Recherche avancée

Sur d’autres sites (6439)

• ffmpeg Could not find input stream [closed]

  17 avril 2013, par Peter Walker

  I've seen lot's of post on streaming video feeds to ffserver but still haven't found what I wanted.

  I use not only on the mp4 file but avi and other formats as well.

      ffmpeg -i test5.mp4 http://localhost:8090/feed1.ffm

  I keep finding this error.

      Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'test5.mp4':
  
      Metadata:
  
             major_brand     : isom
  
             minor_version   : 512
  
             compatible_brands: isomiso2mp41
  
             encoder         : Lavf53.21.1
  
      Duration: 00:05:00.00, start: 0.000000, bitrate: 278 kb/s
  
      Stream #0.0(und): Video: mpeg4 (Simple Profile), yuv420p, 640x320 [PAR 1:1 DAR     2:1], 277 kb/s, 25 fps, 25 tbr, 25 tbn, 25 tbc
  
      Incompatible sample format '(null)' for codec 'mp2', auto-selecting format 's16'
  
      Last message repeated 1 times
  
      Output #0, ffm, to 'http://localhost:8090/feed1.ffm':
  
      Stream #0.0: Video: [0][0][0][0] / 0x0000, yuv420p, q=0-0, 1000k tbn
  
      Stream #0.1: Video: mpeg1video, (null), 32622x226393720, q=2-31, pass 1, pass 2,     226393 kb/s, 1000k tbn
  
      Stream #0.2: Audio: mp2, 22050 Hz, 1 channels, s16, 226391 kb/s
  
      Stream #0.3: Video: msmpeg4, yuv420p, 352x240, q=2-31, 226391 kb/s, 1000k tbn, 15  tbc
  
      Could not find input stream matching output stream #0.2
  
      *** glibc detected *** ffmpeg: free(): invalid pointer: 0x0000000000a53d00 ***
  
      ======= Backtrace: =========
  
      /lib/x86_64-linux-gnu/libc.so.6(+0x7e626)[0x7f6e0d4af626]
  
      /usr/lib/x86_64-linux-gnu/libavformat.so.53(avformat_free_context+0xb0)    [0x7f6e0f2c9c70]
  
      ffmpeg[0x4089bf]
  
      ffmpeg[0x40a52f]
  
      ffmpeg[0x407a04]
  
      /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xed)[0x7f6e0d45276d]

  The config file I used for the ffserver is

      Port 8090
  
      RTSPPort 7654
  
      BindAddress 0.0.0.0
  
      MaxHTTPConnections 2000
  
      MaxClients 1000
  
      MaxBandwidth 500000
  
      CustomLog -
  
      NoDaemon
  
      <feed>
  
              File /tmp/feed1.ffm
  
              FileMaxSize 5M
  
      #       NoAudio
  
      </feed>
  
  
  
      <stream>
  
              Feed feed1.ffm
  
              Format rtp
  
      #       NoAudio
  
              VideoFrameRate 30
  
      </stream>

  Can anyone help me figure out some something or something to look for ?
  I tried googling but almost everyone uses video feeds instead or it just works well for them.
  Any clue or direction/suggestion would help. Thank you so much in advance.

• Encoding a screenshot into a video using FFMPEG

  2 juillet 2013, par mohM

  I'm trying to get the pixels from the screen, and encode the screenshot into a video using ffmpeg. I've seen a couple of examples but they either assume you already have the pixel data, or use image file input. It seems like whether I use sws_scale() or not (which is included in the examples I've seen), or whether I'm typecasting a HBITMAP or RGBQUAD* it's telling me that the image src data is bad and is encoding a blank image rather than the screenshot. Is there something I'm missing here ?

  AVCodec* codec;
  
  AVCodecContext* c = NULL;
  
  AVFrame* inpic;
  
  uint8_t* outbuf, *picture_buf;
  
  int i, out_size, size, outbuf_size;
  
  HBITMAP hBmp;
  
  //int x,y;
  
  
  
  avcodec_register_all();
  
  
  
  printf("Video encoding\n");
  
  
  
  // Find the mpeg1 video encoder
  
  codec = avcodec_find_encoder(CODEC_ID_H264);
  
  if (!codec) {
  
      fprintf(stderr, "Codec not found\n");
  
      exit(1);
  
  }
  
  else printf("H264 codec found\n");
  
  
  
  c = avcodec_alloc_context3(codec);
  
  inpic = avcodec_alloc_frame();
  
  
  
  c->bit_rate = 400000;
  
  c->width = screenWidth;                                     // resolution must be a multiple of two
  
  c->height = screenHeight;
  
  c->time_base.num = 1;
  
  c->time_base.den = 25;
  
  c->gop_size = 10;                                           // emit one intra frame every ten frames
  
  c->max_b_frames=1;
  
  c->pix_fmt = PIX_FMT_YUV420P;
  
  c->codec_id = CODEC_ID_H264;
  
  //c->codec_type = AVMEDIA_TYPE_VIDEO;
  
  
  
  //av_opt_set(c->priv_data, "preset", "slow", 0);
  
  //printf("Setting presets to slow for performance\n");
  
  
  
