By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Chazore said:
Alright lads, I've a question.

For those of you who have played GTA V, which AA settings have you used and which do you think do a better job over the others?.

I'm asking because I've honestly never bothered trying Nvidia's MFAA, but I have tried MSAA mixed with TXAA via GTA V's settings. 2X MSAA doesn't really seem to eliminate jaggies, while 4X ends up being rather taxing at 1440p. FXAA doesn't really seem to do much, but I also haven't tried going with DSR.

Which do you think offers a good balance to performance and visual elimination of jaggies?.

I also noticed that MFAA isn't really supported in many games, though Nvidia's site does contain a list of those that do support the AA setting.

I don't play GTA 5. Or any GTA for that matter. (Still sooking over the glory days of GTA 1+2 top down perspective. haha)

But. MSAA > MFAA > TXAA > FXAA in general.

MSAA or Multisample anti-aliasing is spatial antialiasing. And is usually the preffered approach if you care about quality.

MFAA or Multi-Frame Anti-Aliasing will take a temporal and spatial approach over multiple frames and samples, which means it's lighter on hardware than MSAA. It is a good compromise between image quality and performance.

TXAA or Temporal Anti-Aliasing uses a combination of MSAA, Post-Processing and filtering to achieve it's anti-aliasing.

FXAA or Fast Approximate Anti-Aliasing is basically a post processing filtering, essentially just blurring edges.

Depending on your hardware... Enable DSR or Dynamic Super Resolution for nVidia or Virtual Super Resolution for AMD.
It is hardware intensive.
I personally run games like Overwatch, whack 4xSMAA and downscale from 4k to 1440P for an amazingly sharp image. And Overwatch is a light enough game to get away with that on a peasant Radeon RX 580 and still maintain 60fps.

It doesn't have to be 4k though. Even 1800P (3200x1800) could be achievable and bring some gains.



--::{PC Gaming Master Race}::--