gnome-vrr VS LatencyFleX

https://copr.fedorainfracloud.org/coprs/kylegospo/gnome-vrr/ :

  gsettings set org.gnome.mutter experimental-features "['variable-refresh-rate']"

This probably isn't what you're looking for, but by enabling this: https://copr.fedorainfracloud.org/coprs/kylegospo/gnome-vrr/ you can get vrr/vsync on Wayland for gnome, it can cause flickering but otherwise it's pretty good and got rid of any tearing and stuttering for me.

You can grab the patch from https://github.com/KyleGospo/gnome-vrr and build it.

Source: https://github.com/KyleGospo/gnome-vrr/issues/5

Based on the comments in the AUR repo, I'd assume that the maintainer still invests time trying to update mutter-vrr. There seems to be a non-trivial issue when rebasing on Gnome 44, but I'm not able to assess how complex the problem actually is. (https://github.com/KyleGospo/gnome-vrr/issues/5)

Currently you can either use a Wayland DE that supports VRR, I know KDE has support for it, or use the gnome-vrr copr or build it yourself from his GitHub.

For fedora you can use https://copr.fedorainfracloud.org/coprs/kylegospo/gnome-vrr/

Just wanted to share my updated PKGBUILD to enable VRR in Gnome 43.2. The original author has no time to update his AUR package it seems so i updated it myself with the sources from the Fedora COPR.

Installing LatencyFleX on multiple different games and Proton versions was cumbersome to me. I created an install script to do that for you. Tested on Arch Linux, should work on any distribution, provided LatencyFleX is installed on the system beforehand.

Find liblatencyflex_layer.so, latencyflex_layer.dll, latencyflex_wine.dll from extracted https://github.com/ishitatsuyuki/LatencyFleX files

No and it has a vendor agnostic alternative that works better: \ https://github.com/ishitatsuyuki/LatencyFleX

Strangely, a reflex alternative is available on Linux for AMD cards (third party, open) https://github.com/ishitatsuyuki/LatencyFleX Nvidia and AMD could both implement that as a standard, but NVidia is not keen on standards, and AMD is oblivious.

If you're on linux there's LatencyFleX: https://github.com/ishitatsuyuki/LatencyFleX

https://github.com/ishitatsuyuki/LatencyFleX here you go

And this depends on what actually reflex is and doing - Nvidia love to take a standard api and giving it a brand name and strut around like they invented something new. For example, https://github.com/ishitatsuyuki/LatencyFleX seems to suggest they get similar advantages "just" using the submission timing interfaces already available in the standard graphics APIs, I wouldn't be surprised if reflex is "just" an implementation of that, possibly with an API designed to be a bit easier to integrate in game engines than the gamedev doing it themselves.

As long as you have the most recent driver version and are on Proton Experimental, you shouldn't really have to do anything. You are on a VERY old Proton-GE build, which does not include DXVK 2.0, so that's probably why you're having issues. Maybe look into LatencyFlex, but again, with these bans going around, use at your own risk.

Use LatencyFlex instead. However it might be a bit tricky to get it working. I've not yet had success using it in Overwatch 2.