DeltaMike

Yeah we all get that. There are a couple of different explanations. It could just be framerate mismatch, or it could be your G2 constantly flipping between 90 and 45fps. Either way, as you've seen, locking your framerates at 45 takes care of a lot of that. Which is probably the primary benefit of using MR in your situation (doubt you're getting much "judder" at those framerates). Good news is, there are other ways to lock in your frame rate other than MR, which I agree generally causes as many problems as it solves. Me, I just dialed up my settings until my system naturally runs at 45 and that works well enough for me. The "correct" way to do it is, locking your frame rate in software. You should be able to do it with your GPU driver. Also can lock your frame rates using the Companion App (the OXR toolkit you get from github), which is quicker and easier. @TED does both, with good results. Try locking your frames at 45, without MR, and see what it looks like. To me it looks pretty awful in SVR but in OpenXR it's not bad. When you're real close to the ground, going real fast, things on your 3-9 line will look blurry but it should be smooth. If you like it, you might as well load your GPU up, because your new target is 45fps instead of 90. I hear the 3080 runs better under a load anyway. Finally, you could just run your headset at 60hz. @edmuss gets good results doing that. Many of us get a noticeable flicker at 60hz so ymmv.

The way it works is, the Oculus runtime knows how many physical pixels are in your headset. Acting on that knowledge, it sends a request to DCS to send over some multiple (or fraction) of that. The exact size of the request, you can regulate with OTT supersampling. DCS can either comply with that request, or not, depending on the PD setting. Those two settings combine to result in a certain size image being rendered by DCS. That image is warped and scaled by Oculus to fit inside your HMD. If you rendered more pixels than can be displayed in your headset, the image is scaled down and it looks awesome. If you rendered fewer pixels than your headset can display, it's scaled up and looks not so awesome. As you figured out, how many pixels you render has a big effect on performance. How you arrive at that number doesn't matter, it's the total number that counts. I'll hand it to you-- you're a tough negotiator. "I demand 150%!" "Screw that, I'm only giving you half that much!" I get it; I have that discussion with my supervisor all the time. In your case, though, best to remember you're negotiating against yourself.

Still gotta use Jabbah's fork for now. With ColinM's .ini file. As the first post in this thread says. (My suggestion: do everything in that post, in order. That post has been continually updated and is the authoritative reference

Leaving aside for the moment that you just don't have enough horsepower to do this, I see you have PD at 0.7 and supersampling at 1.75. Thus, pulling yourself in two opposite directions. Net of everything you're running at 86% of native resolution. I'd keep PD at 1.0 and adjust your resolution using OTT but you need to undersample, not supersample. So you need set your "super" sampling to 0.5 or something, and even then I'm not sure it will be enough. I imagine it's already a blurry mess anyway but, it'll be a blurry mess. So, speaking of your next rig .... When I went to build a rig around the G2, I didn't want a tower computer, got no place to put it. Was very tempted to get an alienware laptop but was advised against it; the laptop GPU's never perform at the level of desktop GPU's and are prone to thermal throttling anyway. I put together an itx build which works like a charm. ITX boxes aren't great for overclocking unless you're willing to build a custom loop (and even then the architecture of the box can work against you, the NR200 for example is very hard to do liquid cooling right). I went with an AMD CPU which doesn't need overclocking and runs right chill with a tower cooler. I pretty much had to go with a 6900XT, it was the only thing on the market at the time. 3080 would be about perfect though, their cooling solution works great in those teeny cases (believe it or not). If you don't prefer to build your own, Thermaltake has a great little ITX rig they call the "Reactor." Personally I think they should have called it the "coffeepot" but still. I couldn't have built the thing for what they go for on Amazon and it has decent stuff in it. And it's a decent way to get your hand on a 3080 without paying scalper prices.

Thanks. Agree with your thoughts re: MR. I've been flying the F18 mostly and have my head down in the cockpit most of the time; under those circumstances all I really notice is the artifact.

