Hardware Acceleration / Client not compatible

[hertie]

Dear Jellyfin users,

I have a new Asustor machine (AS6702) which is powered by an Intel CPU with a GPU onboard. Running Jellyfin, I am trying to use hardware accelerated transcoding to reduce the CPU workload (VAAPI and QSV). But I can not get it to work.

I can transcode several files without hardware acceleration (HEVC and AVC codec) but the CPU load is pretty high. When I activate hardware acceleration (basic settings) every client (several) returns an error message on the same files, that the client is not compatible to the media format sent.

Any ideas what I could do differently?

Regards,
hertie

[Nazar78]

My model doesn't support QSV so I can't tell but is your renderer node pointed correctly? E.g. VA-API: /dev/dri/renderD128.

Check the logs to see what's the issue, change it to debug if necessary. Get into the container and you can even run ffmpeg commands shown in the logs to troubleshoot.

I'm not running the docker version from the App Central, instead in chroot but they're similar.

[hertie]

but is your renderer node pointed correctly? E.g. VA-API: /dev/dri/renderD128.

Yes, it is. I am running the App Central docker version.

Excerpt from one transcode log file

[h264_vaapi @ 0x55589d76b580] Driver does not support VBR RC mode (supported modes: CQP).
Error initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height
[libfdk_aac @ 0x55589d7657c0] 2 frames left in the queue on closing
Conversion failed!

Still not knowing what to change.

[Nazar78]

I'm not sure of your model, perhaps QSV will not work so just try using VA-API alone.

https://jellyfin.org/docs/general/admin ... en11-igpus