  // Open the encoder
  
  if (avcodec_open2(c, codec,NULL) < 0) {
  
      fprintf(stderr, "Could not open codec\n");
  
      exit(1);
  
  }
  
  else printf("H264 codec opened\n");
  
  
  
  outbuf_size = 100000 + 12*c->width*c->height;           // alloc image and output buffer
  
  //outbuf_size = 100000;
  
  outbuf = static_cast(malloc(outbuf_size));
  
  size = c->width * c->height;
  
  picture_buf = static_cast(malloc((size*3)/2));
  
  printf("Setting buffer size to: %d\n",outbuf_size);
  
  
  
  FILE* f = fopen("example.mpg","wb");
  
  if(!f) printf("x  -  Cannot open video file for writing\n");
  
  else printf("Opened video file for writing\n");
  
  
  
  /*inpic->data[0] = picture_buf;
  
  inpic->data[1] = inpic->data[0] + size;
  
  inpic->data[2] = inpic->data[1] + size / 4;
  
  inpic->linesize[0] = c->width;
  
  inpic->linesize[1] = c->width / 2;
  
  inpic->linesize[2] = c->width / 2;*/
  
  
  
  
  
  //int x,y;
  
  // encode 1 second of video
  
  for(i=0;itime_base.den;i++) {
  
      fflush(stdout);
  
  
  
  
  
      HWND hDesktopWnd = GetDesktopWindow();
  
      HDC hDesktopDC = GetDC(hDesktopWnd);
  
      HDC hCaptureDC = CreateCompatibleDC(hDesktopDC);
  
      hBmp = CreateCompatibleBitmap(GetDC(0), screenWidth, screenHeight);
  
      SelectObject(hCaptureDC, hBmp);
  
      BitBlt(hCaptureDC, 0, 0, screenWidth, screenHeight, hDesktopDC, 0, 0, SRCCOPY|CAPTUREBLT);
  
      BITMAPINFO bmi = {0}; 
  
      bmi.bmiHeader.biSize = sizeof(bmi.bmiHeader); 
  
      bmi.bmiHeader.biWidth = screenWidth; 
  
      bmi.bmiHeader.biHeight = screenHeight; 
  
      bmi.bmiHeader.biPlanes = 1; 
  
      bmi.bmiHeader.biBitCount = 32; 
  
      bmi.bmiHeader.biCompression = BI_RGB;
  
      RGBQUAD *pPixels = new RGBQUAD[screenWidth*screenHeight];
  
      GetDIBits(hCaptureDC,hBmp,0,screenHeight,pPixels,&bmi,DIB_RGB_COLORS);
  
  
  
  inpic->pts = (float) i * (1000.0/(float)(c->time_base.den))*90;
  
      avpicture_fill((AVPicture*)inpic, (uint8_t*)pPixels, PIX_FMT_BGR32, c->width, c->height);                   // Fill picture with image
  
      av_image_alloc(inpic->data, inpic->linesize, c->width, c->height, c->pix_fmt, 1);
  
      //printf("Allocated frame\n");
  
      //SaveBMPFile(L"screenshot.bmp",hBmp,hDc,screenWidth,screenHeight);
  
      ReleaseDC(hDesktopWnd,hDesktopDC);
  
      DeleteDC(hCaptureDC);
  
      DeleteObject(hBmp);
  
  
  
      // encode the image
  
      out_size = avcodec_encode_video(c, outbuf, outbuf_size, inpic);
  
      printf("Encoding frame %3d (size=%5d)\n", i, out_size);
  
      fwrite(outbuf, 1, out_size, f);
  
  }
  
  
  
  // get the delayed frames
  
  for(; out_size; i++) {
  
      fflush(stdout);
  
  
  
      out_size = avcodec_encode_video(c, outbuf, outbuf_size, NULL);
  
      printf("Writing frame %3d (size=%5d)\n", i, out_size);
  
      fwrite(outbuf, 1, out_size, f);
  
  }
  
  
  
  // add sequence end code to have a real mpeg file
  
  outbuf[0] = 0x00;
  
  outbuf[1] = 0x00;
  
  outbuf[2] = 0x01;
  
  outbuf[3] = 0xb7;
  
  fwrite(outbuf, 1, 4, f);
  
  fclose(f);
  
  free(picture_buf);
  
  free(outbuf);
  
  
  
  avcodec_close(c);
  
  av_free(c);
  
  av_free(inpic);
  
  printf("Closed codec and Freed\n");

• ffmpeg concatenate images in one image

  26 juillet 2016, par drlexa

  I use this to get frames from video and concatenate them in one image :

  ffmpeg -i output.mp4 -vf 'fps=2,tile=1000x1' out.jpg

  But there is a problem : I do not know number of frames that will be fetched. Here I hardcoded tile size 1000x1, but if there will be more than 1000 frames, then will be an error. Before starting ffmpeg I do not know actual size of tile.

  So I want use command like :

  ffmpeg -i output.mp4 -vf 'fps=2,tile=*x1' out.jpg

  That means : I want you to concatenate ALL images that will be fetched in one row, but I cannot use * as an argument for tile.

  Is there some way to solve my problem ?