Recherche avancée

Sur d’autres sites (5861)

• hwcontext_vulkan : initialize semaphores of DMABUF imports

  12 mars 2020, par Lynne

  hwcontext_vulkan : initialize semaphores of DMABUF imports
  There was a recent change in Intel's driver that triggered a driver-internal
  
  error if the semaphore given to the command buffer wasn't initialized.
  
  Given that the specifications require the semaphore to be initialized,
  
  this is within spec. Unlike what's causing it in the first place, which is
  
  that there are no ways to extract/import dma sync objects from DMABUFs,
  
  so we must leave our semaphores bare.
  

  • [DH] libavutil/hwcontext_vulkan.c

• avcodec/cuvid : Add support for P010/P016 as an output surface format

  22 novembre 2016, par Philip Langdale

  avcodec/cuvid : Add support for P010/P016 as an output surface format
  The nvidia 375.xx driver introduces support for P016 output surfaces,
  
  for 10bit and 12bit HEVC content (it’s also the first driver to support
  
  hardware decoding of 12bit content).
  The cuvid api, as far as I can tell, only declares one output format
  
  that they appear to refer to as P016 in the driver strings. Of course,
  
  10bit content in P016 is identical to P010, and it is useful for
  
  compatibility purposes to declare the format to be P010 to work with
  
  other components that only know how to consume P010 (and to avoid
  
  triggering swscale conversions that are lossy when they shouldn’t be).
  For simplicity, this change does not maintain the previous ability
  
  to output dithered NV12 for 10/12 bit input video - the user will need
  
  to update their driver to decode such videos.
  

  • [DH] compat/cuda/dynlink_cuviddec.h
  • [DH] libavcodec/cuvid.c
  • [DH] libavcodec/version.h

• ffmpeg zoompan + crop

  24 février 2021, par andykais

  I need to pan across a video that is also being cropped by ffmpeg. How can I accomplish this in a single ffmpeg command ?

  


  So far I see two possible options :

  


  Option #1 : use crop variables pan across the video. This is choppy, but possibly fixed by upscaling the input.

  


  ffmpeg -i input.mp4 -vf 'crop=w=in_w/2:h=in_h:x=t*10' crop_panned_output.mp4


  


  Option #2 : create an alpha mask in the shape of the crop, and then apply it on top of the input video with a zoompan filter. ffmpeg is doing extra work in this case, because we have to pad the input, then zoom in on it, then pan across it, then apply the alpha mask. This is actually using the "zoompan" feature to "pan" though.

  


  magick -size 1920x1080 xc:none -fill black -draw "rectangle 480,0 1440,1080" rectangular-alpha-mask.png
ffmpeg -i input.mp4  -i rectangular-alpha-mask.png  -filter_complex "
  [0:v]pad=2112:1188,zoompan=z=1.1:px+0.5:d=1:fps=60:s=1920x1080[input_pan];
  [1:v]alphaextract[alf];
  [input_pan][alf]alphamerge[masked];
  color=s=1920x1080:color=red[base];
  [base][masked]overlay=shortest=1
"  mask_panned_output.mp4


  


  Are either of these the right way to do it or are they incredibly inefficient ? Is there another option ? Cropping & panning together feels like a fairly common workflow, but these solutions feel a bit hacky.

  


  Here is a visual description of what panning and cropping a video looks like :

  


  First start with a video. Here is a simple one.


  


  I want to crop the video to a certain width/height, and move from the left to the right across the video. The background is colored red for clarity


  


  The video moves from left to right across the original video, and keeps the same crop ratio.