Recherche avancée

Sur d’autres sites (8239)

• Turn off sw_scale conversion to planar YUV 32 byte alignment requirements

  8 novembre 2022, par flansel

  I am experiencing artifacts on the right edge of scaled and converted images when converting into planar YUV pixel formats with sw_scale. I am reasonably sure (although I can not find it anywhere in the documentation) that this is because sw_scale is using an optimization for 32 byte aligned lines, in the destination. However I would like to turn this off because I am using sw_scale for image composition, so even though the destination lines may be 32 byte aligned, the output image may not be.

  


  Example.

  


  Full output frame is 1280x720 yuv422p10le. (this is 32 byte aligned)
However into the top left corner I am scaling an image with an outwidth of 1280 / 3 = 426.
426 in this format is not 32 byte aligned, but I believe sw_scale sees that the output linesize is 32 byte aligned and overwrites the width of 426 putting garbage in the next 22 bytes of data thinking this is simply padding when in my case this is displayable area.

  


  This is why I need to actually disable this optimization or somehow trick sw_scale into believing it does not apply while keeping intact the way the program works, which is otherwise fine.

  


  I have tried adding extra padding to the destination lines so they are no longer 32 byte aligned,
this did not help as far as I can tell.

  


  Edit with code Example. Rendering omitted for ease of use.
Also here is a similar issue, unfortunately as I stated there fix will not work for my use case. https://github.com/obsproject/obs-studio/pull/2836

  


  Use the commented line of code to swap between a output width which is and isnt 32 byte aligned.

  


  #include "libswscale/swscale.h"
#include "libavutil/imgutils.h"
#include "libavutil/pixelutils.h"
#include "libavutil/pixfmt.h"
#include "libavutil/pixdesc.h"
#include 
#include 
#include 

int main(int argc, char **argv) {

/// Set up a 1280x720 window, and an item with 1/3 width and height of the window.
int window_width, window_height, item_width, item_height;
window_width = 1280;
window_height = 720;
item_width = (window_width / 3);
item_height = (window_height / 3);

int item_out_width = item_width;
/// This line sets the item width to be 32 byte aligned uncomment to see uncorrupted results
/// Note %16 because outformat is 2 bytes per component
//item_out_width -= (item_width % 16);

enum AVPixelFormat outformat = AV_PIX_FMT_YUV422P10LE;
enum AVPixelFormat informat = AV_PIX_FMT_UYVY422;
int window_lines[4] = {0};
av_image_fill_linesizes(window_lines, outformat, window_width);

uint8_t *window_planes[4] = {0};
window_planes[0] = calloc(1, window_lines[0] * window_height);
window_planes[1] = calloc(1, window_lines[1] * window_height);
window_planes[2] = calloc(1, window_lines[2] * window_height); /// Fill the window with all 0s, this is green in yuv.


int item_lines[4] = {0};
av_image_fill_linesizes(item_lines, informat, item_width);

uint8_t *item_planes[4] = {0};
item_planes[0] = malloc(item_lines[0] * item_height);
memset(item_planes[0], 100, item_lines[0] * item_height);

struct SwsContext *ctx;
ctx = sws_getContext(item_width, item_height, informat,
               item_out_width, item_height, outformat, SWS_FAST_BILINEAR, NULL, NULL, NULL);

/// Check a block in the normal region
printf("Pre scale normal region %d %d %d\n", (int)((uint16_t*)window_planes[0])[0], (int)((uint16_t*)window_planes[1])[0],
       (int)((uint16_t*)window_planes[2])[0]);

/// Check a block in the corrupted region (should be all zeros) These values should be out of the converted region
int corrupt_offset_y = (item_out_width + 3) * 2; ///(item_width + 3) * 2 bytes per component Y PLANE
int corrupt_offset_uv = (item_out_width + 3); ///(item_width + 3) * (2 bytes per component rshift 1 for horiz scaling) U and V PLANES

printf("Pre scale corrupted region %d %d %d\n", (int)(*((uint16_t*)(window_planes[0] + corrupt_offset_y))),
       (int)(*((uint16_t*)(window_planes[1] + corrupt_offset_uv))), (int)(*((uint16_t*)(window_planes[2] + corrupt_offset_uv))));
sws_scale(ctx, (const uint8_t**)item_planes, item_lines, 0, item_height,window_planes, window_lines);

/// Preform same tests after scaling
printf("Post scale normal region %d %d %d\n", (int)((uint16_t*)window_planes[0])[0], (int)((uint16_t*)window_planes[1])[0],
       (int)((uint16_t*)window_planes[2])[0]);
printf("Post scale corrupted region %d %d %d\n", (int)(*((uint16_t*)(window_planes[0] + corrupt_offset_y))),
       (int)(*((uint16_t*)(window_planes[1] + corrupt_offset_uv))), (int)(*((uint16_t*)(window_planes[2] + corrupt_offset_uv))));

return 0;


  


  }

  


  Example Output:

//No alignment
Pre scale normal region 0 0 0
Pre scale corrupted region 0 0 0
Post scale normal region 400 400 400
Post scale corrupted region 512 36865 36865

//With alignment
Pre scale normal region 0 0 0
Pre scale corrupted region 0 0 0
Post scale normal region 400 400 400
Post scale corrupted region 0 0 0


  


• hls fails to play m3u8 playlist with videojs

  9 novembre 2020, par charlie

  I created an hls stream from yuv frames with ffmpeg which appears to work just fine. It creates a playlist and updates that, and all .ts files as expected. However even videojs refuses to play it, and I really run out of ideas, probably missing the obvious ?

  


  First, the m3u8 list which is continuously updated :

  


  #EXTM3U
#EXT-X-VERSION:3
#EXT-X-TARGETDURATION:2
#EXT-X-MEDIA-SEQUENCE:190
#EXTINF:2.000000,
test190.ts
#EXTINF:2.000000,
test191.ts
#EXTINF:2.000000,
test192.ts
#EXTINF:2.000000,
test193.ts
#EXTINF:2.000000,
test194.ts


  


  Those .ts files exist and have sizes of approx 250 kBytes.
Then, I open this standard videojs html file (residing in the same directory) :

  


  &#xA;&#xA;&#xA;&#xA;&#xA;&#xA;  &#xA;&#xA;&#xA;  <h1>VSTream</h1>&#xA;&#xA;  &#xA;    <source src="test.m3u8" type="application/x-mpegURL">&#xA;  &#xA;  &#xA;  <code class="echappe-js">&lt;script src='http://stackoverflow.com/feeds/tag/video.min.js'&gt;&lt;/script&gt;&#xA;  &lt;script src='http://stackoverflow.com/feeds/tag/videojs-http-streaming.min.js'&gt;&lt;/script&gt;&#xA;  &#xA;  &lt;script&gt;&amp;#xA;    var player = videojs(&amp;#x27;my_video_1&amp;#x27;);&amp;#xA;  &lt;/script&gt;&#xA;  &#xA;&#xA;&#xA;

  


  I open this in Firefox (or Chrome, or Edge, none works) and use its debugger which in the console shows :

  


  


  VIDEOJS : WARN : A plugin named "reloadSourceOnError" already exists.
You may want to avoid re-registering plugins ! video.min.js:12:977
Specified “type” attribute of “application/x-mpegURL” is not
supported. Load of media resource test.m3u8 failed. vstream.html All
candidate resources failed to load. Media load paused. vstream.html
XML Parsing Error : syntax error Location : file :///C :/test/test.m3u8
Line Number 1, Column 1 : 2 test.m3u8:1:1 MouseEvent.mozPressure is
deprecated. Use PointerEvent.pressure instead. video.min.js:12:9031
VIDEOJS : ERROR : (CODE:3 MEDIA_ERR_DECODE) Playback cannot continue. No
available working or supported playlists. Object code : 3, message :
"Playback cannot continue. No available working or supported
playlists." video.min.js:12:977

  


  


  This doesn't make much sense to me. Why would it throw an XML parsing error for an m3u8 file ? Of course, the MEDIA_ERR_DECODE points in some direction, but the same problem persists with fMP4 files.

  


  The relevant ffmpeg c++ code (format is "hls" and formatContext works well in other scenarios, gop is 2) :

  


      AVDictionary* options = 0;
    if (fragmented)
    {
        // type
        if (codecID == AV_CODEC_ID_H264)
            av_dict_set (&options, "hls_segment_type", "fmp4", 0);
        else
            av_dict_set (&options, "hls_segment_type", "mpegts", 0);

        av_dict_set (&options, "hls_time", "2.0", 0);
        av_dict_set_int (&options, "hls_list_size", 5, 0);
        av_dict_set (&options, "hls_flags", "delete_segments", 0);
    }

    int ret = avformat_write_header (formatContext, &options);


  


  ...which returns success. Then writing frames

  


      int ret = avcodec_send_frame (videoStream->codecContext, frame);
    av_init_packet (avPacket);
    AVRational timeBase = videoStream->getTimeBase ();
    avPacket->stream_index = streams.indexOf (videoStream);
    ret = avcodec_receive_packet (videoStream->codecContext, avPacket);
    avPacket->duration = av_rescale_q (1, videoStream->codecContext->time_base, videoStream->getTimeBase ());
    pts += avPacket->duration;
    avPacket->pts = pts;
    avPacket->dts = pts;    //  AV_NOPTS_VALUE;
    ret = av_interleaved_write_frame (formatContext, avPacket);


  


  (videoStream here is an internal object representing the avstream).
I've omitted error checks for clarity. All return values show success.
Thanks for any help on this !

  


• ffmpeg record screen and save video file to disk as .mpg

  7 janvier 2015, par musimbate

  I want to record the screen of my pc (using gdigrab on my windows machine) and store the saved video file on my disk as an mp4 or mpg file .I have found an example piece of code that grabs the screen and shows it in an SDL window here :http://xwk.iteye.com/blog/2125720 (The code is on the bottom of the page and has an english version) and the ffmpeg muxing example https://ffmpeg.org/doxygen/trunk/muxing_8c-source.html seems to be able to help encode audio and video into a desired output video file.

  I have tried to combine these two by having a format context for grabbing the screen (AVFormatContext *pFormatCtx ; in my code ) and a separate format context to write the desired video file (AVFormatContext *outFormatContextEncoded ;).Within the loop to read packets from the input stream( screen grab stream) I directly encode write packets to the output file as shown in my code.I have kept the SDL code so I can see what I am recording.Below is my code with my modified write_video_frame() function .

  The code builds OK but the output video can’t be played by vlc. When I run the command

  ffmpeg -i filename.mpg

  I get this output

  [mpeg @ 003fed20] probed stream 0 failed
  
  [mpeg @ 003fed20] Stream #0: not enough frames to estimate rate; consider increasing probesize
  
  [mpeg @ 003fed20] Could not find codec parameters for stream 0 (Video: none): unknown codec
  
  Consider increasing the value for the 'analyzeduration' and 'probesize' options
  
  karamage.mpg: could not find codec parameters
  
  Input #0, mpeg, from 'karamage.mpg':
  
    Duration: 19:30:09.25, start: 37545.438756, bitrate: 2 kb/s
  
      Stream #0:0[0x1e0]: Video: none, 90k tbr, 90k tbn
  
  At least one output file must be specified

  Am I doing something wrong here ? I am new to ffmpeg and any guidance on this is highly appreciated.Thank you for your time.

  int main(int argc, char* argv[])
  
  {
  
  
  
      AVFormatContext *pFormatCtx;
  
  
  
      int             i, videoindex;
  
      AVCodecContext  *pCodecCtx;
  
      AVCodec         *pCodec;
  
  
  
      av_register_all();
  
      avformat_network_init();
  
  
  
      //Localy defined structure.
  
      OutputStream outVideoStream = { 0 };
  
  
  
      const char *filename;
  
      AVOutputFormat *outFormatEncoded;
  
      AVFormatContext *outFormatContextEncoded;
  
      AVCodec *videoCodec;
  
  
  
      filename="karamage.mpg";
  
  
  
      int ret1;
  
  
  
      int have_video = 0, have_audio = 0;
  
      int encode_video = 0, encode_audio = 0;
  
  
  
  
  
      AVDictionary *opt = NULL;
  
  
  
  
  
  
  
      //ASSIGN STH TO THE FORMAT CONTEXT.
  
      pFormatCtx = avformat_alloc_context();
  
  
  
      //
  
      //Use this when opening a local file.
  
      //char filepath[]="src01_480x272_22.h265";
  
      //avformat_open_input(&pFormatCtx,filepath,NULL,NULL)
  
  
  
      //Register Device
  
      avdevice_register_all();
  
  
  
      //Use gdigrab
  
      AVDictionary* options = NULL;
  
      //Set some options
  
      //grabbing frame rate
  
      //av_dict_set(&options,"framerate","5",0);
  
      //The distance from the left edge of the screen or desktop
  
      //av_dict_set(&options,"offset_x","20",0);
  
      //The distance from the top edge of the screen or desktop
  
      //av_dict_set(&options,"offset_y","40",0);
  
      //Video frame size. The default is to capture the full screen
  
      //av_dict_set(&options,"video_size","640x480",0);
  
      AVInputFormat *ifmt=av_find_input_format("gdigrab");
  
      if(avformat_open_input(&pFormatCtx,"desktop",ifmt,&options)!=0){
  
          printf("Couldn't open input stream.\n");
  
          return -1;
  
      }
  
  
  
      if(avformat_find_stream_info(pFormatCtx,NULL)<0)
  
      {
  
          printf("Couldn't find stream information.\n");
  
          return -1;
  
      }
  
      videoindex=-1;
  
      for(i=0; inb_streams; i++) 
  
          if(pFormatCtx->streams[i]->codec->codec_type==AVMEDIA_TYPE_VIDEO)
  
          {
  
              videoindex=i;
  
              break;
  
          }
  
      if(videoindex==-1)
  
      {
  
          printf("Didn't find a video stream.\n");
  
          return -1;
  
      }
  
      pCodecCtx=pFormatCtx->streams[videoindex]->codec;
  
      pCodec=avcodec_find_decoder(pCodecCtx->codec_id);
  
      if(pCodec==NULL)
  
      {
  
          printf("Codec not found.\n");
  
          return -1;
  
      }
  
      if(avcodec_open2(pCodecCtx, pCodec,NULL)<0)
  
      {
  
          printf("Could not open codec.\n");
  
          return -1;
  
      }
  
  
  
  
  
      AVFrame *pFrame,*pFrameYUV;
  
      pFrame=avcodec_alloc_frame();
  
      pFrameYUV=avcodec_alloc_frame();
  
  
  
      //PIX_FMT_YUV420P WHAT DOES THIS SAY ABOUT THE FORMAT??
  
      uint8_t *out_buffer=(uint8_t *)av_malloc(avpicture_get_size(PIX_FMT_YUV420P, pCodecCtx->width, pCodecCtx->height));
  
  
  
  
  
      avpicture_fill((AVPicture *)pFrameYUV, out_buffer, PIX_FMT_YUV420P, pCodecCtx->width, pCodecCtx->height);
  
  
  
      //<<<<<<<<<<<-------PREP WORK TO WRITE ENCODED VIDEO FILES-----
  
  
  
      avformat_alloc_output_context2(&outFormatContextEncoded, NULL, NULL, filename);
  
      if (!outFormatContextEncoded) {
  
          printf("Could not deduce output format from file extension: using MPEG.\n");
  
          avformat_alloc_output_context2(&outFormatContextEncoded, NULL, "mpeg", filename);
  
      }
  
      if (!outFormatContextEncoded)
  
          return 1;
  
  
  
      outFormatEncoded=outFormatContextEncoded->oformat;
  
  
  
  
  
       //THIS CREATES THE STREAMS(AUDIO AND VIDEO) ADDED TO OUR OUTPUT STREAM
  
  
  
      if (outFormatEncoded->video_codec != AV_CODEC_ID_NONE) {
  
  
  
          //YOUR VIDEO AND AUDIO PROPS ARE SET HERE.
  
          add_stream(&outVideoStream, outFormatContextEncoded, &videoCodec, outFormatEncoded->video_codec);
  
          have_video = 1;
  
          encode_video = 1;
  
      }
  
  
  
  
  
       // Now that all the parameters are set, we can open the audio and
  
       // video codecs and allocate the necessary encode buffers. 
  
      if (have_video)
  
          open_video(outFormatContextEncoded, videoCodec, &outVideoStream, opt);
  
  
  
       av_dump_format(outFormatContextEncoded, 0, filename, 1);
  
  
  
  
  
        /* open the output file, if needed */
  
      if (!(outFormatEncoded->flags & AVFMT_NOFILE)) {
  
          ret1 = avio_open(&outFormatContextEncoded->pb, filename, AVIO_FLAG_WRITE);
  
          if (ret1 < 0) {
  
              //fprintf(stderr, "Could not open '%s': %s\n", filename,
  
              //        av_err2str(ret));
  
              fprintf(stderr, "Could not open your dumb file.\n");
  
              return 1;
  
          }
  
      }
  
  
  
  
  
      /* Write the stream header, if any. */
  
      ret1 = avformat_write_header(outFormatContextEncoded, &opt);
  
      if (ret1 < 0) {
  
          //fprintf(stderr, "Error occurred when opening output file: %s\n",
  
           //       av_err2str(ret));
  
          fprintf(stderr, "Error occurred when opening output file\n");
  
          return 1;
  
      }
  
  
  
  
  
      //<<<<<<<<<<<-------PREP WORK TO WRITE ENCODED VIDEO FILES-----
  
  
  
      //SDL----------------------------
  
      if(SDL_Init(SDL_INIT_VIDEO | SDL_INIT_AUDIO | SDL_INIT_TIMER)) {  
  
          printf( "Could not initialize SDL - %s\n", SDL_GetError()); 
  
          return -1;
  
      } 
  
      int screen_w=640,screen_h=360;
  
      const SDL_VideoInfo *vi = SDL_GetVideoInfo();
  
      //Half of the Desktop's width and height.
  
      screen_w = vi->current_w/2;
  
      screen_h = vi->current_h/2;
  
      SDL_Surface *screen; 
  
      screen = SDL_SetVideoMode(screen_w, screen_h, 0,0);
  
  
  
      if(!screen) {  
  
          printf("SDL: could not set video mode - exiting:%s\n",SDL_GetError());  
  
          return -1;
  
      }
  
      SDL_Overlay *bmp; 
  
      bmp = SDL_CreateYUVOverlay(pCodecCtx->width, pCodecCtx->height,SDL_YV12_OVERLAY, screen); 
  
      SDL_Rect rect;
  
      //SDL End------------------------
  
      int ret, got_picture;
  
  
  
      AVPacket *packet=(AVPacket *)av_malloc(sizeof(AVPacket));
  
  
  
      //TRY TO INIT THE PACKET HERE
  
       av_init_packet(packet);
  
  
  
  
  
      //Output Information-----------------------------
  
      printf("File Information---------------------\n");
  
      av_dump_format(pFormatCtx,0,NULL,0);
  
      printf("-------------------------------------------------\n");
  
  
  
      struct SwsContext *img_convert_ctx;
  
      img_convert_ctx = sws_getContext(pCodecCtx->width, pCodecCtx->height, pCodecCtx->pix_fmt, pCodecCtx->width, pCodecCtx->height, PIX_FMT_YUV420P, SWS_BICUBIC, NULL, NULL, NULL); 
  
      //------------------------------
  
      //
  
      while(av_read_frame(pFormatCtx, packet)>=0)
  
      {
  
  
  
          if(packet->stream_index==videoindex)
  
          {
  
              //HERE WE DECODE THE PACKET INTO THE FRAME
  
              ret = avcodec_decode_video2(pCodecCtx, pFrame, &got_picture, packet);
  
              if(ret < 0)
  
              {
  
                  printf("Decode Error.\n");
  
                  return -1;
  
              }
  
              if(got_picture)
  
              {
  
  
  
                  //THIS IS WHERE WE DO STH WITH THE FRAME WE JUST GOT FROM THE STREAM
  
                  //FREE AREA--START
  
                  //IN HERE YOU CAN WORK WITH THE FRAME OF THE PACKET.
  
                  write_video_frame(outFormatContextEncoded, &outVideoStream,packet);
  
  
  
  
  
                  //FREE AREA--END
  
                  sws_scale(img_convert_ctx, (const uint8_t* const*)pFrame->data, pFrame->linesize, 0, pCodecCtx->height, pFrameYUV->data, pFrameYUV->linesize);
  
  
  
  
  
                  SDL_LockYUVOverlay(bmp);
  
                  bmp->pixels[0]=pFrameYUV->data[0];
  
                  bmp->pixels[2]=pFrameYUV->data[1];
  
                  bmp->pixels[1]=pFrameYUV->data[2];     
  
                  bmp->pitches[0]=pFrameYUV->linesize[0];
  
                  bmp->pitches[2]=pFrameYUV->linesize[1];   
  
                  bmp->pitches[1]=pFrameYUV->linesize[2];
  
                  SDL_UnlockYUVOverlay(bmp); 
  
                  rect.x = 0;    
  
                  rect.y = 0;    
  
                  rect.w = screen_w;    
  
                  rect.h = screen_h;  
  
                  SDL_DisplayYUVOverlay(bmp, &rect); 
  
                  //Delay 40ms----WHY THIS DELAY????
  
                  SDL_Delay(40);
  
              }
  
          }
  
          av_free_packet(packet);
  
      }//THE LOOP TO PULL PACKETS FROM THE FORMAT CONTEXT ENDS HERE.
  
  
  
      //AFTER THE WHILE LOOP WE DO SOME CLEANING
  
  
  
      //av_read_pause(context);
  
  
  
  
  
       av_write_trailer(outFormatContextEncoded);
  
       close_stream(outFormatContextEncoded, &outVideoStream);
  
       if (!(outFormatContextEncoded->flags & AVFMT_NOFILE))
  
          /* Close the output file. */
  
          avio_close(outFormatContextEncoded->pb);
  
  
  
      /* free the stream */
  
      avformat_free_context(outFormatContextEncoded);
  
  
  
  
  
  
  
      //STOP DOING YOUR CLEANING
  
      sws_freeContext(img_convert_ctx);
  
  
  
  
  
  
  
      SDL_Quit();
  
  
  
      av_free(out_buffer);
  
      av_free(pFrameYUV);
  
      avcodec_close(pCodecCtx);
  
      avformat_close_input(&pFormatCtx);
  
  
  
      return 0;
  
  }
  
  
  
  
  
  
  
  /*
  
   * encode one video frame and send it to the muxer
  
   * return 1 when encoding is finished, 0 otherwise
  
   */
  
  static int write_video_frame(AVFormatContext *oc, OutputStream *ost,AVPacket * pkt11)
  
  {
  
      int ret;
  
      AVCodecContext *c;
  
      AVFrame *frame;
  
      int got_packet = 0;
  
  
  
      c = ost->st->codec;
  
  
  
      //DO NOT NEED THIS FRAME.
  
      //frame = get_video_frame(ost);
  
  
  
      if (oc->oformat->flags & AVFMT_RAWPICTURE) {
  
  
  
          //IGNORE THIS FOR A MOMENT
  
          /* a hack to avoid data copy with some raw video muxers */
  
          AVPacket pkt;
  
          av_init_packet(&pkt);
  
  
  
          if (!frame)
  
              return 1;
  
  
  
          pkt.flags        |= AV_PKT_FLAG_KEY;
  
          pkt.stream_index  = ost->st->index;
  
          pkt.data          = (uint8_t *)frame;
  
          pkt.size          = sizeof(AVPicture);
  
  
  
          pkt.pts = pkt.dts = frame->pts;
  
          av_packet_rescale_ts(&pkt, c->time_base, ost->st->time_base);
  
  
  
          ret = av_interleaved_write_frame(oc, &pkt);
  
      } else {
  
  
  
              ret = write_frame(oc, &c->time_base, ost->st, pkt11);
  
  
  
      }
  
  
  
      if (ret < 0) {
  
          fprintf(stderr, "Error while writing video frame: %s\n");
  
          exit(1);
  
      }
  
  
  
  
  
      return 1;
  
  }