Recherche avancée

Sur d’autres sites (10935)

• retriving video details with ffprobe in php

  15 juillet 2014, par Roohbakhsh Masoud

  i’m using this command for retrieve video information in ubuntu terminal :

  ffprobe -show_streams sample.mp4

  this command show all information about video in terminal, but when i use this command in php and use exec function to retrieve result , return

  string(0) ""
  
  {}

  how i can retrieve result in php ?

  my code :

      public function information($ffmpeg,$file)
  
      {
  
        $info = exec("$ffmpeg -show_streams $file");
  
        return $info;
  
      }

• FFMPEG Drag and Drop sh file

  29 avril 2021, par WFCDefi

  On Windows I have a simple batch file which I drop video files onto to convert to webms, it saves a lot of time as I prefer to just use the same configuration and don't care much about the names.

  


  @echo off 
echo. 
ffmpeg -i %1 -c:v libvpx-vp9 -quality good -cpu-used 2 -b:v 5000k -qmin 15 -qmax 45 -maxrate 500k -bufsize 1500k -framerate 60 -threads 8 -vf scale=-1:1080 -c:a libvorbis -b:a 192k -f webm %1.webm 
pause


  


  I know the .bat file won't really work in Linux (I'm on pop os so pretty much Ubuntu) so with the other lines removed and the %1 changed to $1 it works. It won't do anything if I try dragging and dropping a video file onto it though.

  


  I can type sudo then drag and drop the .sh file followed by a video into a terminal and press enter and it will have the same effect as dragging a video file onto a bat file in Windows.

  


  Is there a way to recreate dragging and dropping a file directly onto another and it executing in Linux or is the terminal the only way ?

  


• Memory leak when using ffmpeg

  21 janvier 2017, par se210

  I have implemented a class which spawns a thread for reading and queuing frames, and the main thread displays these frames via OpenGL. I try to free the allocated memory after binding the image data to a OpenGL texture, but it seems some memory is not freed properly. The memory usage keeps growing until the system runs out of memory and eventually the frame reader thread cannot grab new frames due to memory allocation failure. Would someone please help me on what I might have missed ? Thank you.

  This is the code for the frame reader thread :

  void AVIReader::frameReaderThreadFunc()
  
  {
  
      AVPacket packet;
  
  
  
      while (readFrames) {
  
          // Allocate necessary memory
  
          AVFrame* pFrame = av_frame_alloc();
  
          if (pFrame == nullptr)
  
          {
  
              continue;
  
          }
  
  
  
          AVFrame* pFrameRGB = av_frame_alloc();
  
          if (pFrameRGB == nullptr)
  
          {
  
              av_frame_free(&pFrame);
  
              continue;
  
          }
  
  
  
          // Determine required buffer size and allocate buffer
  
          int numBytes = avpicture_get_size(AV_PIX_FMT_RGB24, pCodecCtx->width,
  
              pCodecCtx->height);
  
          uint8_t* buffer = (uint8_t *)av_malloc(numBytes * sizeof(uint8_t));
  
  
  
          if (buffer == nullptr)
  
          {
  
              av_frame_free(&pFrame);
  
              av_frame_free(&pFrameRGB);
  
              continue;
  
          }
  
  
  
          // Assign appropriate parts of buffer to image planes in pFrameRGB
  
          // Note that pFrameRGB is an AVFrame, but AVFrame is a superset
  
          // of AVPicture
  
          avpicture_fill((AVPicture *)pFrameRGB, buffer, AV_PIX_FMT_RGB24,
  
              pCodecCtx->width, pCodecCtx->height);
  
  
  
          if (av_read_frame(pFormatCtx, &packet) >= 0) {
  
              // Is this a packet from the video stream?
  
              if (packet.stream_index == videoStream) {
  
                  // Decode video frame
  
                  int frameFinished;
  
                  avcodec_decode_video2(pCodecCtx, pFrame, &frameFinished, &packet);
  
  
  
                  if (frameFinished) {
  
                      // Convert the image from its native format to RGB
  
                      sws_scale(sws_ctx, (uint8_t const * const *)pFrame->data,
  
                          pFrame->linesize, 0, pCodecCtx->height,
  
                          pFrameRGB->data, pFrameRGB->linesize);
  
  
  
                      VideoFrame vf;
  
                      vf.frame = pFrameRGB;
  
                      vf.pts = av_frame_get_best_effort_timestamp(pFrame) * time_base;
  
                      frameQueue.enqueue(vf);
  
  
  
                      av_frame_unref(pFrame);
  
                      av_frame_free(&pFrame);
  
                  }
  
              }
  
              //av_packet_unref(&packet);
  
              av_free_packet(&packet);
  
