Newest 'ffmpeg' Questions - Stack Overflow

Les articles publiés sur le site

• Huge latency in HTTP streaming desktop to browser

  29 janvier 2019, par fineman

  I'm trying to put together a reliable, reasonably low (<2s) latency desktop window share to browser solution. Currently I have:

  client sender using FFMPEG:

  ffmpeg -f gdigrab -i "title=notepad.exe" -r 10 -framerate 10  -c:v libx264 -g 50  -preset fast -tune zerolatency -f rtp rtp://192.168.1.85:1234
  

  server re-stream to HTTP using VLC:

  vlc -vv test.sdp  --sout=#transcode{vcodec=theo,vb=1600,scale=1,channels=1,acodec=none}:http{dst=:8080/webcam.ogg} :no-sout-rtp-sap :no-sout-standard-sap :sout-keep 
  

  where the sdp file is generated from the output of the ffmpeg command

  Client browser:

  This works and gives good quality. But the latency is terrible (around 10s) and I'm at a loss to know how to tune it. I know that the latency is in the VLC transcoding/restreaming - displaying the RTP stream from the client on the server only has around 1s lag.

  I guess there are two questions - can this approach be sensibly tuned, or is the approach wrong to start with?

• How to execute commands through PHP ?

  29 janvier 2019, par SuperSimplePimpleDimple

  I am trying to convert videos into MP4 using FFMPEG. I have it set up this way:

  .
  .
  private $ffmpegPath;
  
  public function __construct($con) {
      $this->con = $con;
      $this->ffmpegPath = realpath("ffmpeg/bin/ffmpeg.exe");
  }
  .
  .
  public function convertVideoToMp4($tempFilePath, $finalFilePath){
      $cmd = "$this->ffmpegPath -i $tempFilePath $finalFilePath 2>&1";
  
      $outputLog = array();
      exec($cmd, $outputLog, $returnCode);
  
      if($returnCode != 0){
          foreach ($outputLog as $line){
              echo $line."
  ";
              return false;
          }
      }
  
      return true;
  }
  

  And in the browser i get the following error: 'C:\xampp\htdocs\Thinksmart First Sprint' is not recognized as an internal or external command".

  In my constructor i have it set up to give me the realpath and i suspect that this is what it does in the command line:

  C:/xampp/htdocs/Thinksmart FIrst Sprint/ffmpeg/bin/ffmpeg.exe -i (file temp name) (file name i want)

  And this should work, but i dont know why it wont. Any ideas? Its my first time working with video conversions.

• Error : ENOENT : no such file or directory ( AWS Lambda function)

  29 janvier 2019, par Arun

  I am trying to convert the video file to audio using FFMPEG. But I keep getting this error while converting video to audio in AWS Lambda function. I searched a lot of googles but I can't figure out a suitable solution. If anyone knows the answer please share your solution. I referred this video to audio convertion method from this post.

  Error:

  { Error: ENOENT: no such file or directory, lstat '/var/task/tmp/c82f117b7841f1c2a4c9cd86cd93aad9.mp3'
  at Error (native)
  at Object.fs.lstatSync (fs.js:994:11)
  at Object.byteLength (/var/task/node_modules/aws-sdk/lib/util.js:175:30)
  at Request.SET_CONTENT_LENGTH (/var/task/node_modules/aws-sdk/lib/event_listeners.js:161:40)
  at Request.callListeners (/var/task/node_modules/aws-sdk/lib/sequential_executor.js:106:20)
  at Request.emit (/var/task/node_modules/aws-sdk/lib/sequential_executor.js:78:10)
  at Request.emit (/var/task/node_modules/aws-sdk/lib/request.js:683:14)
  at Request.transition (/var/task/node_modules/aws-sdk/lib/request.js:22:10)
  at AcceptorStateMachine.runTo (/var/task/node_modules/aws-sdk/lib/state_machine.js:14:12)
  at /var/task/node_modules/aws-sdk/lib/state_machine.js:26:10
  
  message: 'ENOENT: no such file or directory, lstat 
  \'/var/task/tmp/c82f117b7841f1c2a4c9cd86cd93aad9.mp3\'',
  errno: -2,
  code: 'ENOENT',
  syscall: 'lstat',
  path: '/var/task/tmp/c82f117b7841f1c2a4c9cd86cd932332.mp3'}
  