^Interesting, I would have thought the Index would have run circles around the G2 in SVR Your settings fall into three main categories: Those that affect your CPU times. Shadows mainly but water and clouds also. Tree visibility, not so much any more (used to be). Civ traffic. You're fine there; just remember, anything that affects your CPU time will affect your GPU time also and it's really easy to get yourself CPU limited in DCS so, don't go too crazy with that stuff. Those that affect your capacity to map textures. I get the feeling AF and detail factors fall into this category. You're fine there. Those that affect how many pixels you're rendering. This would include pixel density, resolution settings, scale factor (if you're using FSR or NIS) and MSAA. You're better off than you used to be. I'm not sure what I think about vis range any more. Used to be a factor but not anymore, not with my new rig and probably not with yours either. Crank it up. Optimal settings depends on the map and mission. You can get away with stuff on the Caucasus map that you can't get away with on Syria or Marianas; and stuff in SP that you can't get away with in MP. I'd suggest starting with a solo mission on the Caucasus map. You should be able to crank up textures, vis range, and AF all the way. You can drive either HMD at native resolution. PD 1.0, resolution (in OpenXR Developer Tools) 100%, and you don't have to use FSR or NIS if you don't want to. With the Index, you might consider over-sampling. I'd try 150 or 200% (using the ODT slider, leave PD alone), which would fool the eye into thinking your HMD is higher resolution, and will have a bit of an anti-aliasing effect. Far as anti-aliasing is concerned, with the Index, I feel supersampling is gonna be superior to MSAA in pretty much every respect. How much you supersample depends on how much you want in terms of shadows and clouds. You should be able to add a fair amount of eye candy. Let us know what you come up with With the G2, you have some tricky decisions to make. I think there's a strong argument for running it at native resolution -- the clarity is remarkable, makes it easy to read instruments and easy to spot things. As you dial down the resolution, your "sweet spot" gets smaller and blurrier, and your sweet spot is small enough to begin with. The only problem is, you won't have much left over. You set water at whatever suits your needs; it'll be higher if you're primarily flying off the carrier, and lower if you're mainly over land. Once you've done that, between shadows, clouds, and MSAA, you're only going to have room to crank up one. There's a great temptation to undersample the G2 to make room for at least two of those things, with MSAA being the greatest temptation of all, because the G2 image is pretty hard and shimmery. This is where FSR and NIS come in. Just be careful with that stuff. You can get away with a scale factor of 77% (which is about 60% resolution) without affecting game play to any appreciable extent. Go much lower than that and it'll hurt your spotting distance. Some of us strongly feel the G2 should be should be run at native resolution, and one should eschew scaling and learn to ignore the shimmering. Others feel strongly FSR or NIS should be brought to bear. Don't let people "should" all over ya. Pick which one works for you. Once you're tuned for the Caucasus map you should be OK on PG or NTTR (long as you're not there just to buzz Dubai or the Strip). If you're flying on Syria, you'll have to make an adjustment. I back off on textures, which makes sense for me because my VRAM bandwidth isn't as good as yours. You might back of on FSR or NIS (if you're using it) because it's easier to adjust on the fly. Pick your poison. If there's an answer for Marianas, I'll be danged if I know. If you're in a complex single player mission, like the ones Liberation can crank up, frankly I think the answer is to run the mission on a server. If you're going in to multiplayer, who knows. You're gonna be CPU (or RAM) limited so I would personally look mainly at your shadows settings. What used to work with me back in the Rift S days was to turn shadows on for SP, off for MP and that seemed to work for me. I haven't been doing MP since I got my G2 and haven't seen a lot of discussion on that topic here.

Is your performance crashing? I would personally start by looking at my GPU driver. My GPU isn't all that stable at factory settings, if it resets I run into problems. Not that exactly but might be worth considering? I've also got in the habit of restarting DCs before a big mission

You should get a little boost from open xr. You can reduce resolution in the toolkit. Between that and cutting back on textures seems like you could get something going until you can figure out what's going on with the companion app. Sometimes openxr doesn't update, you may have to uninstall/reinstall that to get the latest version. There's a guide to that somewhere around here.

What were your resolution settings when you were in svr? Were you using a scaling mod? Shader mod?

I'd start with clouds, shadows and water either off or low. Msaa off. One player suggested backing off on textures. I'd try low terrain textures and dial down detail factors. Try to keep cockpit textures high if you can. You'll probably have to undersample. Either using resolution setting in oxr toolkit or preferably by using FSR or nis through the companion app. I'd try 77%. Don't mess with pixel density, keep that at 1.0 Tune from there Let us know what you come up with

Whoa. Liberation is basically unplayable in VR on Syria map, CPU times up to 37ms once all the units are up, FPS below 20. Running it on the server I'm getting CPU time 14, GPU time 16.

@TED -- what settings are you running on the Syria map these days? Here's mine I just started flying on Syria (gorgeous map) and had to adjust something... Following a tip I lowered my textures. Put terrain textures to low and put the detail sliders in the middle and that did the trick for me. Which surprised me a little, we aren't exactly short on VRAM or TMU's but it works

^I can't get back into DCS at all if my HMD goes to sleep. I think that might be an OXR bug but I'll try the menu thing next time

Yeah backing off on terrain textures does the trick on the Syria map. Doesn't do squat on Marianas, which seems to be flogging my CPU primarily. Presently testing some monster missions on the Syria map.

That's the best time to use the rudder, when you're flying at high AoA. That's when the rudder actually does something. Problem with banking is, you lose lift. You see this when refueling, for example. No matter how gently you ease the stick over, you wind up low and slow when you're done. A teeny touch of rudder might work better. Granted, you are correct, you still have to mind the stick due to dihedral effect. You'll get a little roll. Point being, it's a little. When you're flying the pattern, you're fairly close to stalling at any given time. You really don't want to lose much lift if you can help it. If you do, pulling on the stick won't get it back. Quite the opposite usually. So, when you're turning to final, you have the option to flatten out your bank angle a little and use some rudder to make up the difference. When you're on final, as when you're refueling, you have the option to use a little rudder to get yourself lined up (and, when there's a cross wind, to keep yourself that way). I guess you don't have to use the rudder if you don't want to. I'll say this though, if it takes more than a little rudder to get yourself lined up on final, you probably made a mistake two turns ago.

