Another thing, I tested it on a non gaming laptop as well. What I mentioned.
- Data My laptop has:
- Intel Xe GPU,
- i5-12500H CPU.
- Results Results were the same. Visual change is there (looks like a better version of TAA), but performance is unchanged.
Conclusion: Obviously my GPU is far worse than my CPU, GPU is bad for gaming and CPU is quite good and fast. So it's not me being CPU limited.
For testing I even disabled all other running processes, even if my laptop is very clean. No antivirus, no crap. CPU was not even sweating.
Visual change is there, I won't argue about it. It looks just like it should (helps aliasing just like TAA would), but performance is absolutely dead unchanged (maybe minimally lower). Checking all performance modes.
Realistic theory is that the mod author simply COULDN'T test it under ALL scenarios. And that under our scenarios there are issues with his mod. Another person wrote to me that he has very similar issues, but only on iGPU. I observed this issue also iGPUs as well(!) So most likely the mod itself doesn't handle iGPU scenarios well.
So our only hope would be to reach to him and ask him for improvements. Practically impossible, as I never saw him responding to his comment section. I may become his Patron supporter if this offers a way to report issues properly.
I'm going to check Lossless Scaling, as I hear that LS1 is pretty good. I want to compare it with FSR1.