Recherche avancée

Sur d’autres sites (7510)

• avformat/mov : Add support for exporting Video Extension Usage info

  30 mai 2024, par Derek Buitenhuis

  avformat/mov : Add support for exporting Video Extension Usage info
  This box is provided by files created by the Apple Vision Pro, as well
  
  as the iPhone 15+ when capture for Vision Pro is enabled.
  The boxes are a mix of things documented by Apple in some PDFs, their
  
  API docs, and reverse engineering. Ideally we will have a real spec
  
  one day.
  Links :
  
    * https://developer.apple.com/av-foundation/Stereo-Video-ISOBMFF-Extensions.pdf
  
    * https://developer.apple.com/documentation/videotoolbox/kvtcompressionpropertykey_horizontaldisparityadjustment
  
    * https://developer.apple.com/documentation/videotoolbox/kvtcompressionpropertykey_stereocamerabaseline
  
    * https://developer.apple.com/documentation/videotoolbox/kvtcompressionpropertykey_heroeye
  Signed-off-by : Derek Buitenhuis <derek.buitenhuis@gmail.com>
  

  • [DH] libavformat/mov.c

• ffmpeg drawtext box height variies depending on text

  13 mars 2019, par NFpeter

  I use ffmpeg with zoompan and drawtext filter. I’m writing 4 lines of text in a box.
  Now I’ve the problem that the space between the boxes variies depending on the text height.

  Is there a way to set a fixed height for the text box ?
  Or is it possible to calculate the text position depending on the current text ?

  Here my call with 2 lines of text

  ffmpeg -ss 0 -t 10 -i white.png -y -filter_complex [0:v]zoompan='z=min(zoom+0.10143,1.217217)':x=iw/2-(iw/zoom/2):y=ih/2-(ih/zoom/2):d=25*6.0760000000000005:,trim=duration=6.0760000000000005[v0];[v0]drawtext=text=f:fontcolor=white:fontsize=40:box=1:boxcolor=black@0.8:boxborderw=15:alpha=1:x=w/20:y=(h/1.11)-th*0.5[v2];[v2]drawtext=text=f:fontcolor=white:fontsize=40:box=1:boxcolor=black@0.8:boxborderw=15:alpha=1:x=w/20:y=(h/1.246)-th*0.5 -c:v libx264 -r 25 -pix_fmt yuv420p -strict -2 out.mp4

  

• FFMPEG with deepstream-app rtsp stream fails

  30 mars 2022, par YeonwooSung

  I am currently working with a video streaming project by using deepstream sdk with node-rtsp-stream which uses ffmpeg internally.
For the deepstream part, I am simply re-using the sample codes that the nvidia provided (sample_apps/deepstream-app).

  


  So, the camera records a video, and streams the video via rtsp stream.
Then, the deepstream-app will read the stream, use YOLOv3 model for object detection, add bounding boxes, and streams the output via new rtsp stream. The deepstream-app uses the h264 codec for output.
Next, the node.js program will get the rtsp stream that the deepstream-app generates, and uses the node-rtsp-stream for streaming it to the web page.

  


  However, the ffmpeg keeps generate error "[h264 @ 0x5571499250] no frame !", "[h264 @ 0x5571499250] decode_slice_header error", and "[h264 @ 0x5571499250] non-existing PPS 0 referenced".

  


  The rtsp stream that is generated by deepstream-app actually works fine, I tested it with SMPlayer app, and it works properly.
However, when I test the rtsp stream with ffmpeg by using "ffmpeg -rtsp_transport tcp -i 'rtsp ://localhost:8123/ds-test' -an -vcodec h264 -re -f rtsp", it keeps generate the error message above.

  


  Am I missing something ? or doing something stupid thing ?