Recherche avancée

Sur d’autres sites (10621)

• Getting error "WebAssembly.Memory() : could not allocate memory" when running ffmpeg.wasm on android chrome browser

  27 juin 2022, par Ravi Kundu

  Error in detail :

  


  WebAssembly.Memory(): could not allocate memory
at https://*******/ffmpeg-core.js:22:82
at https://*******/ffmpeg.min.js:1:6506
at f (https://*******/ffmpeg.min.js:1:11322)
at Generator._invoke (https://*******/ffmpeg.min.js:1:11110)
at Generator.next (https://*******/ffmpeg.min.js:1:11747)
at i (https://*******/ffmpeg.min.js:1:4295)
at c (https://*******/ffmpeg.min.js:1:4498)


  


  Code for ffmpeg :

  


  const downloadWithFFMPEG = async () =>{
  const sourceBuffer = await fetch(recordingURL).then(r => r.arrayBuffer());
  await ffmpeg.load();
  await ffmpeg.FS(
    "writeFile",
    "input.webm",
    new Uint8Array(sourceBuffer, 0, sourceBuffer.byteLength)
  );
  await ffmpeg.run("-i", "input.webm", "-c", "copy", "output.mp4")
  const output = ffmpeg.FS("readFile", "output.mp4");
  var link = document.createElement('a')
  link.href = URL.createObjectURL(new Blob([output.buffer], { type: 'video/mp4;codecs=H264' }));
  link.download = this.data;
  link.click();
  recording = false;
}


  


  Brief about problem :
The error only comes for android chrome browser. The same code works fine on pc/laptop chrome.
Have also enabled Webassembly-thread on chrome ://flags for android browser as someone suggested me to do it but still same error. Can someone help me ?

  


• Using ffmpeg to "re-time" an RTSP stream ?

  29 mars 2021, par Tom S

  I recently set up a birdbox in my garden, and have been using a cheapie camera that I bought, along with ffmpeg on a raspberry pi to stream from the camera to YouTube.

  


  But YouTube keeps telling me that I am sending data too fast - multiple seconds of video per second, and at that point the stream seems to stall and I just get a buffering spinner on YouTube and have to restart ffmpeg to restore the stream.

  


  This is happening very frequently, infact even more frequently now than when I originally got everything up and running for some reason... So much so that Ive had to setup ffmpeg to run as a service, and restart it once an hour via cron. But even this is not enough and now it seems that it can go bonkers within 10-15 minutes of a restart - almost like the more often I restart the quicker it gets...

  


  I dont know where the issue is, or how to find it (does ffmpeg simply restream "verbatim" what ever the camera supplies ?), but I wonder if it might be possible to have ffmpeg re-time the stream so that I only send to YouTube the 25fps that it is happy with (i.e. dropping any frames in excess of this) ?

  


  I have key frames set to 1 second on the camera.

  


  Thanks !

  


  ---

  


  Here is my service file which contains the command that I use to start ffmpeg :

  


  [Unit]
Description=BirbCam 1 streaming service
After=network.target
StartLimitIntervalSec=0

[Service]
Type=simple
Restart=always
RestartSec=1
User=root
# Uncomment one ExecStart line below, then run `sudo systemctl daemon-reload` before restart
# With audio
ExecStart=/usr/local/bin/ffmpeg -rtsp_transport tcp -i rtsp://172.25.144.35:554/user=admin&password=&channel=1&stream=0.sdp?real_stream -tune zerolatency -vcodec libx264 -pix_fmt + -c:v copy -c:a aac -strict experimental -f flv rtmp://a.rtmp.youtube.com/live2/xxxx-xxxx-xxxx-xxxx-xxxx

# Without audio
# ExecStart=/usr/local/bin/ffmpeg -f lavfi -i anullsrc -rtsp_transport tcp -i rtsp://172.25.144.35:554/user=admin&password=&channel=1&stream=0.sdp?real_stream -tune zerolatency -vcodec libx264 -pix_fmt + -c:v copy -c:a aac -strict experimental -f flv rtmp://a.rtmp.youtube.com/live2/xxxx-xxxx-xxxx-xxxx-xxxx

[Install]
WantedBy=multi-user.target


  


• typeError when using ffmpeg with buffer in NodeJS ["argument must be of type string or an instance of Buffer"]

  16 mars 2021, par coolps811

  I am trying to covert buffer data into the correct mp4 video format. However I am getting an error : "UnhandledPromiseRejectionWarning : TypeError [ERR_INVALID_ARG_TYPE] : The "data" argument must be of type string or an instance of Buffer, TypedArray, or DataView. Received an instance of FfmpegCommand". How can I fix this ?

  


  router.post("/download", (req, res, next) => {
  axios({
    method: "get",
    url: req.body.url,
    responseType: "arraybuffer",
  }).then(function (response) {
    const data = new Uint8Array(Buffer.from(response.data));

    const proc = new ffmpeg(data)
      .videoCodec("libx264")
      .outputOptions(["-movflags isml+frag_keyframe"])
      .toFormat("mp4")
      //.seekInput(offset) this is a problem with piping
      .on("error", function (err, stdout, stderr) {
        console.log("an error happened: " + err.message);
        console.log("ffmpeg stdout: " + stdout);
        console.log("ffmpeg stderr: " + stderr);
      })
      .on("end", function () {
        console.log("Processing finished !");
      })
      .on("progress", function (progress) {
        console.log("Processing: " + progress.percent + "% done");
      });

    fs.writeFile("Assets/test.mp4", proc, callback);
  });

  const callback = (err) => {
    if (err) throw err;
    console.log("It's saved!");
  };
});