Changes between Version 41 and Version 42 of HWAccelIntro


Ignore:
Timestamp:
Oct 30, 2016, 3:40:24 PM (3 years ago)
Author:
oromit
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • HWAccelIntro

    v41 v42  
    134134}}}
    135135
     136== CUDA/CUVID ==
     137
     138CUVID, which is also called nvdec by Nvidia now, can be used for decoding on Windows and Linux.
     139In combination with nvenc it offers full hardware transcoding.
     140
     141CUVID offers decoders for H264, HEVC, H263, MJPEG, mpeg1/2/4, vp8/9, vc1.
     142Codec support varies by hardware. The full set of codecs being available only on Pascal hardware, which adds VP9 and 10 bit support.
     143
     144While decoding 10 bit video is supported, CUVID is only able to output 8 bit video, so the additional 2 bits are lost in the process.
     145
     146Sample decode using CUVID, the cuvid decoder copies the frames to system memory in this case:
     147{{{
     148ffmpeg -c:v h264_cuvid -i input output.mkv
     149}}}
     150
     151Full hardware transcode with CUVID and NVENC:
     152{{{
     153ffmpeg -hwaccel cuvid -c:v h264_cuvid -i input -c:v h264_nvenc -preset slow output.mkv
     154}}}
     155
     156If ffmpeg was compiled with support for libnpp, it can be used to insert a GPU based scaler into the chain:
     157{{{
     158ffmpeg -hwaccel_device 0 -hwaccel cuvid -c:v h264_cuvid -i input -vf scale_npp=-1:720 -c:v h264_nvenc -preset slow output.mkv
     159}}}
     160The {{{-hwaccel_device}}} option can be used to specify the GPU to be used by the cuvid hwaccel in ffmpeg.
     161
    136162== Intel QSV ==
    137163