Recherche avancée

Sur d’autres sites (7054)

• Making youtube-dl download mp3's faster

  12 novembre 2017, par sciencelord

  It takes a long time to download mp3 songs using youtube-dl, and it takes even longer when downloading a bunch of mp3 songs. Is there a way to make it significantly faster ? I don’t mind reducing quality. I’ve been using the command below.

  youtube-dl --extract-audio --audio-format "mp3" --output "%(title)s.%(ext)s" "https://www.youtube.com/watch?v={videoid}

  I also took a look at this post :
  ffmpeg command for faster encoding at a decent bitrate with smaller file size

  But I wasn’t sure how to change the above command using the ffmpeg modifications.

  Thanks !

• How to download .m3u8 in once time

  11 octobre 2024, par Nabi K.A.Z.

  I have a .m3u8 file on remote host, with contain fixed numbers of chunk .ts file name, and not stream :

  



  #EXTM3U
#EXT-X-VERSION:3
#EXT-X-TARGETDURATION:11
#EXT-X-MEDIA-SEQUENCE:0
#EXTINF:9.736,
media_0.ts
#EXTINF:9.96,
media_1.ts
#EXTINF:10.0,
media_2.ts
#EXTINF:10.0,
media_3.ts
#EXTINF:10.0,
media_4.ts
#EXTINF:10.2,
media_5.ts
#EXTINF:10.0,


  



  When I use this command :

  



  # ffmpeg -i "http://example.com/chunklist.m3u8" file.mp4

frame=  582 fps=9.4 q=28.0 size=    1536kB time=00:00:23.21 bitrate= 542.1kbits/s dup=2 drop=4 speed=0.375x


  



  It works. But It get frame by frame video and very long time needed. (It takes time almost to playback the video.)

  



  But since the path of all the .ts files are known. (http://example.com/media_0.ts, http://example.com/media_1.ts, ...) There must be a way to get and merge them all at the same time.

  



  But How in ffmpeg directly ?!

  



  EDIT (try a solution) :

  



  For one solution, I know how can concatenation files with ffmpeg.

  



  ffmpeg -i "concat:0.ts|1.ts|2.ts|3.ts|4.ts|5.ts" -c copy output.mp4


  



  This ffmpeg command was great, and works in less 1 sec time !

  



  So try to download all .ts files with CURL with this command :

  



  curl \
http://example.com/media_0.ts -o 0.ts \
http://example.com/media_1.ts -o 1.ts \
http://example.com/media_2.ts -o 2.ts \
http://example.com/media_3.ts -o 3.ts \
http://example.com/media_4.ts -o 4.ts \
http://example.com/media_5.ts -o 5.ts


  



  But you can see result :

  



    % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  687k  100  687k    0     0  75108      0  0:00:09  0:00:09 --:--:-- 74111
100  652k  100  652k    0     0  59404      0  0:00:11  0:00:11 --:--:-- 53400
100  673k  100  673k    0     0  48675      0  0:00:14  0:00:14 --:--:-- 55781
100  657k  100  657k    0     0  63573      0  0:00:10  0:00:10 --:--:-- 62494
100  671k  100  671k    0     0  39019      0  0:00:17  0:00:17 --:--:-- 40863
100  692k  100  692k    0     0  63480      0  0:00:11  0:00:11 --:--:-- 80049


  



  See, total download time was 72 sec, while the total duration of all parts is 59 sec ! that this time is very long !

  



  So sorry, download all parts and then concat that, was not good solution.

  



  EDIT 2

  



  I try for another .m3u8 file on the another server with difference URL :

  



  Download and concat together :

  



  ffmpeg -i "concat:\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_0.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_1.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_2.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_3.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_4.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_5.ts\
" -c copy -y output.ts


  



  Another command with input.txt URLs file.

  



  ffmpeg -f "concat" -i "input.txt" -c copy -y output.ts


  



  input.txt file :

  



  file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_0.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_1.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_2.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_3.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_4.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_5.ts'


  



  Or this command some time if needed :

  



  ffmpeg -f "concat" -safe "0" -protocol_whitelist "file,http,https,tcp,tls" -i "input.txt" -c copy -y output.ts


  



  Finally, for that download speed was good, MAYBE my server target has limited bandwidth. :-(

  


• Bash : bash script to download trimmed mp3 from youtube url

  25 août 2017, par Bhishan Poudel

  I would like to download the initially x seconds trimmed mp3 from a video url of youtube.
  I found that youtube-dl can download the video from youtube to local machine. But, when I looked at the man pages of youtube-dl, I could not find any trim options.

  So I tried to use the ffmpeg to trim downloaded mp3 file.
  Instead of doing this is two steps, I like to write one bash script which does the same thing.
  My attempt is given below.

  However, I was stuck at one place :
  "HOW TO GET THE VARIABLE NAME OF OUTPUT MP3 FILE FROM YOUTUBE-DL ?"
  The script is given below :

  # trim initial x seconds of mp3 file
  
  # e.g. mytrim https://www.youtube.com/watch?v=dD5RgCf1hrI 30
  
  function mytrim() {
  
      youtube-dl --extract-audio --embed-thumbnail --audio-format mp3 -o "%(title)s.%(ext)s" $1
  
      ffmpeg -ss $2 -i $OUTPUT_MP3 -acodec copy -y temp.mp3
  
      mv temp.mp3 $OUTPUT_MP3
  
      }

  How to get the variable value $OUTPUT_MP3 ?
  echo "%(title)s.%(ext)s" gives the verbatim output, does not give the output filename.

  How could we make the script work ?

  The help will be appreciated.