DeltaMike

Nothing wrong with openfsr anyway.

Seems like a lot of really bright people, the developers post frequently. not a lot of hand holding

Check and make sure you have automatic context switching turned off in WMR settings. You want to have to manually press Win-Y to switch the mouse from the desktop to the headset.

Based on what I've been able to surmise, the default settings are thought to be optimal. Ton of discussion in the ReShade forums if you're interested. Tough crowd over there.

1.. Neither. They both do the same thing. The big picture is, performance correlates with pixel count. Pixel count increases linearly with Steam resolution, and with the square of pixel density. For a G2, the formula is (3100x3100)*g*a*p^2 Where g = Steam General Resolution, a = Per App resolution, p = pixel density Doesn't matter how you get there, long as the numbers add up. Best to just adjust one, and set the others to 1.0 or 100%. With Steam resolution being the more granular parameter. There's some debate as to whether adjusting one is better than the other. If there's a difference in quality, it is subtle indeed. It's OK to ignore that debate until we come up with some solid data. Meanwhile, pick one. Resolution is king in DCS. Get a high res headset, and shoot for native. 2. In terms of antialiasing, in decreasing order of performance hit: Supersampling has theoretical benefits, especially for low resolution headsets, but in DCS at least, eliminating jaggies and shimmering isn't one of them. Flogs the daylights out of your GPU. MSAA looks great but it crushes your framerate. Ditto. I think MSAA is quite a bit more efficient than supersampling at getting rid of jaggies and shimmering. FXAA and SMAA accomplish something at least, at about the third of the cost of MSAA. (SMAA is probably the better technology and the ReShade version of it is pretty impressive imo) 3. It all sucks. People were kind of butt-hurt that DCS went to deferred rendering, which killed MSAA. But time marches on, and there's new technology we are begging DCS for. Meanwhile, take a ride around NTTR and then run the same mission in MSFS. Which one is better? I imagine your answer is like mine, "it depends." Do you want to look at the scenery, or blow it up? Valid question I think

I found this influential I wound up getting a G2, it was on sale for like $450 which made it a steal and I really like it. Problem with the G2 is, the interface between SVR and WMR. There are a lot of moving parts, lots of ways to screw up the install. You would think there would be no way to screw up the 3090 but as you can see there are plenty of posts saying stuff like "wuh I'm only getting 10fps wtf" etc. Flip side is, people make it more complicated than it needs to be. You don't have to do that. The interface between SVR and WMR is very inefficient but you do have enough horsepower to where you shouldn't notice it. The G2 has a narrow sweet spot but I feel that's OK for DCS. I was this close to getting a Quest 2. Oculus software is highly evolved and personally I felt its motion smoothing was superior to WMR's. I was using a Rift S back then, so perhaps that's apples and oranges. Also I didn't have Zuckerberg looking over my shoulder. I don't do Light Sabre so didn't need that functionality. . Pimax 5K is about the same number of pixels as G2 but with a wider FOV. Sweet spot should be wider, but not quite as sweet (fewer pixels per degree). Pick your poison. Pimax 8K sounds awesome on paper, but question in my mind is, whether the 3090 is even enough to drive it at native resolution. You'd have to decide how you feel about undersampling. I don't mind it in MSFS but don't like it much in DCS. Dunno about Pimax software, whether or not it's a PIA, I'd check on that. Growling loves his. I think the varjo aero is the wave of the future, sounds like it's really "early access" and it's expensive. I found this entertaining (same dude reviewed the Aero later). If I had the money for a varjo and a 3090, and could live with the external tracking, that's what I would do. But I LOVE fiddling with things, YMMV. Here's Tyriel's comparison. Hopefully they've done something about that chromatic aberration. G2 has the bang for the buck, especially if you can get a deal on it. Quest 2 has the ease of use. Varjo has the cutting edge tech. Personally I'm not a fan of Pimax. That's my take on it.

I think I see what you're saying. You don't need a bajillion triangles to define an object's profile; just enough. If you blur the edges, eg with FXAA, you would need more pixels. Likewise, with lower pixels-per-degree, you would need more pixels to resolve the object, as if you were looking at a low-res monitor. Now I feel guilty about casting aspersions on ED. I don't normally do that. I think they do good work.

