Ever since it was discovered that NVIDIA DLSS .dll files could be manually updated to possibly fix some issues (such as ghosting, for example), TechPowerUP has maintained a database with all of the versions released through the various games so that users may attempt to find out which one works best. Of course, that requires manually downloading and copying files every single time.
Now, though, Australian coder Brad Moore has crafted a handy tool called DLSS Swapper that will do the job for you. Do note that it doesn't have an executable yet, which means it'll require a manual compile.
What is this?
Tool for downloading, managing, and swapping out DLSS (Deep Learning Super Sampling) dlls in games on your computer.
But why?
DLSS gets updates from nVidia, but often studios are not updating DLSS within their own games. See this video from Digital Foundry as to why you may want to manually update DLSS yourself.
What games are supported?
Any Steam game that suppoorts DLSS should work.
My #dlss swapper tool is finally coming together. Just need to make it verify dll hashes for a touch of security and I'll probably go full send on it. (ping @Dachsjaeger) pic.twitter.com/Fesco4ybKh
— Brad (@beeradmoore) August 18, 2021
Do note that this NVIDIA DLSS Swapper tool certainly can't automatically filter which .dll files are optimal for a specific game; it'll still have to be a trial-and-error process, as Brad Moore pointed out in a tweet. Some may not even work, as NVIDIA mentioned when users first started tinkering with DLSS versions.
It doesn’t. There isn’t even any guarantee that a game that came with 2.2.15 will work with 2.2.16. Much more likely to be an issue of v1 <-> v2
It’s probably a good idea to add a hard separation of v1 to v2 of the dlls at least.
The main difference is that you won't have to go through the pain of manually copying files, as the NVIDIA DLSS Swapper tool lets you quickly download a new version, test it in the game, and then roll back to the old one if at all needed.
In recent DLSS news, NVIDIA released the SDK so that developers may easily integrate Deep Learning Super Sampling in their custom/proprietary engines.
