I get the following error: Splitting the commandline. Hopefully that helps someone else.Ffmpeg -init_hw_device qsv=hw -filter_hw_device -f rawvideo -pix_fmt yuv420p -s:v 2560x1440 -i normal-desktop-use-1440p.mkv -c:v h264_qsv. I did not yet reinstall TOP or any driver. Now, I can't use the iGPU or the Nvidia, Plex goes straight to CPU/software encoding.ĮDIT 2: Removing the container and image, uninstalling Nvidia and Intel TOP and drivers, and then recreating the container has worked for me. Were you able to find a resolution? I thought it might be a permission issue, but I've tried 777 to no availĮDIT: Also running an nvidia GPU as well, but before changing my cache drive, I had no issues (Using the HardwareDevicePath field in Preferences.XML to select my iGPU for quicksynk, in my case D129). I am having a similar issue after changing the Cache drive my appdata folder was on. Here is also a screenshot from my logs with HDR Tonemapping on. (I reverted the changes to value /dev/dri/) I also tried Adding HardwareDevicePath="/dev/dri/renderD128" to the Preferences.xml but nothing changed. the HDR Tonemapping stopped working via QuickSync the normal Transcoding works fine. I started all my dockers and also Plex and than I was wondering why my CPU is near 100% when I am watching an HDR/DV Movie/Series on my Tablet. Than I switched to a new Cache NVMe, so I moved the appdata folder from the old NVMe to the new one. I switched my hardware from an old 7700k with GTX1650 to a new i5 13500.Įverything worked fine with the additional value for the container: /dev/dri/, HW Transcoding and HDR Tonemapping, I had no problems. Not sure what I'm missing here but hopefully I'm just overlooking something. Nothing seems to actually get the Plex docker container to use the iGPU but switching my extra parameter back to "-runtime=nvidia" and hardware transcoding works fine again. I still have my Nvidia Quadro P400 installed in my system as I'd like to keep using that for Tdarr and just have Plex on the Intel iGPU Only thing that I can think of is ls in /dev/dri populates card0, card1, card2, renderD128, and renderD129 I've tried adding the extra parameter "-device=/dev/dri" in multiple different types of ways from every thread I could find with no luck. Edited February 20 by DemoRicįor some reason I can't seem to get this to work for me. This easy tutorial wouldn't be possible without your work. Chaz with his video showing how to do this for linuxserver's PLEX app. Standing on the Shoulders of Giants: Thanks b3rs3rk, ich777 for your work on the Apps, and Cpt. If you're using ARC or multiple GPU's you may have to do some additional changes (see thread below) On UNRAID server go to Settings, and then click on GPU Statistics in the User Utilities area. Then on the left panel go to Settings > Transcoder. Enable "User hardware acceleration when available" Then go to settings (the little wrench icon at the top right). Use PLEX's WebUI and login to the admin account. Then "Add another Path, Port, Variable, Label or Device", name it (whatever you'd like Intel iGPU or -device /dev/dri it doesn't really matter) and give it the value: /dev/dri/ On your UNRAID Server, Goto DOCKER and left-click on the PLEX icon and select Edit. Once you've configured PLEX to work already you will need to add the ability to use hardware encoding. GPU Statistics (GPU Statistics isn't required but is nice to have)ĬONFIGURE PLEX ON UNRAID: (I won't cover everything on how to install PLEX, just how to enable QSV hardware encoding.) Install the following apps on your UNRAID server: NOTE: Not all intel CPU's have QUICK SYNC encoding capabilities (most do, but you'll have to check)
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |