Recherche avancée

Sur d’autres sites (6565)

• Xuggler Encoding video of Desktop With Audio - audio has gaps

  2 novembre 2012, par Chris

  I am using Xuggler to convert images captured from the java Robot class and sound read from TargetDataLine class and encoding this into a video. I am then attempting to http stream this video data (after writing my header) to a flash client via http (Socket OutputStream) but it plays and stutters (never just playing smoothly) no matter what buffer value I use on the client side.

  I am asking for help and showing my java code because I suspect it might be to do with how I am encoding the video or something about sending data via http socket which i am not getting..

  ByteArrayURLHandler ba = new ByteArrayURLHandler();
  
  final IRational FRAME_RATE = IRational.make(30); 
  
  final int SECONDS_TO_RUN_FOR = 20; 
  
  final Robot robot = new Robot(); 
  
  final Toolkit toolkit = Toolkit.getDefaultToolkit(); 
  
  final Rectangle screenBounds = new Rectangle(toolkit.getScreenSize()); 
  
  IMediaWriter writer;
  
  
  
  writer = ToolFactory.makeWriter(
  
      XugglerIO.map( 
  
          XugglerIO.generateUniqueName(out, ".flv"), 
  
          out 
  
      )); 
  
  
  
  writer.addListener(new MediaListenerAdapter() { 
  
      public void onAddStream(IAddStreamEvent event) { 
  
          event.getSource().getContainer().setInputBufferLength(1000);
  
          IStreamCoder coder = event.getSource().getContainer().getStream(event.getStreamIndex()).getStreamCoder();
  
          if (coder.getCodecType() == ICodec.Type.CODEC_TYPE_AUDIO) { 
  
              coder.setFlag(IStreamCoder.Flags.FLAG_QSCALE, false);   
  
              coder.setBitRate(32000); 
  
              System.out.println("onaddstream"+ coder.getPropertyNames().toString());
  
          } 
  
          if (coder.getCodecType() == ICodec.Type.CODEC_TYPE_VIDEO) { 
  
              // coder.setBitRate(64000); 
  
              // coder.setBitRateTolerance(64000); 
  
          } 
  
      } 
  
  });
  
  
  
  writer.addVideoStream(videoStreamIndex, videoStreamId, 1024, 768);
  
  final int channelCount = 1;       
  
  
  
  int audionumber =   writer.addAudioStream(audioStreamIndex, audioStreamId,1, 44100);
  
  int bufferSize = (int)audioFormat.getSampleRate()   *audioFormat.getFrameSize();//*6;///6;
  
  byte[] audioBuf;// = new byte[bufferSize]; 
  
  
  
  int i = 0;
  
  
  
  final int audioStreamIndex = 1;
  
  final int audioStreamId = 1;
  
  BufferedImage screen, bgrScreen;
  
  long startTime = System.nanoTime();
  
  while(keepGoing)
  
  { 
  
  
  
      audioBuf = new byte[bufferSize]; 
  
      i++;
  
  
  
      screen = robot.createScreenCapture(screenBounds); 
  
  
  
      bgrScreen = convertToType(screen, BufferedImage.TYPE_3BYTE_BGR); 
  
      long nanoTs = System.nanoTime()-startTime; 
  
      writer.encodeVideo(0, bgrScreen, nanoTs, TimeUnit.NANOSECONDS);
  
      audioBuf = new byte[line.available()];
  
      int nBytesRead = line.read(audioBuf, 0, audioBuf.length); 
  
  
  
      IBuffer iBuf = IBuffer.make(null, audioBuf, 0, nBytesRead);
  
  
  
      IAudioSamples smp = IAudioSamples.make(iBuf,1,IAudioSamples.Format.FMT_S16);
  
      if (smp == null) {
  
          return;
  
      }
  
  
  
      long numSample = audioBuf.length / smp.getSampleSize();
  
  
  
      smp.setComplete(true, numSample,(int)
  
      audioFormat.getSampleRate(), audioFormat.getChannels(),
  
      IAudioSamples.Format.FMT_S16, nanoTs/1000);
  
  
  
      writer.encodeAudio(1, smp);
  
  
  
      writer.flush();
  
  }

• Streaming full HD using lighttpd streaming

  2 juin 2015, par HD Quality

  We have 1000 of HD full length movies and streaming using lighttpd all the files are converted using FFMPEG .FLV format, and we got VPS(Cent OS) 2TB storage & CPU : 6 cores & 30 GB RAM & 1 Gbit/s port.

  when we get traffic player buffering very long time means 10min to start play,i spend searching google and tried all the possible way to improve but no luck.

  can any one suggest how other sites are running without buffer issue,

  1.http://cloudy.ec

  3.http://putloader.com

• FFmpeg H264 parsing

  20 mai 2015, par Yoohoo

  I am using FFmpeg to decoding H.264 video, for the following program :

      #include 
  
      #include 
  
      #include 
  
      #include 
  
  
  
