edmuss

With open composite I get better performance and image quality running lower resolution with MSAA. Running the reverb at 2600 wide resolution with 2xMSAA + MFAA gives less shimmering than the reverb at 3160 wide resolution with no MSAA.

I've not seen it myself, I do have image stabilisation set to zero though. The only issue I was aware of with the image stabilisation is it causing the waterboarding effect when reprojection is enabled.

I wouldn't say that's necessarily the case, for 99% of titles current hardware is more than enough. Using DCS as a benchmark isn't an accurate indicator because the software simply isn't written from the ground up for VR. As and when the new engine is implemented, I would expect a step change in VR performance (some of which might get swallowed up by new bells and whistles) that will likely make a significant difference. That said current top end hardware is quite capable of running DCS VR at high settings when properly tuned.

Not speaking from direct experience but... For clarity I would suggest the aero but FOV is similar to the G2 I think. For FOV I believe the pimax is technically highest but stretching the pixels over all that FOV results in a reduction of clarity. Performance wise, I don't think they're really going to be any better than the G2. It's the computer itself that gives the performance, if you want it faster you have to downgrade the number of pixels you're pushing.

Have you checked for spare registry keys? It doesn't manifest exactly the same for me and also doesn't stop openxr working at all. However generally if I use openkneeboard it can throw up error warnings on the second tab. Have a look here.

I run maximum performance on the DCS:Blackshark profile but not normally on the global tab. As a test I've just enabled it on global and the GPU is sitting idle at 39°C (ambient is 28°C) and the board power draw according to GPUz is 24W, fans are all completely stopped by the fan profile. If I fire up WMRP the draw jumps to between 60 and 90W, switching between normal/high performance in NVCPL doesn't seem to affect the power draw whilst idling.

^^^ This is the stuff I use but I've not bothered to fully wrap a length, just the shorter contact patches as needed. It's great for cable management (obviously) and we have miles of the stuff at work Note that my cable is actually close to double weight as well because I have a 3m USB extension lead twisted around the reverb cable to feed the leap motion.

Tried running it not as admin? So does DCS not show up in the switcher app at all? What headset are you using?

Bring it over your shoulder and into your lap, that way the weight of the cable is supported by your legs and your head is free to rotate without snagging the cable. I'd tried routing over the back but like you say the drag and weight pulls it about.

You've not got DCS installed in program files have you? You can simply copy and paste it out to the root of C:\

I use a cobra clubman seat which is also pretty reclined, the back is about 30° from vertical. I have the cable mostly coiled and hanging on the wall to the side of the chair, there is enough free cable to loop down into my lap, up over my right shoulder to the back of the headset. I don't have any issues with the cable snagging but when new they seem to have a slightly tacky feel to them. You can sheath the cable with something a bit more slippery, I use spiral wrap which acts as both a strain relief and it's more slippery. https://uk.rs-online.com/web/p/cable-spiral-wrapping/8117673

Where is the open composite being run from? I think if it's in the program files folder then windows might prevent write access. I run it from a seperate HDD that I use for storage.

Just so when you can't find the .etl trace file like I couldn't, use windows search to look in the windows\system32 folder, it should be in there

Hmmm, same settings as mine in that respect. As an outside chance, try resetting the openxr runtimes to WMR; it's a trick that worked for me in the past and also a couple of other people reported it helping. It shouldn't do anything though! Have a trawl through the steamvr settings and in there somewhere (possibly developer/advanced settings) is an option to set steam as the current openxr runtime. Click that and close steamvr, then return to WMRP and it should display a banner at the top warning that WMR is not the current openxr runtime, there should be a button saying "Fix It". As above it shouldn't do anything but it has made a difference for me in the past. Worth a try

Yup that's the one, what pre-rendered vr frames do you have in Nvidia control panel?

Just installed the updated firmware and even in the control panel it's visibly a good deal smoother! edit: just tested in the A10C and it's infinitely better, the physical stick gives more control simply by being longer, the firmware makes it so much better - top work @Deltaalphalima1

I've got an asus 3080ti TUF OC, I don't have a case as such so the GPU is open to ambient, I do have a 140mm fan feeding cool air to the GPU though.

Still seems hot, mines idling at 30/31°C with a 23°C ambient. Have you got a good air flow through the case?

Same here but at the same time maximise the performance. Are you saying that at the card is idling at 60°C? That is incredibly hot, even with my fans stopped it's sitting below 40. Do you have a particularly small case or poor airflow? If your idle temps are that high then load temps will be through the roof which will likely be a cause for instability.

I am indeed air cooled, also in a small enclosed space so the ambient ramps up to almost 30 pretty quickly once the GPU starts acting as a space heater Good shout with cyberpunk, it's installed so no issue to go test! I am only on 1080p though so I suspect that DCS vr will be so much more taxing.

I know it's probably been said before but you do have movement damping on the inputs tab in openxr toolkit set to zero don't you?

Well I've spent the last 90 minutes watching runs of heaven benchmark Result is, 1845mhz @875mv +1000 memory is best bang for buck and at least 5°C cooler than any other settings. I tried up to 1980mhz @975mv (in 15/25 increments) and it just didn't hold the core clocks but was getting up to 86°C. I've already proved this setting DCS stable for a few hours so I guess I just carry on and see what happens

It will eventually auto update but you can force it by starting the MS app store, searching for openxr for WMR and hitting the update button.

Yeah I'll be winding the VRAM back for higher core testing, it's a quick and easy one to change compared to adjusting the curve if afterburner is deciding to be flaky

I did try quickly at 950mv but it seemed to crash heaven without too much hassle. I'll give it a bit more of a thorough test to see if I can eke some more core out of it! The asus tuf cards do have a separate VRAM heatsink so I guess that might help decouple the thermals from the core, not too sure how much difference it would actually make or is it just asus marketing? I've not played with gddr6x before but when overclocking the gddr6 on the 3070 it would produce artifacts on anything much above +1200mhz. I know that it has ecc functionality and will error correct in the background whilst slowing down slightly, also it seems to have a false ceiling that will produce more errors and slow down before then speeding up. It could be that at +1000mhz (which gives ~205fps at the start of heaven) is below the ceiling and at +1500mhz (which gives ~220fps at the start of heaven) is above the ceiling.