Recherche avancée

Sur d’autres sites (4819)

• ffmpeg extract first subtitle for given language or fix mismapped tracks

  11 mars 2023, par iarwain8a

  I'm developing a server to encode any given video and I need to extract its subtitles.

  


  I've tried something like this which worked :

  


  ffmpeg -i pipe:0 -map 0:s:1 -c:s webvtt -f segment -segment_list_flags +live -segment_time 60 -segment_list subs_eng.m3u8 -segment_format webvtt subs_eng.vtt -loglevel debug


  


  This encodes from srt, into webvtt, the second track which happens to be in this video an English subtitle. But very quickly I ran into a problem in which the tracks mappings are not numbered correctly for this video, for example, the (eng) track should be "-map 0:s:2", but it's "-map 0:s:1" as I've said before, and the (bul) track, which should be "-map 0:s:4" is actually "-map 0:s:2".

  


  Stream #0:0: Video: h264 (High), yuv420p(tv, bt709, progressive), 1920x1080 [SAR 1:1 DAR 16:9], 23.98 fps, 23.98 tbr, 1k tbn (default)
  Stream #0:1(eng): Audio: eac3, 48000 Hz, 5.1(side), fltp, 768 kb/s (default)
  Stream #0:2(eng): Subtitle: subrip
    Metadata:
      title           : English
  Stream #0:3(eng): Subtitle: subrip
    Metadata:
      title           : English [SDH]
  Stream #0:4(bul): Subtitle: subrip
    Metadata:
      title           : Bulgarian (Bulgaria)


  


  Given said problem I tried something like this :

  


  ffmpeg -i pipe:0 -map 0:s:m:language:eng -c:s webvtt -f segment -segment_list_flags +live -segment_time 60 -segment_list subs_eng.m3u8 -segment_format webvtt subs_eng.vtt -loglevel debug


  


  Which didn't work because I can't or don't know how, to actually select the first result of this "-map 0:s:m:language:eng" if there is more than 1 "eng" sub track. Is there a way to do this ?
Or the alternative should be : Is there a reason why the metadata of this file is mismapped ? If so is there a way to detect it, to then do something about it ? or have it fixed ?

  


• ffmpeg extract first subtitle for given language or fix missmaped tracks

  11 mars 2023, par iarwain8a

  I'm using developing a server to encode any given video and I need to extract its' subtitles.

  


  I've tried something like this which worked :

  


  ffmpeg -i pipe:0 -map 0:s:1 -c:s webvtt -f segment -segment_list_flags +live -segment_time 60 -segment_list subs_eng.m3u8 -segment_format webvtt subs_eng.vtt -loglevel debug


  


  This encodes from srt, into webvtt, the second track which happens to be in this video an english subtitle. But very quickly I ran into a problem in which the tracks mappings are not numbered correctly for this video, for example, the (eng) track should be "-map 0:s:2", but it's "-map 0:s:1" as I've said before, and the (bul) track, wich should be "-map 0:s:4" is actually "-map 0:s:2".

  


  Stream #0:0: Video: h264 (High), yuv420p(tv, bt709, progressive), 1920x1080 [SAR 1:1 DAR 16:9], 23.98 fps, 23.98 tbr, 1k tbn (default)
  Stream #0:1(eng): Audio: eac3, 48000 Hz, 5.1(side), fltp, 768 kb/s (default)
  Stream #0:2(eng): Subtitle: subrip
    Metadata:
      title           : English
  Stream #0:3(eng): Subtitle: subrip
    Metadata:
      title           : English [SDH]
  Stream #0:4(bul): Subtitle: subrip
    Metadata:
      title           : Bulgarian (Bulgaria)


  


  Given said problem I tried something like this :

  


  ffmpeg -i pipe:0 -map 0:s:m:language:eng -c:s webvtt -f segment -segment_list_flags +live -segment_time 60 -segment_list subs_eng.m3u8 -segment_format webvtt subs_eng.vtt -loglevel debug


  


  Which didn't work because I can't or don't know how, to actually select the first result of this "-map 0:s:m:language:eng" if there is more than 1 "eng" sub track. Is there a way to do this ?
Or the alternative should be : Is there a reason why the metadata of this file is missmapped ? If so is there a way to detect it, to then do something about it ? or have it fixed ?

  


• Extract subtitle by language code via ffmpeg

  7 mai 2023, par TrustFound

  I have a simple task - extract subtitle for exact language from tvshows.
For example, I want to extract English subtitles from Netflix's show.
As you know there're a few different types of subtitles : forced, full and SDH.
So I want to extract all of them if it has eng language code.

  


  To extract 1 subtitle from file I used this code for windows :

  


  FOR %%i IN (*.mkv) DO (ffmpeg.exe -i "%%i" -map 0:s:m:language:eng -c copy "%%~ni".eng.srt)


  


  It worked fine with 1 english subtitle per file. But if it contains 2, ffmpeg shows error

  


  


  SRT supports only a single subtitles stream

  


  


  ---

  


  MI is...

  


  

  • Stream #0:2(eng) : Subtitle : subrip

  


  • Stream #0:3(eng) : Subtitle : subrip

  


  • Stream #0:4(ara) : Subtitle : subrip

  


  • ...

  


  


  ---

  


  So I should set 2 or more output files. I tried to figure out how to do this and found similar threads on reddit and stacksoverflow. They said there's no way to do this without ffprobe.
So I used ffprobe to parse all subtitle tracks and their language code.

  


  FOR %%i IN (*.mkv) DO (ffprobe -loglevel error -select_streams s -show_entries stream=index:stream_tags=language -of csv=p=0 -i %%i > subs.txt)


  


  File contains this info :

  


  

  • 2,eng

  


  • 3,eng

  


  • 4,ara

  


  • ...

  


  


  ---

  


  As I understand I should use integers and set them values 2 and 3. I want to get output like this

  


  


  

  • MovieName.2.eng.srt

  


  • MovieName.3.eng.srt

  


  


  


  If it easier to extract all subs, let it be. I tried to do this too but I dont know how to set integers and use them :(
So what I should do ?
Thanks in advance