      #include <sys></sys>time.h>
  
      #include 
  
  
  
  #include <libavcodec></libavcodec>avcodec.h>
  
  #include <libavutil></libavutil>mathematics.h>
  
  
  
  #include <sdl></sdl>SDL.h>
  
  
  
  void sigint_handler(int signal) {
  
      printf("\n");
  
      exit(0);
  
  }
  
  
  
  const char *window_title;
  
  SDL_Surface *screen;
  
  SDL_Overlay *yuv_overlay;
  
  
  
  #define INBUF_SIZE 80000
  
  
  
  /*
  
   * Video decoding example
  
   */
  
  
  
  static long get_time_diff(struct timeval time_now) {
  
     struct timeval time_now2;
  
     gettimeofday(&amp;time_now2,0);
  
     return time_now2.tv_sec*1.e6 - time_now.tv_sec*1.e6 + time_now2.tv_usec - time_now.tv_usec;
  
  }
  
  
  
  int video_open(AVCodecContext *avctx, const char *filename){
  
      int flags = SDL_HWSURFACE|SDL_ASYNCBLIT|SDL_HWACCEL;
  
      int w,h;
  
  
  
      flags |= SDL_RESIZABLE;
  
  
  
      if (avctx->width){
  
          w = avctx->width;
  
          h = avctx->height;
  
      } else {
  
          w = 640;
  
          h = 480;
  
      }
  
  
  
      if(SDL_Init(SDL_INIT_VIDEO) &lt; 0) {
  
      fprintf(stderr, "SDL_INIT_VIDEO failed!\n");
  
      exit(1);
  
      }
  
  
  
      screen = SDL_SetVideoMode(w, h, 0, flags);
  
  
  
      if (!screen) {
  
          fprintf(stderr, "SDL: could not set video mode - exiting\n");
  
          return -1;
  
      }
  
      if (!window_title)
  
          window_title = filename;
  
      SDL_WM_SetCaption(window_title, window_title);
  
  
  
      yuv_overlay = SDL_CreateYUVOverlay(w, h, SDL_YV12_OVERLAY, screen);
  
  
  
      if (yuv_overlay->hw_overlay) {
  
      fprintf(stderr, "Using hardware overlay!\n");
  
      }
  
  
  
      return 0;
  
  }
  
  
  
  int main(int argc, char **argv) {
  
      AVCodec *codec;
  
      AVCodecContext *c= NULL;
  
      AVCodecParserContext *parser = NULL;
  
      int frame, got_picture, len2, len;
  
      const char *filename;
  
      FILE *f;
  
      AVFrame *picture;
  
      char *arghwtf = malloc(INBUF_SIZE);
  
      char *luma = NULL;
  
      char *chroma = NULL;
  
      int i=0;
  
      uint64_t in_len;
  
      int pts, dts;
  
      struct timeval t;
  
      float inv_fps = 1e6/23.98;
  
      AVPacket avpkt;
  
      SDL_Rect rect;
  
  
  
  
  
      /* register all the codecs */
  
      avcodec_register_all();
  
  
  
      filename = argv[1];
  
  
  
      av_init_packet(&amp;avpkt);
  
  
  
      printf("Decoding file %s...\n", filename);
  
  
  
      /* find the H.264 video decoder */
  
      codec = avcodec_find_decoder(CODEC_ID_H264);
  
      if (!codec) {
  
          fprintf(stderr, "codec not found\n");
  
          exit(1);
  
      }
  
  
  
      c = avcodec_alloc_context3(codec);
  
      picture = avcodec_alloc_frame();
  
  
  
      c->skip_loop_filter = 48; // skiploopfilter=all
  
  
  
      if (avcodec_open(c, codec) &lt; 0) {
  
          fprintf(stderr, "could not open codec\n");
  
          exit(1);
  
      }
  
  
  
      /* the codec gives us the frame size, in samples */
  
      parser = av_parser_init(c->codec_id);
  
      parser->flags |= PARSER_FLAG_ONCE;
  
  
  
      f = fopen(filename, "rb");
  
      if (!f) {
  
          fprintf(stderr, "could not open %s\n", filename);
  
          exit(1);
  
      }
  
  
  
      frame = 0;
  
      gettimeofday(&amp;t, 0);
  
      if(fread(arghwtf, 1, INBUF_SIZE, f) == 0) {
  
      exit(1);
  
      }
  
      in_len = 80000;
  
          while (in_len > 0 &amp;&amp; !feof(f)) {
  
          len = av_parser_parse2(parser, c, &amp;avpkt.data, &amp;avpkt.size, arghwtf, in_len,
  
                                     pts, dts, AV_NOPTS_VALUE);
  
  
  
              len2 = avcodec_decode_video2(c, picture, &amp;got_picture, &amp;avpkt);
  
              if (len2 &lt; 0) {
  
                  fprintf(stderr, "Error while decoding frame %d\n", frame);
  
                  exit(1);
  
