
Input #0, rtsp, from 'rtsp ://192.168.40.44:554/1' :
  
Metadata :
  
title : 10
  
Duration : N/A, start : 0.000000, bitrate : N/A
  
Stream #0:0 : Video : h264, yuvj420p(pc, bt709), 1280x720, 15 fps, 15 tbr, 90k tbn, 180k tbc
  
Stream #0:1 : Audio : aac, 48000 Hz, stereo, fltp
  
[rtmp @ 0x7f0b4af13840] Ignoring unsupported var reason
  
rtmp ://someurl : I/O error
  
Exiting normally, received signal 2.
  
Input #0, rtsp, from 'rtsp ://192.168.40.44:554/1' :
  
Metadata :
  
title : 10
  
Duration : N/A, start : 0.000000, bitrate : N/A
  
Stream #0:0 : Video : h264, yuvj420p(pc, bt709), 1280x720, 15 fps, 15 tbr, 90k tbn, 180k tbc
  
Stream #0:1 : Audio : aac, 48000 Hz, stereo, fltp
  
[rtmp @ 0x7fc61ef8b880] Ignoring unsupported var reason

  


• lavfi/concat : allow to support inputs with different frame rates

  30 août 2019, par Calvin Walton

  lavfi/concat : allow to support inputs with different frame rates
  Right now, the concat filter does not set the frame_rate value on any of
  
  the out links. As a result, the default ffmpeg behaviour kicks in - to
  
  copy the framerate from the first input to the outputs.
  If a later input is higher framerate, this results in dropped frames ; if
  
  a later input is lower framerate it might cause judder.
  This patch checks if all of the video inputs have the same framerate, and
  
  if not it sets the out link to use '1/0' as the frame rate, the value
  
  meaning "unknown/vfr".
  A test is added to verify the VFR behaviour. The existing test for CFR
  
  behaviour passes unchanged.
  

  • [DH] libavfilter/avf_concat.c
  • [DH] tests/fate/filter-video.mak
  • [DH] tests/filtergraphs/concat-vfr
  • [DH] tests/ref/fate/filter-concat-vfr