Jump to content

QuickSync Transcoding in Emby on a Proxmox Cluster


Go to solution Solved by Luke,

Recommended Posts

PixelWizard
Posted

Hello everyone,

I'm in the process of migrating my homelab to Proxmox, where I have a cluster of three nodes, each with different GPUs. I'm currently using Emby for transcoding, but I'm facing an issue with hardware acceleration.

By default, Emby uses VA-API for transcoding, but I prefer to use QuickSync. I tried changing this by using the advanced transcoding settings, reordering the options so that QuickSync has priority over VA-API. However, when I migrate the LXC container between nodes, since each node has a different GPU and the drivers have different names, the advanced transcoding settings appear unchecked, causing transcoding to stop working.

Additionally, when I switch the container back to the original node, the issue persists. Emby does not remember the selected transcoders from the first node, and they remain unchecked as well.

On the other hand, if I don't use the advanced settings and just enable the general "Transcoding" option, everything works fine, but Emby continues to use VA-API instead of QuickSync.

My questions are:

  1. Is there a configuration file where I can modify the default behavior of the "Transcoding enabled" option, so that Emby uses QuickSync whenever available, without relying on the advanced settings?
  2. Is there any other solution to make Emby prioritize QuickSync when available, even after migrating the container between nodes with different GPUs?

Thanks in advance for any help you can provide.

 

Fist use in first node:

image.png.cbe841c5e586e8ea3b24c2017c4d3e86.png

 

After migrate Emby to other node: 

image.png.3adf09139a524bb6af1724a693a8974d.png

 

Back to first node: 

image.png.90c081ad82348fc9093b41ff80bdf826.png

  • Solution
Posted
Quote

However, when I migrate the LXC container between nodes, since each node has a different GPU and the drivers have different names, the advanced transcoding settings appear unchecked, causing transcoding to stop working.

Additionally, when I switch the container back to the original node, the issue persists. Emby does not remember the selected transcoders from the first node, and they remain unchecked as well.

Hi, the reason why this happens is because the changes causes it to appear as different hardware to Emby Server. 

You're probably better off not using the advanced config. The 4.9 beta channel prioritizes quicksync over vaapi in most cases, so eventually that change will be in the stable release.

PixelWizard
Posted
On 18/08/2024 at 23:49, Luke said:

Hi, the reason why this happens is because the changes causes it to appear as different hardware to Emby Server. 

You're probably better off not using the advanced config. The 4.9 beta channel prioritizes quicksync over vaapi in most cases, so eventually that change will be in the stable release.

To make matters worse, it would be ideal to prioritize dedicated graphics over integrated graphics.
If the computer has 2 graphics cards, with the option “YES” it will prioritize the iGPU over the large GPU.

And then you get back to a similar problem with the one at the beginning of the post.

  • Thanks 1

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...