My 3090 Settings for my G2 (2.7.11)

[ShaunX]

Nope, I completley forgot about that in my excitement.

 

Just tried it and it certainly makes a difference..to both cpu and gpu

 

GPU Frame Time: 9.21ms

CPU Frame Time: 6.10ms

 

Thanks, I'll leave it there for now as I think we're hijacking S-O-H's thread a bit.

 

Thanks again @speed-of-heat for putting this all together for us.

[speed-of-heat]

All good guys, but, I know a lot of people have moved from Intel to AMD need this sort of help !

[ShaunX]

Just for completeness, I repeated the benchmark for Test 26 today, then installed the sharper eye consolidation and ran with it disabled and then enabled with defaults and 1.8 sharpening.

 

 

..so it appears that just having it installed and disabled has a some impact, and obviously greater when enabled which confirms your findings

 

Brightness=1.002000
CASContrast=1.000000
Contrast=1.008000
MaxDelta=0.070000
Saturation=1.100000
Sharpening=1.800000
SharpeningRadius=0.500000
SingleImage=0

 

[speed-of-heat]

Thanks @ShaunX , that was some work that you saved me there!

 

[zildac]

2 hours ago, ShaunX said:

Just for completeness, I repeated the benchmark for Test 26 today, then installed the sharper eye consolidation and ran with it disabled and then enabled with defaults and 1.8 sharpening.

 

 

..so it appears that just having it installed and disabled has a some impact, and obviously greater when enabled which confirms your findings

 

Brightness=1.002000
CASContrast=1.000000
Contrast=1.008000
MaxDelta=0.070000
Saturation=1.100000
Sharpening=1.800000
SharpeningRadius=0.500000
SingleImage=0

 

Very interesting, thanks for taking the time. I hadn't really thought about it but it kind of makes sense that even with it installed and NOT active there will be some latency added as DCS will still be using the modded local copy of DXGI.DLL in its render process.

Edited April 29, 2021 by zildac

[ShaunX]

Damn! Realised I ran those tests with MSAA off. I thought those GPU numbers looked remarkebly low particularly as I haven't overclocked it, and I don't have the VR Mask.

 

Repeated test 26 with MSAAx2

 

 

Big hits! particularly my FPS average.

 

 

Edited April 29, 2021 by ShaunX

[speed-of-heat]

yeah test 4 onwards assume MSAA x 2 as part of the High preset, but about 1 to 1.5ms seems about right as a delta 

Edited April 29, 2021 by speed-of-heat

[ShaunX]

Noticed that I got a SteamVR Beta update this morning so decided to rerun the last test again. Ran it twice. (MSAA is still x2)

 

 

coincidence ?

[speed-of-heat]

that's right on the edge of my "is it though"  0.17 - 0.15ms   variance just checking you were just re-running 26 not Sharper EYE ? 

[ShaunX]

5 minutes ago, speed-of-heat said:

that's right on the edge of my "is it though"  0.17 - 0.15ms   variance just checking you were just re-running 26 not Sharper EYE ? 

 

Definately 26, flat shadow removal, clear water, shadows reduced frame rate, reprojection disabled, NO sharper eye

(msaa still x2)

 

Edited April 30, 2021 by ShaunX

[speed-of-heat]

OK I will check later today as well, as my Steam VR just got updated

[Mathieas]

Hi all,

 

I realized I also forgot the 2XMSAA. I do however have the new steam VR beta.

 

Test 26 (Beta 1.17.7):

GPU Frametime: 10.40

CPU Frame Time: 6.11

[DavePastry]

so wait is the kegetys shaders mod working in 2.7? 

[speed-of-heat]

Yes…

[DavePastry]

1 minute ago, speed-of-heat said:

Yes…

ah! I see it now, excellent, thank you for your hard work my kind sir! 

[ShaunX]

Following some Windows updates yesterday I'm suddenly seeing the VR Mask for the first time !

 

So I was able to repeat the above (test 26 settings) and confirm that the VR Mask does improve framerates and FPS..

 

 

 

[freehand]

I think it would be useful if people quoted there current windows version before giving tests results so we are all on the same page as differences can vary.

[ShaunX]

51 minutes ago, freehand said:

I think it would be useful if people quoted there current windows version before giving tests results so we are all on the same page as differences can vary.

Sorry, Windows 10, version 2004. My DCS settings, SteamVR SS, and installed mods are as per S-O-H's test 26. I'm on SteamVR and WMR for SteamVR betas. I have Resizable bar turned on though not sure it makes any difference. I have basic PBO OC enabled on my 5900X.

 

..the only additional app I have running is Viacom for Voice Attack. It doesn't seem to have any impact at all running idle during the test.

 

 

Edited May 4, 2021 by ShaunX

[Panzertard]

An absolutely stunning study of the effects of 'everything' affecting VR performance in DCS, in a scientific manner.

And one that has taken the time to explain how & why the effects occur. Much kudos to you sir!

I'll definitely spend a weekend going through the list and improving my own rig.

 

Thank you Dr. Speed. o7

 

Edited May 4, 2021 by Panzertard

[speed-of-heat]

@Panzertard Glad it helps!  Thanks for the nice comments! 

[speed-of-heat]

The latest patch (2.7.0.5659) had very little performance impact one way or the other, so that's a good thing I got about 0.2ms worse for GPU and 0.2ms improvement in CPU , which is close enough to be run to run variance that I  dont care very much  ... the 0.2 ms improvement could be some other changes i have made, i would need to back them out and re-baseline.  the CPU improvement could significant depending on the cause.  

[Burt]

Wow !!! I came straight to page 12 and haven’t been in here in a few weeks and to just see what’s up and looks like you have updated some things  I guess and everyone is excited so going back a couple of pages to see what’s up... hell I’m excited just reading the last few post as of now. 

[speed-of-heat]

i completely re-wrote the post for 2.7

[Burt]

Thanks ! Wow that takes a lot of time I appreciate it. 

[speed-of-heat]

thanks for the comments, and support