Recherche avancée

Sur d’autres sites (6931)

• FFmpeg : concatenation problems

  13 décembre 2016, par CryGuy

  I’d like to concat two mpeg files. Both have an MP2 audio stream that I select, but even though copying is intended, an AC3 stream is written.

  Input files :

  Input #0, mpeg, from 'D:\a.mpg':
  
    Duration: 01:25:54.05, start: 0.500000, bitrate: 4528 kb/s
  
      Stream #0:0[0x1bf]: Data: dvd_nav_packet
  
      Stream #0:1[0x1c0]: Audio: mp2, 48000 Hz, stereo, s16p, 192 kb/s
  
      Stream #0:2[0x80]: Audio: ac3, 48000 Hz, stereo, fltp, 384 kb/s
  
      Stream #0:3[0x1e0]: Video: mpeg2video (Main), yuv420p(tv, top first), 720x576 [SAR 64:45 DAR 16:9], 25 fps, 25 tbr, 90k tbn, 50 tbc
  
  Input #1, mpeg, from 'D:\b.mpg':
  
    Duration: 00:12:53.11, start: 0.500000, bitrate: 4486 kb/s
  
      Stream #1:0[0x1c0]: Audio: mp2, 48000 Hz, stereo, s16p, 192 kb/s
  
      Stream #1:1[0x80]: Audio: ac3, 48000 Hz, stereo, fltp, 384 kb/s
  
      Stream #1:2[0x1e0]: Video: mpeg2video (Main), yuv420p(tv, top first), 720x576 [SAR 64:45 DAR 16:9], 25 fps, 25 tbr, 90k tbn, 50 

  ffmpeg args :

  -i "d:\a.mpg" -i "d:\b.mpg" -filter_complex "[0:3][0:1][1:2][1:0]concat=n=2:v=1:a=1[v][a]" -map [v] -map [a] -c copy -target pal-dvd "d:\c.mpg"

  As you can see, I selecting the video and the mp2 stream for both input files.

  Output :

  Stream mapping:
  
    Stream #0:1 (mp2) -> concat:in0:a0
  
    Stream #0:3 (mpeg2video) -> concat:in0:v0
  
    Stream #1:0 (mp2) -> concat:in1:a0
  
    Stream #1:2 (mpeg2video) -> concat:in1:v0
  
    concat:out:v0 -> Stream #0:0 (mpeg2video)
  
    concat:out:a0 -> Stream #0:1 (ac3)

  1. As you can see in the last line, the concat filter outputs an AC3 stream. Why ?
  2. The input is interlaced material. In the output file, the interlaced flag is not set correctly so that it’s played back interlaced.

  I’ve been trying to solve this for a couple of hours now, with and without different mapping options.

  Thanks for reading.

• MP3 Audio requiring overlays at specific intervals to ensure I dont break copyright rules

  15 novembre 2012, par user1828008

  I have started a project as a DJ to record my online radio shows to mp3. The station I play on is licenced so I am not breaking any copyright rules there. I have had the idea to upload the recordings onto Youtube.

  Clearly that needed some form of adjustment to make a video over the top of the audio which I have achieved using ffmpeg so I have an mp4 file with a picture overlay for the duration of the audio. I did a quick test run to upload the video to Youtube and it tracked that I was playing copyrighted material. It was exactly right in its statement and yes as my own channel on youtube I am not licenced to do so. Either way I have spoke to the copyright owners and they have agreed it is fine.

  However in order to not have to do this every week if I overlay a 5 second 'jingle' every 4 minutes over the audio (fading down if possible) the main volume slightly to play the jingle and fade up after (once again if possible) then I am 100% in the clear. I could do this manually but would rather something I can script in bash to do it for me.

  I am using Centos 6.3 operating system and FFMPEG solved the merge of video and audio together, but I couldnt find anything to do a predefined audio overlay at specific intervals.

  I have looked at programs like Avisynth and Sox but dont believe from the documentation that they have anything that can do regular interval overlays during a single track. The closest thing I could find is on this post :

  Add multiple audio files to video at specific points using FFMPEG

  But it would be nice to do everything in a single sweep with ffmpeg of the audio file (video and audio overlay) if thats possible somehow or any better options that are out there ?

  If you need any more info from me then please let me know.
  Thanks

• Cutting and fading ts segment with ffmpeg ?

  6 juillet 2017, par Matt W

  I’m trying to cut a ts segment and apply a fade from white at the initial point of the cut segment. This is used in a situation where I need to "crop" some material from the beginning of a video and apply a fade for a smoother entry. In my test, I am cutting an 8 second segment at the 5.5s mark to yield a 2.5 second segment that will fade from white over the first second.

  The cut command by itself works fine :

  ffmpeg  -i test.ts  -ss 5.5 -c:v libx264 -profile:v baseline -c:a aac -map 0 -mpegts_copyts 1  -preset ultrafast -f ssegment -initial_offset 5.5 -segment_format mpegts   ~/Desktop/cut%d.ts

  I’ve successfully used this fade filter syntax before :

  -filter:v fade=t=in:st=0:d=1:color=0xffffff

  But I can’t seem to make the whole thing work :

  ffmpeg  -i test.ts  -filter:v fade=t=in:st=5.5:d=1:color=0xffffff  -ss 5.5 -c:v libx264 -profile:v baseline -c:a aac -map 0 -mpegts_copyts 1  -preset ultrafast -f ssegment -initial_offset 5.5 -segment_format mpegts   ~/Desktop/cut%d.ts

  I’m getting the following error :

  x264 [error] : baseline profile doesn’t support 4:4:4

  [libx264 @ 0x7fd9db002400] Error setting profile baseline.

  [libx264 @ 0x7fd9db002400] Possible profiles : baseline main high high10
  high422 high444

  Error initializing output stream 0:1 — Error while opening encoder for output stream #0:1 - maybe incorrect parameters such as bit_rate, rate, width or height

  [aac @ 0x7fd9db001200] Qavg : nan

  Conversion failed !

  My knowledge of ffmpeg just isn’t deep enough to figure out why this is happening. Any thoughts ?

  Thanks,
  Matt