Skip to main content
4 votes
Accepted

How to extract a video frame using NVIDIA card

Since the video bitstream is being decoded on the GPU, CUDA upload isn't required. Command with format conversion is ffmpeg -hwaccel cuvid -c:v h264_cuvid -resize 1024x70 -i video_source \ -...
Gyan's user avatar
  • 33.4k
3 votes
Accepted

hardware cuda decoding error

Stream 0 is Stream #0:0[0x1006]: Video: mpeg2video for which you need -c:v mpeg2_cuvid
Gyan's user avatar
  • 33.4k
1 vote
Accepted

cuvid - slow performance when used 4 devices

Problem was resolved by replace scale_npp=-2:360 to scale_npp=-2:360,hwdownload,format=nv12 now decoders fully loaded: nvidia-smi dmon -c 1 # gpu pwr gtemp mtemp sm mem enc dec mclk ...
Hett's user avatar
  • 121
1 vote

cuvid - slow performance when used 4 devices

NVidia boards has a limit number of concurrent encoding tasks, according number of encode engine chips on board. Look at specifications of boards you are using. Even if you have a board with ...
PRu's user avatar
  • 36
1 vote

How to extract a video frame using NVIDIA card

Actually we can extract a video frame using only NVIDIA card via thumbnail_cuda filter. For it is necessary to configure FFmpeg with: --enable-cuda-sdk --enable-filter=scale_cuda --enable-filter=...
Ivan Kolesnikov's user avatar
1 vote
Accepted

FFmpeg. CUDA decode. Past duration too large. Variable FPS

The key -drop_second_field 1 fixed this issue. And result command looks as follow: ffmpeg -y -hwaccel cuvid -c:v h264_cuvid -deint 2 -drop_second_field 1 -i "udp://IP:PORT" -vcodec h264_nvenc -...
Ivan Kolesnikov's user avatar
1 vote

hardware cuda decoding error

It's a bug in ffmpeg. When you explicitly specify a type of the HW acceleration (i.e. cuvid), ffmpeg assigns 0 to CUVIDDECODECREATEINFO::ulNumDecodeSurfaces and to CUVIDDECODECREATEINFO::...
facetus's user avatar
  • 111

Only top scored, non community-wiki answers of a minimum length are eligible