Closed RockPolish closed 2 years ago
You need to specify that the input pipe is a Y4M, since it cannot be inferred from the file extension (add --y4m 1
to the cmd). See vvencapp --fullhelp
for details.
Oh I see, I figured it would be derived from the header. Sorry for the inconvenience
Hi all, I can't get vvenc to correctly read the input parameters like resolution and framerate when piping y4m input to it, while it does work when reading from a .y4m file. Instead, when piping y4m, it defaults to 1080p 60 FPS. The following script converts the input file (a 1 second 720p30 clip) to y4m in two different ways which produce identical output. It then calls vvenc in three different ways first passing the .y4m file, then piping either from ffmpeg or directly from the .y4m file, in each case decoding the produced output to a new .y4m file.
The md5 output is as follows:
Only 1.vvc & 1.y4m are correct 720p30 video and not the 720p pixels but viewed as a 1080p clip:
Platform: Ubuntu 18.04 (WSL 2) Version: ver. 1.4.0-rc1