Recherche avancée

Sur d’autres sites (14816)

• FFMPEG to OpenGL Texture

  23 avril 2014, par Spamdark

  I was here to ask, how can I convert an AVFrame to an opengl texture. Actually, I created a renderer the outputs me the audio (Audio is working) and the video, but the video is not outputing. Here is my code :

  Texture creation :

  glGenTextures(1,&_texture);
  
  glBindTexture(GL_TEXTURE_2D,_texture);
  
  glPixelStorei(GL_UNPACK_ALIGNMENT, 1);
  
  
  
  glTexParameteri( GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR ); 
  
  glTexParameteri( GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR );

  Code Info : _texture variable is a GLuint that keeps the texture ID

  Function that gets the AVFrame and convert it to OpenGL Texture :

  int VideoGL::NextVideoFrame(){
  
  // Get a packet from the queue
  
  AVPacket *videopacket = this->DEQUEUE(VIDEO);
  
  int frameFinished;
  
  if(videopacket!=0){
  
      avcodec_decode_video2(_codec_context_video, _std_frame,&frameFinished,videopacket);
  
  
  
      if(frameFinished){
  
  
  
          sws_scale(sws_ctx, _std_frame->data, _std_frame->linesize, 0, _codec_context_video->height, _rgb_frame->data, _rgb_frame->linesize);
  
  
  
          if(_firstrendering){
  
          glBindTexture(GL_TEXTURE_2D,_texture);
  
          glTexImage2D(GL_TEXTURE_2D, 0, GL_RGB, _codec_context_video->width,_codec_context_video->height,0,GL_RGB,GL_UNSIGNED_BYTE,_rgb_frame->data[0]);
  
  
  
          _firstrendering = false;
  
  
  
          }else{
  
  
  
              glActiveTexture(_texture);
  
              glBindTexture(GL_TEXTURE_2D,_texture);
  
              glTexSubImage2D(GL_TEXTURE_2D,0,0,0,_codec_context_video->width,_codec_context_video->height,GL_RGB,GL_UNSIGNED_BYTE,_rgb_frame->data[0]);
  
  
  
          }
  
          av_free_packet(videopacket);
  
          return 0;
  
      }else{
  
  
  
          av_free_packet(videopacket);
  
          return -1;
  
      }
  
  
  
  }else{
  
      return -1;
  
  }
  
  return 0;
  
  }

  Code Information : There is a queue where a thread store the AVFrames, this function is frequently called to get the AVFrames, until it gets a NULL it stops to being called.

  That’s actually not working. (I tried to look at some questions in stack overflow, it’s still not working)
  Any example, or someone that helps me to correct any error there ?

  Additional Data : I tried to change the GL_RGB to GL_RGBA and started to play with the formats, anyway it crashes when I try GL_RGBA (Because the width and height are very big, anyway I tried to resize them). I have tried to change the sizes to Power Of 2, stills not working.

  1 Edit :

  Thread function :

  DWORD WINAPI VideoGL::VidThread(LPVOID myparam){
  
  
  
  VideoGL * instance = (VideoGL*) myparam;
  
  instance->wave_audio->Start();
  
  
  
  int quantity=0;
  
  
  
  AVPacket packet;
  
  while(av_read_frame(instance->_format_context,&packet) >= 0){
  
      if(packet.stream_index==instance->videoStream){
  
          instance->ENQUEUE(VIDEO,&packet);
  
      }
  
      if(packet.stream_index==instance->audioStream){
  
          instance->ENQUEUE(AUDIO,&packet);
  
      }
  
  }
  
  
  
  instance->ENQUEUE(AUDIO,NULL);
  
  instance->ENQUEUE(VIDEO,NULL);
  
  
  
  return 0;
  
  }

  Thread creation function :

  CreateThread(NULL, 0, VidThread, this, NULL, NULL);

  Where this refers to the class that contains the NextVideoFrame, and the _texture members.

  Solved :

  I followed some of the datenwolf tips, and now the video is displaying correctly with the audio/video :

  

• ffmpeg Sound going out of sync with -concat or -ss

  6 août 2012, par Jared Glass

  I have a tool that spits out video from a 3D application and then concats the individual videos to make a sequence. But the sound seems to go out of sync in the sequence (the inividual files are fine) and it stutters in VLC and Quicktime. Windows media player seems to handle it bes to my supprise, yet it still goes out of sync. I have two senarios, one works and one doesn't but i need both working :

  Working :
  get already created out movs...

  convert to avi :
  os.system( ffmpeg + " -i C:\clip.mov -sameq -r 24 -y C:\clip.avi")

  concat to avi sequence :
  os.system( ffmpeg + ''' -i concat: C:\clip.avi|C:\clip1.avi|C:\clip2.avi -sameq -r 24 -y C:\sequence.avi''' )

  convert sequence to mov :
  os.system( ffmpeg + " -i C:\sequence.avi -sameq -r 24 -y C:\sequence.mov")

  Not Working :
  create individual avi's from 3D program...

  cut down to correct length :
  os.system(ffmpeg + " -i C:\clip.avi -sameq -r 24 -ss " + startTime + " -vframes " + totalFrames + " -y C:\clip.avi" )

  concat to avi sequence :
  os.system( ffmpeg + ''' -i concat: C:\clip.avi|C:\clip1.avi|C:\clip2.avi -sameq -r 24 -y C:\sequence.avi''' )

  convert sequence to mov :
  os.system( ffmpeg + " -i C:\sequence.avi -sameq -r 24 -y C:\sequence.mov")

  convert individual avi's to mov :
  os.system( ffmpeg + " -i C:\clip.avi-sameq -r 24 -y C:\clip.mov")

  Please let me know where I've gone wrong ?

• Faster frame extraction than ffmpeg

  23 janvier 2013, par Ben Ford

  I have a video analytics program that processes assorted frames from a video. (Several hours long)
  The video is likely going to be an MP4 but may be other formats going forwards.

  At the moment, I have a C# wrapper around an ffmpeg call to extract an individual frame at the requested time. (I'm using the ffmpeg.exe binary. Not the libraries directly)
  At the moment, this all works. But it's slow. Very slow.

  I've found ways to improve the speed by storing the extracted frames in a ramdisk while they're being processed. Changing the stored image format etc...

  I just wanted to check if anyone could think of any way to pull individual frames out. At split-second accuracy.
  I know this is probably possible with DShow etc... I went straight to FFMPEG as I've used it before. But if DShow is likely to be faster I'll gladly change !