I want to share with you a few modifications i’ve made to the Nvidia control panel settings that have produced a very noticable difference in clarity for me. It’s not a dramatic difference, but it is VERY obvious. I don’t think this should noticably influence the visuals when running MSFS on a monitor, but in VR it does. I also haven’t noticed anyone pointing out these exact settings before, so perhaps it might help some of you.
Antialiasing - Mode: Enhance the application setting
Antialiasing - Setting: 8x
Texture filtering - Anisotropic sample optimization: Off
Texture filtering - Negative LOD bias: Clamp
Texture filtering - Quality: Quality (might as well set this to high quality)
Texture filtering - Trilinear optimization: OFF
I am not entirely sure which one of these the chief suspect for an increase of clarity is but my strongest suspicion is the trilinear optimization set to off. I usually tinker a lot with various settings to try to produce some increase in either clarity or performance and i was too tired from re-running the game to exactly pinpoint which one has noticable effect out of these. But please do try and let me know if you notice any change and if someone has the will to exclude them one by one to see which one has the most effect. I should also add i haven’t noticed any change in performance using these changes.
I will be testing this shortly. Thanks for helping out! Can you edit the title to indicate that the main focus is on Enhancing AA in the NCP?
I am curious to see if this works with TAA. Normally, when you choose Enhance AA, you match the Transparency setting to the AA setting In Game. For example, if you are using multisample x2 in game you set the enhance to x2 and set transparency to multisample. However, if you use Texture Supersampling x2 in game, you choose Supersample x2 in the NCP.
Texture Filtering -Quality is a set of presets for all four texture filtering settings. All it does is change the other 3 settings. However, you are correct in your choices for Anisotropic sample optimization and Trilinear Optimization.
I am curious again about LOD bias.
Also, to pick up these setting changes do we go in and out of VR mode or do we require a bounce of the Sim? I am leaning toward bouncing the sim to pick up NCP changes, but I am new to VR.
Hi. I’m pretty sure all nvidia CP settings should be set and applied before running the sim. Don’t get your hopes up about enchancing AA setting. Since it doesn’t tax away at performance i have my doubts it has any effects since as you pointed it’s questionable if it has any affect on TAA. But either one, or the combination of these settings have noticably affected my experience. Please also be aware i am running the sim by using OXR resolution 150% and ingame resolution at 75% which equals 1995 x 1995 for me. I’m using an older WMR headset (Dell Visor). I doubt that it should make much difference if someone is using a G2 for example. I have no idea if this should have any effect on steam users though. I do still believe strongly that one of the texture filtering optimizations are the key. The only setting i’ve tested solo was the enchanced antialiasing but i couldn’t conclude if there’s a tangible difference or a psychological perception in difference.
Yes, you would need to try a similar setup to really compare. I have OXR 100%, In Game 100%, no NIS scaling, no AI, Bing Maps, but no Photogrammetry, no rolling cache, clear skies, and almost all settings set to Low or Off, including Terrain and Building LOD’s at 10 each. No motion smoothing or vsync or frame limited at all.
FXAA is a post-processing technique that is done in the GPU, not in the game frame rendering. Therefore, I suspect that setting AA to FXAA in game only tells the Nvidia card to turn on FXAA. However, I am not sure if you require FXAA to be enabled in the 3d settings tab of the NCP. I am testing both ON with great results. I can have Texture Resolution = Ultra, Shadow Maps = 2048 and Ambient Occlusion = Ultra with FXAA. I find this improves the shadows and cockpit textures without killing performance.
I do not think you can “enhance the AA” with FXAA or TAA. DLAA is supposed to be like MSAA x4 to x6. For the purposes of testing, you should only be able to “enhance” DLAA, in theory. But, if you set the enhancement to x2 does this mean you get DLAA equivalent to MSAA x10? or x8?
Also, can you “enhance” Texture Supersampling in game? If this is set to x8 in game do you get x16 or x64 supersampling with x8 enhancement? These are “expensive levels” of AA.
For this reason I tested DLAA in Game with Texture SuperSampling OFF in game and Enhancement x2 in NCP. My thinking is that I get the equivalent of MSAA x8 or x10 and I don’t pay for expensive SuperSampling.
I can’t be sure if Texture Supersampling only impacts Terrain textures, not cockpit textures.
Hi. I think you may have mistaken the enchanced AA feature to be the main focus here when it’s not. As i said, the main benefit is from one or more of the texture filtering options being turned off.
No worries. It was fun testing AA settings. BTW, trilinear filtering doesn’t sharpen the edges, it blends them. So, yes, turning it off should sharpen edges.
I run a fairly middle-of-the-road, although modern, rig, GPU is GTX1650, at mostly medium settings with a few high and the occasional ultra where it doesn’t seem to affect performance. I generally get between 30 and 40fps on my 1080p monitor with which I’m quite happy.
Not really into experimenting but I thought I’d try those settings and was pleasantly surprised to get what seemed to be a smoother and sharper display with no apparent effect on performance. May just have been subjective of course but I’ll take it! Thanks for that!
I would be grateful to understand the differences between the new NVIDIA in built NIS scaler option and the OPEN XR NIS scaler program and whether they can/should be used together and with best settings. Many thanks for any help on this.
Thank you for the link. I understand the difference now although I am still not sure if I would get the best (or worst) of both worlds by setting OPEN XR NIS for VR and NVIDIA driver NIS for monitor, I’ll ask in that thread.
I believe that it can be distilled down to:
NVIDIA Built in NIS - Only works with 2D (does nothing for VR)
OpenXR NIS - API layer that uses the NIS scaler with OpenXR (does nothing for 2D)
Thanks for your tips. However, my conclusion is exactly opposite to yours regarding trilinear optimization. In my case, to set ON is the key for terrain sharpness and clarity. RTX 2060, quest 2, oculus app resolution max (slide to the far right), OTT ss1.4, ingame rendering 65%, and no other tools.