          }
  
      }
  
  }

  This is the code that grabs the queued frames and binds it to an OpenGL texture. I explicitly save the previous frame until I switch it out with the next frame. Otherwise, it seems to cause a segfault.

  void AVIReader::GrabAVIFrame()
  
  {
  
      if (curFrame.pts >= clock_pts)
  
      {
  
          return;
  
      }
  
  
  
      if (frameQueue.empty())
  
          return;
  
  
  
      // Get a packet from the queue
  
      VideoFrame videoFrame = frameQueue.top();
  
      while (!frameQueue.empty() && frameQueue.top().pts < clock_pts)
  
      {
  
          videoFrame = frameQueue.dequeue();
  
      }
  
  
  
      glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, width, height, GL_RGB, GL_UNSIGNED_BYTE, videoFrame.frame->data[0]);
  
  
  
      // release previous frame
  
      if (curFrame.frame)
  
      {
  
          av_free(curFrame.frame->data[0]);
  
      }
  
      av_frame_unref(curFrame.frame);
  
  
  
      // set current frame to new frame
  
      curFrame = videoFrame;
  
  }

  The frameQueue is a thread-safe priority queue that holds VideoFrame defined as :

  class VideoFrame {
  
  public:
  
      AVFrame* frame;
  
      double pts;
  
  };

  Update : There was a silly error in the ordering of setting current frame to new frame. I forgot to switch it back after trying some things out. I also incorporated @ivan_onys’s suggestion, but that does not seem to fix the problem.

  ---

  Update 2 : I adopted @Al Bundy’s suggestion to release pFrame and packet unconditionally, but the issue still persists.

  Since buffer is what contains the actual image data which needs to be used in glTexSubImage2D(), I cannot release it until I am done displaying it on the screen (otherwise I get a segfault). avpicture_fill() assigns frame->data[0] = buffer, so I think calling av_free(curFrame.frame->data[0]) ; on the previous frame after texture mapping the new frame should release the allocated buffer.

  Here is the updated frame reader thread code :

  void AVIReader::frameReaderThreadFunc()
  
  {
  
      AVPacket packet;
  
  
  
      while (readFrames) {
  
          // Allocate necessary memory
  
          AVFrame* pFrame = av_frame_alloc();
  
          if (pFrame == nullptr)
  
          {
  
              continue;
  
          }
  
  
  
          AVFrame* pFrameRGB = av_frame_alloc();
  
          if (pFrameRGB == nullptr)
  
          {
  
              av_frame_free(&pFrame);
  
              continue;
  
          }
  
  
  
          // Determine required buffer size and allocate buffer
  
          int numBytes = avpicture_get_size(AV_PIX_FMT_RGB24, pCodecCtx->width,
  
              pCodecCtx->height);
  
          uint8_t* buffer = (uint8_t *)av_malloc(numBytes * sizeof(uint8_t));
  
  
  
          if (buffer == nullptr)
  
          {
  
              av_frame_free(&pFrame);
  
              av_frame_free(&pFrameRGB);
  
              continue;
  
          }
  
  
  
          // Assign appropriate parts of buffer to image planes in pFrameRGB
  
          // Note that pFrameRGB is an AVFrame, but AVFrame is a superset
  
          // of AVPicture
  
          avpicture_fill((AVPicture *)pFrameRGB, buffer, AV_PIX_FMT_RGB24,
  
              pCodecCtx->width, pCodecCtx->height);
  
  
  
          if (av_read_frame(pFormatCtx, &packet) >= 0) {
  
              // Is this a packet from the video stream?
  
              if (packet.stream_index == videoStream) {
  
                  // Decode video frame
  
                  int frameFinished;
  
                  avcodec_decode_video2(pCodecCtx, pFrame, &frameFinished, &packet);
  
  
  
                  if (frameFinished) {
  
                      // Convert the image from its native format to RGB
  
                      sws_scale(sws_ctx, (uint8_t const * const *)pFrame->data,
  
                          pFrame->linesize, 0, pCodecCtx->height,
  
                          pFrameRGB->data, pFrameRGB->linesize);
  
  
  
                      VideoFrame vf;
  
                      vf.frame = pFrameRGB;
  
                      vf.pts = av_frame_get_best_effort_timestamp(pFrame) * time_base;
  
                      frameQueue.enqueue(vf);
  
                  }
  
              }
  
          }
  
          av_frame_unref(pFrame);
  
          av_frame_free(&pFrame);
  
          av_packet_unref(&packet);
  
          av_free_packet(&packet);
  
      }
  
  }

  ---

  Solved : It turned out the leaks were happening when the packet was from a non-video stream (e.g. audio). I also needed to free resources on frames that are skipped in the while-loop of GrabAVIFrame().