  Code

  const child_process = require('child_process');
  const fs = require('fs');
  const path = require('path');
  
  const AWS = require('aws-sdk');
  const request = require('request');
  const tempy = require('tempy');
  
  const s3 = new AWS.S3();
  
  
  exports.handler = (event, context, callback) => {
    // We're going to do the transcoding asynchronously, so we callback immediately.
    callback();
  
    // Extract the event parameters.
    const { mp3Key, url } = event;
  
    const filename = event.filename || path.basename(mp3Key);
    const logKey = event.logKey || `${mp3Key}.log`;
    const s3Bucket = event.s3Bucket || 'bucket-name;
  
    // Create temporary input/output filenames that we can clean up afterwards.
    const inputFilename = tempy.file();
    const mp3Filename = tempy.file({ extension: 'mp3' });
  
    // Download the source file.
    Promise.resolve().then(() => new Promise((resolve, revoke) => {
      const writeStream = fs.createWriteStream(inputFilename);
      writeStream.on('finish', resolve);
      writeStream.on('error', revoke);
      request(url).pipe(writeStream);
    }))
    // Perform the actual transcoding.
    .then(() => {
      // Use the Exodus ffmpeg bundled executable.
      const ffmpeg = path.resolve(__dirname, 'exodus', 'bin', 'ffmpeg');
  
      // Convert the FLV file to an MP3 file using FFmpeg.
      const ffmpegArgs = [
        '-i', inputFilename,
        '-vn', // Disable the video stream in the output.
        '-acodec', 'libmp3lame', // Use Lame for the mp3 encoding.
        '-ac', '2', // Set 2 audio channels.
        '-q:a', '6', // Set the quality to be roughly 128 kb/s.
        mp3Filename,
      ];
      const process = child_process.spawnSync(ffmpeg, ffmpegArgs);
      console.log("process ", process.stdout);
      // return process;
    // return process.stdout.toString() + process.stderr.toString();
    })
    // Upload the generated MP3 to S3.
    .then(logContent => new Promise((resolve, revoke) => {
      console.log("inside s3 upload", mp3Filename)
      s3.putObject({
        Body: fs.createReadStream(mp3Filename),
        Bucket: s3Bucket,
        Key: mp3Key,
        ContentDisposition: `attachment; filename="${filename.replace('"', '\'')}"`,
        ContentType: 'audio/mpeg',
      }, (error) => {
        if (error) {
          revoke(error);
        } else {
          // Update a log of the FFmpeg output.
          const logFilename = path.basename(logKey);
          console.log("log file upload")
          s3.putObject({
            Body: logContent,
            Bucket: s3Bucket,
            ContentType: 'text/plain',
            ContentDisposition: `inline; filename="${logFilename.replace('"', '\'')}"`,
            Key: logKey,
          }, resolve);
        }
      })
    }))
    .catch(console.error)
    // Delete the temporary files.
    .then(() => {
      [inputFilename, mp3Filename].forEach((filename) => {
        if (fs.existsSync(filename)) {
          fs.unlinkSync(filename);
        }
      });
    });
  };
  

• Want to overlay video over an Image - FFMPEG

  29 janvier 2019, par Siddharth

  I'm willing to overlay video over an Image just as below:

  The Car is the image and rest orange or background you are seeing is the video.

  Now for overlaying video over the image I'm using the following command:

  ffmpeg -i image.jpg -i input.mp4
     -filter_complex "[1]split=2[color][alpha];
                      [color]crop=iw/2:ih:0:0[color];[alpha]crop=iw/2:ih:iw/2:0[alpha];
                      [color][alpha]alphamerge[ovrly];
                      [0][ovrly]overlay=0:0" output.mp4
  

  and its throwing me an error as below:

  At least one output file must be specified
  zsh: command not found: -filter_complex
  

  What am I doing wrong here?

  Any help will be highly appreciated.

• How can I capture and record the desktop and also the sound from speakers using ffmpeg ?

  29 janvier 2019, par Dubi Duboni

  This is how I'm recording the video of the desktop without sound: And it's working good:

  ffmpeg -f gdigrab -framerate 24 -i desktop -preset ultrafast -pix_fmt yuv420p out.mp4
  

  But now I want to capture and record also the sound so I tried:

  ffmpeg -f gdigrab -framerate 24 -i audio="Stereo Mix (Realtek High Definition Audio)"  desktop -preset ultrafast -pix_fmt yuv420p testing002.mp4
  

  But getting error:

  [gdigrab @ 0000000000df2cc0] Please use "desktop" or "title=" to specify your target. audio=Stereo Mix (Realtek High Definition Audio): I/O error

  This is a screenshot of my speakers properties: