Recherche avancée

Sur d’autres sites (7565)

• Problems in my sample C code using FFmpeg API

  11 juillet 2019, par Tina J

  I’ve been trying to change an FFmpeg’s example code HERE to call other filters using its C APIs. Say the filter be freezedetect=n=-60dB:d=8 which normally runs like this :

  ffmpeg -i small.mp4 -vf "freezedetect=n=-60dB:d=8" -map 0:v:0 -f null -

  And prints outputs like this :

  [freezedetect @ 0x25b91c0] lavfi.freezedetect.freeze_start: 5.005
  
  [freezedetect @ 0x25b91c0] lavfi.freezedetect.freeze_duration: 2.03537
  
  [freezedetect @ 0x25b91c0] lavfi.freezedetect.freeze_end: 7.04037

  However, the original example displays frames, not these metadata information. How can I change the code to print this metadata information (and not the frames) ?

  I’ve been trying to change the display_frame function below into a display_metadata function. Looks like the frame variable has a metadata dictionary which looks promising, but my attempts failed to use it. I’m also new to C language.

  Original display_frame function :

  static void display_frame(const AVFrame *frame, AVRational time_base)
  
  {
  
      int x, y;
  
      uint8_t *p0, *p;
  
      int64_t delay;
  
  
  
      if (frame->pts != AV_NOPTS_VALUE) {
  
          if (last_pts != AV_NOPTS_VALUE) {
  
              /* sleep roughly the right amount of time;
  
               * usleep is in microseconds, just like AV_TIME_BASE. */
  
              delay = av_rescale_q(frame->pts - last_pts,
  
                                   time_base, AV_TIME_BASE_Q);
  
              if (delay > 0 && delay < 1000000)
  
                  usleep(delay);
  
          }
  
          last_pts = frame->pts;
  
      }
  
  
  
      /* Trivial ASCII grayscale display. */
  
      p0 = frame->data[0];
  
      puts("\033c");
  
      for (y = 0; y < frame->height; y++) {
  
          p = p0;
  
          for (x = 0; x < frame->width; x++)
  
              putchar(" .-+#"[*(p++) / 52]);
  
          putchar('\n');
  
          p0 += frame->linesize[0];
  
      }
  
      fflush(stdout);
  
  }

  My new display_metadata function that needs to be completed :

  static void display_metadata(const AVFrame *frame)
  
  {
  
  
  
  //    printf("%d\n",frame->height);
  
  
  
      AVDictionary* dic = frame->metadata;
  
  
  
      printf("%d\n",*(dic->count));
  
  
  
  //    fflush(stdout);
  
  } 

• FFmoeg : What does the source code do when input file is picture ?

  22 août 2022, par Eynnzerr

  As is known, in command lines, FFmpeg can receive pictures as input and output an encoded video, for example,

  


  ffmpeg -i ./pictures/%3d.png output.mp4


  


  Here I'm curious about what happened in source code, i.e. how FFmpeg reads the pictures and in what format it stores them. I think to get output video, input pictures should first be converted to AVFrame, but I have no idea how to do that. I tried to read the source code of ffmpeg, but it's too hard for me to figure out now.

  


  I'd appreciate any helpful advice to solve my question !

  


• Request aid to translate from an old avconv code to ffmpeg

  27 septembre 2022, par Azankiew

  I am working on an old script which uses avconv to manipulate some videos. I wanted to ask help to convert the following code into ffmpeg as I aknowledged avconv is not supported anymore.

  


  The code is :

  


  subprocess.call('avconv -i %s -an -filter_complex "select=between(n\,%s\,%s),setpts=PTS-STARTPTS, crop=%s:%s:%s:%s, scale=-2:224" %s'%(
            input_video_folder+clip.video_name,
            clip.start_frame,
            clip.end_frame,
            crop_width,
            crop_height,
            crop_left_distance,
            crop_top_distance,
            video),
                        shell=True)