That said, you can try to establish your own calibration between the current versions of the encoding libraries in your ffmpeg by running the following command, which executes two popular video quality metrics: ffmpeg -i encoded-video.mp4 -i reference-video.mp4 -lavfi 'ssim psnr' -f null. But x265 is not yet as mature in its development as x264, so take that CRF equivalence with a pinch of salt. x265 CRF 28 is supposed to be e quivalent to x264 CRF 23. When selecting CPU encoding there is a detailed dropown for the pixel format (4:2:0 10bit, 4:2:2 10bit, 4:4:4 10bit and so on) - but encoding on the CPU takes ages.The CRF scales for x264 and x265 do not correspond. Then I end up with a 4:2:0 encoded 10bit video, loosing both quality in color depth as well as chroma subsampling. I realize that it will cost more disk space in such an Incompatible pixel format event, but at least I do not loose any quality.īut when I select HEVC (NVIDIA NVENC) In Voukoder the only thing I can select is 8bit or 10bit. >Incompatible pixel format 'yuv444p12le' for codec 'hevc_nvenc', auto-selecting format 'yuv444p16le' That means that even if I have 4:2:2 12bit footage like yuv444p12le nvenc will auto-switch to the higher pixel-format without quality loss:įfmpeg.exe -i in.mov -c:v hevc_nvenc -preset p7 -tune hq -rc vbr -multipass:v fullres -qmin:v 15 -qmax:v 20 -pix_fmt yuv444p12le -an 'out.mkv'

When using ffmpeg from command line I get an output that has 4:4:4. >Supported pixel formats: yuv420p nv12 p010le yuv444p p016le yuv444p16le bgr0 rgb0 cuda d3d11 I have an Nvidia 1080Ti which in higher than 8 bit color depth not only supports YUV 4:2:0 hardware encoding but also 4:4:4.

It puts the default encoding capabilities in various video editing software to shame.

Ffmpeg hevc