Recherche avancée

Sur d’autres sites (5563)

• how to remove PTS gap within a file when transcoding with ffmpeg ?

  18 juillet 2017, par jsBaek

  i have a video which is from rtmp streaming.

  Since the broadcasting is on and off frequently,

  the archived file has PTS like below

  (assume that this is sec)

  0—10—20—30—40 120—130—140

  there’s discontinuity between 40 and 120 sec.

  duration of this file must be 60sec since there’s 80sec gap between 40 120.

  but when i transcoded this file, final duration became 140sec with 80sec of pausing parts.

  how can i transcode this file without "not existing" 80 sec so that output file became 60sec without redundant pausing 80 sec.

  i tried "+getpts" or "+igndts" options but they don’t work at all.

• Pipe output of ffmpeg using nodejs stdout

  21 mai 2014, par rughimire

  I am not being able to pipe the output of the ffmpeg over a stdout.

  Following are the block of code what I coded so far.

      var http = require('http')
  
      , fs = require('fs') 
  
      var child_process = require("child_process")
  
  
  
      http.createServer(function (req, res) {
  
      console.log("Request:", dump_req(req) , "\n")
  
  
  
      // path of the 
  
      var path = 'test-mp4.mp4'  //test-mp4-long.mp4
  
      , stat = fs.statSync(path)
  
      , total = stat.size
  
  
  
  
  
      var range = req.headers.range
  
      , parts = range.replace(/bytes=/, "").split("-")
  
      , partialstart = parts[0]
  
      , partialend = parts[1]
  
      , start = parseInt(partialstart, 10)
  
      , end = partialend ? parseInt(partialend, 10) : total-1
  
      , chunksize = (end-start)+1
  
  
  
  
  
      console.log('RANGE: ' + start + ' - ' + end + ' = ' + chunksize +  "\n")
  
  
  
  
  
      var ffmpeg = child_process.spawn("ffmpeg",[
  
              "-i", path,             // path
  
              "-b:v" , "64k",         // bitrate to 64k
  
              "-bufsize", "64k", 
  
              "-"                     // Output to STDOUT
  
          ]);
  
  
  
  
  
      //set header
  
      res.writeHead(206
  
      , { 'Content-Range': 'bytes ' + start + '-' + end + '/' + total
  
      , 'Accept-Ranges': 'bytes', 'Content-Length': chunksize
  
      , 'Content-Type': 'video/mp4'
  
      })
  
  
  
      stdout[ params[1] ] = ffmpeg.stdout
  
  
  
      // Pipe the video output to the client response
  
      ffmpeg.stdout.pipe(res);
  
  
  
      console.log("Response", dump_res(res), "\n")
  
      }).listen(1337)

  When i replaced the ffmpeg stuffs from above code, all works fine. Following is the part of the code when i replace the ffmpeg stuffs.

   var file = fs.createReadStream(path, {start: start, end: end})

  And piping like :

  file.pipe(res)

  What wrong I am running ?

  Edit :
  The ffmpeg command works fine. I have tested this through the command line and generating proper output.

• ffmpeg transcoding in python with aws-lambda limits

  14 juillet 2017, par Michael

  I want to transcode audio file into HLS with segments in aws-lambda environment :

  ffmpeg -i 2540.mp3 -f segment -segment_time 20 -segment_list battle.m3u8 -segment_format mpegts battle%05d.ts

  But there are cases when I need to transcode files larger than 500 MB (disc space limit for aws-lambda).

  How can I trancode audio file parts with segments ? Is it possible to redirect segments to pipe:1 (stdout), and upload these segments and manifes from stdout to s3 in separated files ?