              }
  
              if (got_picture) {
  
          if(!screen) {
  
              video_open(c, filename);
  
  
  
              rect.x = 0;
  
              rect.y = 0;
  
              rect.w = c->width;
  
              rect.h = c->height;
  
              inv_fps = av_q2d(c->time_base);
  
              fprintf(stderr, "w:%i h:%i\n", rect.w, rect.h);
  
  
  
              luma = malloc(c->width*c->height);
  
              chroma = malloc(c->width*c->height/4);
  
  
  
              SDL_DisplayYUVOverlay(yuv_overlay, &amp;rect);
  
  
  
              signal(SIGINT, sigint_handler);
  
          }
  
                  fprintf(stderr, "\rDisplaying %c:frame %3d (%02d:%02d)...", av_get_pict_type_char(picture->pict_type), frame, frame/1440, (frame/24)%60);
  
                  fflush(stderr);
  
  
  
          SDL_LockYUVOverlay(yuv_overlay);
  
  
  
                  for(i=0;iheight;i++) {
  
                    memcpy(luma + i * c->width, picture->data[0] + i * picture->linesize[0], c->width);
  
                  }
  
          memcpy(yuv_overlay->pixels[0], luma, c->width * c->height);
  
                  for(i=0;iheight/2;i++) {
  
                    memcpy(chroma + i * c->width/2, picture->data[2] + i * picture->linesize[2], c->width/2);
  
                  }
  
          memcpy(yuv_overlay->pixels[1], chroma, c->width * c->height / 4);
  
                  for(i=0;iheight/2;i++) {
  
                    memcpy(chroma + i * c->width/2, picture->data[1] + i * picture->linesize[1], c->width/2);
  
                  }
  
          memcpy(yuv_overlay->pixels[2], chroma, c->width * c->height / 4);
  
  
  
          SDL_UnlockYUVOverlay(yuv_overlay);
  
          SDL_DisplayYUVOverlay(yuv_overlay, &amp;rect);
  
  
  
          while(get_time_diff(t) &lt; inv_fps) {
  
              sleep(1000);
  
          }
  
                  frame++;
  
          gettimeofday(&amp;t, 0);
  
              }
  
          memcpy(arghwtf, arghwtf + len, 80000-len);
  
          fread(arghwtf + 80000 - len, 1, len, f);
  
          }
  
  
  
      /* some codecs, such as MPEG, transmit the I and P frame with a
  
         latency of one frame. You must do the following to have a
  
         chance to get the last frame of the video */
  
      avpkt.data = NULL;
  
      avpkt.size = 0;
  
      len = avcodec_decode_video2(c, picture, &amp;got_picture, &amp;avpkt);
  
      if (got_picture) {
  
          printf("saving last frame %3d\n", frame);
  
          fflush(stdout);
  
  
  
      /* Display last frame here, same code as in the decoding loop above. */
  
  
  
          frame++;
  
      }
  
  
  
      fclose(f);
  
  
  
      avcodec_close(c);
  
      av_free(c);
  
      av_free(picture);
  
      printf("\n");
  
  }

  It gives error :

  ||=== H264_decoding, Debug ===|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp||In function ‘int main(int, char**)’:|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|107|error: invalid conversion from ‘void*’ to ‘char*’ [-fpermissive]|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|136|warning: ‘AVFrame* avcodec_alloc_frame()’ is deprecated (declared at /home/yoohoo/ffmpeg_build/include/libavcodec/avcodec.h:3629) [-Wdeprecated-declarations]|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|136|warning: ‘AVFrame* avcodec_alloc_frame()’ is deprecated (declared at /home/yoohoo/ffmpeg_build/include/libavcodec/avcodec.h:3629) [-Wdeprecated-declarations]|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|138|error: invalid conversion from ‘int’ to ‘AVDiscard’ [-fpermissive]|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|140|error: ‘avcodec_open’ was not declared in this scope|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|163|error: invalid conversion from ‘char*’ to ‘const uint8_t* {aka const unsigned char*}’ [-fpermissive]|
  
  /home/yoohoo/ffmpeg_build/include/libavcodec/avcodec.h|4465|error:   initializing argument 5 of ‘int av_parser_parse2(AVCodecParserContext*, AVCodecContext*, uint8_t**, int*, const uint8_t*, int, int64_t, int64_t, int64_t)’ [-fpermissive]|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|181|error: invalid conversion from ‘void*’ to ‘char*’ [-fpermissive]|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|182|error: invalid conversion from ‘void*’ to ‘char*’ [-fpermissive]|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|188|error: ‘av_get_pict_type_char’ was not declared in this scope|
  
  /home/yoohoo/codeblocks/Real-time video streaming/H264_decoding/main.cpp|210|error: ‘sleep’ was not declared in this scope|
  
  ||=== Build finished: 9 errors, 2 warnings ===|

  I am wondering it is an incompatible problem due to I move this C program to C++, but I finally need to use it in C++. How to solve these invalid conversion problem ? Thanks in advance !