Recherche avancée

Sur d’autres sites (12879)

• adding overlay with ffmpeg is very slow

  15 juillet 2024, par Exorcismus

  Am trying to append some PNGs on a video, I've used overlay filter for that, but it has increase processing time 10X !!! am not sure if am doing it wrong or this is expected,

  


  


  


  async function processVideo3() {

  let ffmpegProcess = new ffmpeg("./data/input.mp4")

    //Add your args here
    .withVideoCodec("libx264")
    .withAudioCodec("copy")
    .withOutputFormat("mp4")
    .outputOptions("-preset faster")

  let filterChain = [
    {
      filter: "scale",
      options: "-1:720",
      inputs: '0', outputs: 'a1'
    },
    {
      filter: 'subtitles',
      options: `./data/1721070513371.ass`,
      inputs: 'a1', outputs: 'a2'
    }
  ];

  let lastOutput = 'a2';


  const emojiJson = await extractJsonFromFile(`./data/1721070513371.ass`)


  const timedEmojis = convertTimestampsForFfmpeg(emojiJson.emojis)
  // Add filters for each PNG with its respective timestamps
  timedEmojis.forEach((pngWithTimestamp, index) => {
    const pngInputIndex = index + 1;
    const { emoji, start, end } = pngWithTimestamp;

    // Input the PNG URL
    ffmpegProcess.input(emoji);

    // Scale filter for the PNG
    filterChain.push({
      filter: 'scale',
      options: { w: 50, h: 50 },
      inputs: pngInputIndex.toString(), outputs: `b${pngInputIndex}`
    });

    // Overlay filters for each timestamp
    const outputLabel = `c${pngInputIndex}_${1}`;
    const overlayFilter = {
      filter: 'overlay',
      options: {
        eval: "init",
        format: "auto",
        x: "(main_w-overlay_w)/2",
        y: "(main_h-overlay_h)/2",
        enable: `between(t,${start},${end})`
      },
      inputs: [lastOutput, `b${pngInputIndex}`],
      outputs: outputLabel
    };
    filterChain.push(overlayFilter);
    lastOutput = outputLabel;
  });

  // Remove the output label from the last filter
  filterChain[filterChain.length - 1].outputs = undefined;


  await ffmpegProcess.complexFilter(filterChain)
    .save("output.mp4")



  ffmpegProcess.on("progress", (progress) => console.log(progress))

  ffmpegProcess.on('error', async function (err, stdout, stderr) {
    if (err) {
      console.log(err.message);
      console.log("stdout:\n" + stdout);
      console.log("stderr:\n" + stderr);
      reject("Error");
    }
  }).on('end', async () => {

    console.log('Finished processing');

  })

}

  


  


  



  can some one advise on this ?

  


• Extracting thumbnails with FFMPEG is super slow on large video files ? [duplicate]

  5 octobre 2015, par vaid

  This question already has an answer here :

  • Fastest way to extract a specific frame from a video (PHP/ffmpeg/anything)
    
    1 answer
    

  I extract thumbnails from a .MOV file using FFMPEG on Linus (Debian 64bit).

  The file I extract the thumbnail from is about 430 Megabytes large.

  I use the following command to do so :

  ffmpeg -i 'largeVideoFile.mov' -ss 00:14:37 -vframes 1 'thumbnail.jpg'

  It takes well over 3 minutes for a single frame to be extracted.

  How can I speed it up ?

• Ffmpeg decode H264 video from RTP stream very slow

  3 octobre 2013, par Dmitry Bakhtiyarov

  I have H264 RTP stream over network and I need to decode it. I use libx264 to encode stream and ffmpeg to decode. When I connect to server using VLC, it correct play video without any problem. But when I connect to server using my application I have a long period, when widget, which draw video from this stream, draw only one image.

  I check log file and found, that avcodec_decode_video2() set got_image into 1 very rarely ! Decoder give me new decoded frame average every 1-2 seconds, but on the server I have 12-15 fps on encoder !

  What the reasons of this delays on decoder and how I can fix its ?

  avcodec_register_all();
  
  av_init_packet(&m_packet);
  
  m_decoder = avcodec_find_decoder(CODEC_ID_H264);
  
  if (!m_decoder)
  
  {
  
      QString str = QString("Can't find H264 decoder!");
  
      emit criticalError(str);
  
  }
  
  m_decoderContext = avcodec_alloc_context3(m_decoder);
  
  
  
  m_decoderContext->flags |= CODEC_FLAG_LOW_DELAY;
  
  m_decoderContext->flags2 |= CODEC_FLAG2_CHUNKS;
  
  AVDictionary* dictionary = nullptr;
  
  if (avcodec_open2(m_decoderContext, m_decoder, &dictionary) < 0)
  
  {
  
      QString str = QString("Failed to open decoder!");
  
      emit criticalError(str);
  
  }
  
  qDebug() << "H264 Decoder successfully opened";
  
  m_picture = avcodec_alloc_frame();
  
  ...
  
      while(m_packet.size > 0)
  
      {
  
          int got_picture;
  
          int len = avcodec_decode_video2(m_decoderContext, m_picture, &got_picture, &m_packet);
  
          if (len < 0)
  
          {
  
              QString err("Decoding error");
  
              qDebug() << err;
  
              //emit criticalError(err);
  
              return false;
  
          }
  
          if (got_picture)
  
          {
  
                  //create from frame QImage and send it into GUI thread
  
              qDebug() << "H264Decoder: frame decoded!";

  I try to change some options of m_decoderContext (i.e. thread_count) but this not changing anything.