now I try since some months for the ffmpeg video de-encoding or transcoding the best setup with it the videos, however, become small with the best quality.
I have loaded the Cuda of stack for Visual studio and have installed or executed with VERSUS (Cublas & Matricmul) The video Sdk I have also installed or executed in VERSUS.
With ffmpeg order:
-c:v nvenc_hevc-c:v hevc_nvenc tunes film-rc:v vbr_minqp-qmin:v 10-qmax:v 21-b:v 2500K-maxrate:v 8000K-pixel_format yuv444p-pixel_format cuda-c:a copy-hide_banner-acodec ac3-b:a 384-K are 48000 Ac 2
if the coding goes rather fast, with files of 2.5-3,0 Gb I come on Gb 1.3-1.7 per file
indeed, does not come with all possible settings on the same quality like with “-x265”.
however-x265 lasts very long there ffmpeg only fps=0.3 - max. 2-3fps computed.
Can give to me one maybe a few tips like I a good result for “-c:v nvenc_hevc-c:v hevc_nvenc…” got?.
Nvidia 1050 t
AMD quad / 3.8 Ghz
Ram 8 Gb the GDR 4