Recherche avancée

Sur d’autres sites (12689)

• Controlling ffmpeg at runtime with zmq

  3 avril 2023, par Gavin

  I want to dynamically change the rectilinear view (eg its yaw) of a 360 video as it plays.

  


  basic command

  


  To take a 360 video and show a flat/normal view at a yaw perspective of 60 degrees

  


  ffmpeg -i input360.mp4 -vf "v360=input=e:rectilinear:yaw=60,scale=iw/4:-1" out60.mp4
This works fine.

  


  Changing the yaw during playback

  


  I understand ffmpeg has two methods to change filter params at runtime ; sendcmd (file based) and zmq (message based)

  


  I got sendcmd method working, but am struggling to understand zmq syntax and use. ffmpeg's zmq docs are pretty sparse.
I am using a local Windows 10 PC

  


  sendcmd

  


  ffplay -i input360.mp4 -vf "sendcmd=f=cmd.txt,v360=input=e:rectilinear:reset_rot=1,scale=iw/4:-1"

  


  with a cmd.txt file

  


  0-5 [expr] v360 yaw 'lerp(0,90,TI)';
5-10 [expr] v360 yaw 'lerp(90,0,TI)';


  


  result : yaw changes from 0 to 90 degrees from t=0-5s and then 90 to 0 degrees from t=5-10s. Perfect

  


  zmq

  


  ffplay -i input360.mp4 -vf "v360=input=e:rectilinear:reset_rot=1,zmq,scale=iw/4:-1"

  


  I got zmqsend from ffmpeg-tools.zip and added to my ffmpeg bin directory

  


  


  1. execute above ffplay command from terminal window #1 - and see video playing

     


  


  2. Open terminal window #2 and execute : echo v360 yaw 90 > zmqsend

     


  


  


  result : no change to video yaw. No errors either

  


  What am I doing wrong ? Im not sure if my ffmpeg command is wrong, or if my zmq message does not reach ffmpeg (or both). I checked my ffmpeg v2022-10-10 config has —enable-libzmq

  


• Copy C++ FFmpeg AVFrame to C# WritableBitmap

  2 janvier 2023, par Spuriga

  I have a C++ function definition with ffmpeg scaler The last two parameter is my question (buffer and stride) :

  


  int scale_decoded_video_frame(void* handle, void* scalerHandle, void* scaledBuffer, int scaledBufferStride) 

sws_scale(
    scalerContext->sws_context,
    srcData,
    context->frame->linesize,
    0, 
    scalerContext->source_height,
    reinterpret_cast(&scaledBuffer),
    &scaledBufferStride);


  


  When I declared it in C# I use the following code :

  


  [DllImport(LibraryName, EntryPoint = "scale_decoded_video_frame", CallingConvention = CallingConvention.Cdecl)]
public static extern int ScaleDecodedVideoFrame(IntPtr handle, IntPtr scalerHandle, IntPtr scaledBuffer, int scaledBufferStride);


  


  And the C# call :

  


  WriteableBitmap w = new WriteableBitmap(1920, 1080, 96, 96, PixelFormats.Pbgra32, null);
RenderOptions.SetBitmapScalingMode(w, BitmapScalingMode.NearestNeighbor);
w.Lock();

Int32Rect rect = new Int32Rect(0, 0, 1920, 1080);

resultCode = FFmpegVideoPInvoke.ScaleDecodedVideoFrame(decoderPtr, scalerHandle, w.BackBuffer, w.BackBufferStride);
if (resultCode == 0)
   w.AddDirtyRect(rect);

w.Unlock();


  


  It works perfect, but I want to skip the scale function and only want to get the decoded image to the WritableBitmap buffer. I try to use this code, change the sws_scale function to memcpy, but it shows nothing. The context->frame is AVFrame*.

  


  memcpy(scaledBuffer, context->frame->data, AV_NUM_DATA_POINTERS * sizeof(uint8_t));
memcpy(&scaledBufferStride, context->frame->linesize, AV_NUM_DATA_POINTERS * sizeof(int));


  


  How can I copy the AVFrame buffer to the WritableBitmap buffer ?

  


• phantomjs screenshots and ffmpeg frame missing

  20 octobre 2015, par Bussiness Way

  I have problem making video from website screenshots taken from phantomjs.

  the phantomjs did not make screenshots for all frames within the same second and even not all seconds there , there is huge missing frames .

  the result is high speed video playing with many jumps in video effects .

  test.js :

  var page = require('webpage').create(),
  
      address = 'http://raphaeljs.com/polar-clock.html',
  
      duration = 5, // duration of the video, in seconds
  
      framerate = 24, // number of frames per second. 24 is a good value.
  
      counter = 0,
  
      width = 1024,
  
      height = 786;
  
          frame = 10001;
  
  
  
  page.viewportSize = { width: width, height: height };
  
  
  
  page.open(address, function(status) {
  
      if (status !== 'success') {
  
          console.log('Unable to load the address!');
  
          phantom.exit(1);
  
      } else {
  
          window.setTimeout(function () {
  
              page.clipRect = { top: 0, left: 0, width: width, height: height };
  
  
  
              window.setInterval(function () {
  
                  counter++;
  
                  page.render('newtest/image'+(frame++)+'.png', { format: 'png' });
  
                  if (counter > duration * framerate) {
  
                      phantom.exit();
  
                  }
  
              }, 1/framerate);
  
          }, 200);
  
      }
  
  });

  this will create 120 image , this is correct count , but when you see the images one by one you will see many duplicate the same contents and many missing frames

  ffmpeg :

  fmpeg -start_number 10001 -i newtest/image%05d.png -c:v libx264 -r 24 -pix_fmt yuv420p out.mp4

  I know this script and ffmpeg command not perfect , because I did hundred of changes without lucky, and I lost the correct setting understanding .

  an anyone guide me to fix this ?.

  thank you all