I think you get better metrics using the OpenXR Toolkit Companion App. Use the detailed overlay. You want to look more at render times than FPS anyway.

Everybody who plays DCS has performance problems in missions that have a lot of moving parts. VR players have it worse, since we have to render each scene more or less twice. All tests done with G2, OpenXR, MR off (FPS allowed to float). Measurements taken 7,300 feet above Batumi heading north. First I wanted to check a few things I haven't tested in a long time: Vis range Texture quality (including "detail factors," which I assume has something to do with textures) Shadows Baseline tests done in the caucasus "free flight" mission over Batumi. I started with everything either off, or at its lowest setting, and then tested one variable at time. To my surprise, texture quality and vis range had no discernible effect on performance. This probably depends on the GPU I imagine. People with older GPU's will almost certainly have different results. Hot tip: turn down cockpit texture resolution last. None of those settings seriously affect game play -- the game plays just fine at med vis range for example -- but low cockpit textures look awful. I was also surprised to see that tree visibility did not affect performance. Hooray for the new trees. They do look pretty cool at high detail factor. Next, I tested these variables in a complex single player mission. I used the Liberation Dynamic Campaign to generate a mission on the Caucasus map. Not the easiest, not the busiest, somewhere in the middle. Although my GPU and CPU times went up, once again vis range and textures had no discernible effect. Here, shadows seemed to affect the CPU primarily. Finally, I turned all three on (medium shadows, high textures, ultra range) and compared performance running the same mission on my game computer, compared with running on a dedicated server. I used a Beelink SER3, which is kind of an AMD based NUC computer for this test. Four cores with a boost clock of 4Ghz, 16Gb RAM and a little touch of Vega. Nothing to scream about but even that was overkill. There wasn't a huge difference in the numbers but enough to keep me consistently above 45fps, which is the secret to life if you're running a G2. Definitely takes care of the stutters even when you're flying low to the ground. Just for giggles, I took a spin over Guam on my server. That map performs as crappily on the server as it does on the local machine. Which leads me to think, the main benefit of running your mission on the server is, letting the server handle the scripts. CONCLUSION -- Turning off shadows is a quick and easy way to mitigate CPU load as you swap from simple singer player missions to complex environments with lots of moving parts. -- If you enjoy playing heavy missions with a lot of scripts running, playing off the dedicated server has something to recommend it. Doesn't take much oomph to run, you can dust off an old laptop or something. No problems connecting; I queried the router to see what address it assigned my server, punched that in the multiplayer screen, boom done. (There are a couple of extra steps involved if you want to play Liberation on your server, hmu with any questions)

Even among those of us who prefer to build our own, pre-builds have been attractive lately as it may be the only way to get your hands on a decent GPU. I was *this close* to buying a Thermaltake Reactor 3080 rig, it could use more RAM but otherwise is a fine rig for DCS. Look at the 3080 if you're interested in VR, the 3070 should be more than enough for flat screen Price is about right, I couldn't build it for what they are asking Cool thing about sporting an AMD CPU and NVIDIA GPU is, you don't technically have to tune it. Should work fine for DCS right out of the box. That case doesn't work for the space I've allocated but it's a cool case anyway. Seems like it should be able to make you a nice pot of coffee, no?

Interested to know what your frametimes look like with the Index. With the G2 you're primarily trying to get out from under WMR4SVR. Seems like the Index shouldn't have that problem. Should be a good match for your 1080ti no?

make sure you disabled automatic context switching, ie set it so you have to manually press win-y to use the mouse in VR vs on desktop

Agree. We could all use CPU level efficiencies. But, we already have a potential dual-CPU option -- running SP missions on a separate server. Hoping ED will consider supporting dual GPU setups when NVIDIA and/or AMD step up to the plate.

Depends on how we define "stutter" -- in OpenXR it's more like a blur than the ratchety effect I got in SVR. What Tzeer is probably calling "ghosting." @TZeer glad you got it sorted! Haven't tried locking FPS in the toolkit. Although last time I was looking at how smooth stuff was on the 3-9 line I was showing in the low 50's on the DCS FPS counter, so I imagine I was at 45 most of the time anyway. Far as actual gameplay is concerned, lately I've just been letting the GPU do what it wants and the game is playing fine, and for the most part I don't notice any stutters or glitches. That's in complex SP games on the PG map, Raven One and Liberation. Only notice problems when Adrenaline resets itself. Which it does a lot. My next round of testing is gonna be looking at CPU times in big missions. So far OpenXR isn't any worse than SVR. DCS either runs or crashes; when it's running, it's fine. I've noticed, if I let my headset go to sleep, I can wake it up, and get dropped into WMR home but can't get back into DCS. Wonder if that's a bug.

Hm. So, we aren't really looking precisely for a radar reflection, we are looking for a doppler shift. So for a distant target right off the nose, the faster they are going, the easier they are to see. Not as true at medium prf but I think it's still at least somewhat true, no? Doppler shift varies with angle. So, seems to me we should be comparing the look-up scenario with trying to find something at your altitude, but at some angle off your nose?