Interesting, so we could do both then. Supersample via steam or DCS, and then do a pass with TAA. Interesting. I'd try that in MSFS, except I'm scared my GPU will catch on fire Regarding draw distance, I guess that's been a topic of discussion for years. Good discussion here for example. When I talk about spotting distance in my tests, I'm referring to the pixel-sized sprites. Way I look at it, in VR, we have the same probletunity as people with low-res flat screen monitors. Even though you might have high-res monitors in your headset, when you put em right on your eyeball and blow them up to iMax proportions, each pixel looks pretty big, and I think we do come out ahead by turning AA off. As for once you're close enough to get a 3D object, I still wonder about level of detail as a function of draw distance, but I wouldn't know one way or the other

OK I set up this test. Obviously these tanks are a lot closer than an aircraft would be. The pic is taken at 100/100/1.0 with no anti-aliasing, and it's actually a fairly good representation of what I see through the G2. The third tank is shimmery but identifiable as a tank. The fourth, I can't consistently see the gun so I might know it's armor, but wouldn't know it was a tank. 5 and 6 are just blobs, I couldn't tell 6 from any other vehicle. SSAAx2 and MSAAx2 look about the same. Running a pixel density of 1.4 (equivalent to Steam at 196%), I think tank #4 looks a little clearer. Shimmery but I think I could identify it as a tank. To my surprise setting steam resolution at 200%, with SuperDooperSampling turned on, doesn't look as good as PD 1.4. I think the supersampling techniques (SSAA, MSAA) have some potential here, as you are increasing the amount of information available to your brain, averaged over time. FXAA of course does not increase information available and seems to be reducing it actually. Unclear to me which way TAA swings. Biggest concern I have is the resolution of objects as a function of draw distance. I think it's possible the information just isn't there.

Good point. It's tempting to say, we are up against a hard limit in terms of information available, based on the number of physical pixels-per-degree. But, that's the probletunity of supersampling in VR; if you keep your head moving, there's more information at hand than the display can render. You just have to rely on your brain to stitch it all together. In other words, your CPU makes a map, and your brain makes a map. All the GPU has to do is, act as an intermediary between the two. With that in mind, maybe we should set the matter of shimmering aside and see if supersampling (one way or the other) enhances object identification at intermediate distances.

Everybody is CPU (or RAM) -limited in MP. Nobody is CPU limited in Caucasus Free Flight, which is where us benchmark nerds like to do our thing. (You might could get yourself that way by maxing out stuff like clouds, trees, clutter etc. All of which affect CPU times at least as much as GPU. Best not to do that when benchmarking.) Primarily, what limits your GPU is primarily how many pixels you're trying to make, keeping in mind that supersampling and MSAA both affect that number. Texture quality may matter depending on your GPU. Here, on the other hand, we are looking specifically at a software inefficiency in the SVR-WMR interface. Discussion in this Reddit post.

Ah. That explains a lot. Based on your explanation, ReShade's "TAA" isn't really TAA then. I will say, it's kind of a nifty hack. It is depth-aware, so for example you can set it to where your cockpit isn't blurry but stuff "out there" is. Clearly not made for VR, I imagine it works best with a static camera. But in the end, all it really does it make the image blurry, with ghosting besides. I hear what you're saying about MSAA. I feel much the same way about generic supersampling. As I went from Rift, to Rift S, to G2, seems the cost of supersampling has increased exponentially but the benefit hasn't kept up. To the point where I'm starting to think it's pointless. MSFS's TAA implementation does work, but overall the scene is kind of a blurry mess. It's pretty, so let's call it a hot mess. It's not always that hazy out there, especially out over the desert. When I say "traffic in sight" what I mean is, I see a blob flashing red and green. Is it a Cessna? A 747? Flying saucer? No way to be sure. Which is fine. Apparently the only thing that crashes in MSFS is MSFS Far as DCS is concerned, my feeling is, pretty is as pretty does. The game plays great in a high-resolution headset. I guess to summarize my feelings on anti-aliasing: Supersampling is the bomb if you have a low-res headset; kind of disappointing given the cost, if you have a high-res HMD. To me, anyway. MSAA is fine, if you have the GPU to drive it. VR Toolkit or SMAA if fine, if you don't None are necessary to play the game, which plays best at native resolution imo Aaaaand I'm spent

I remember watching this video before making up my mind which HMD to buy. Kind of gives us an idea of what we are up against. Note how the Quest2 in general has a softer image. Lots of shimmering in the G2 and the Index; those are hard images. The challenge with the G2 is softening it up, which is neither easy nor inexpensive. I finally got to take MSFS out for a spin. Their implementation of TAA is pretty good. Might be worth taking another look at this in ReShade. That'll be next. Greater men than me have tried. It appears Lordbean was never able to get rid of the ghosting, hence his HQAA effect. Not holding out a lot of hope.

Shouldn't have to. The whole point is to get around the bottleneck between SVR and WMR. Way I look at it, ED didn't cause this problem, and they can't solve it. Fubar.

With a reverb, if you're running higher than 3100x3100 pixels using any combination of settings, you are anti-aliasing via supersampling. MSAA basically does the same thing only I think it's slightly more efficient and looks nicer. Supersampling is more granular, so if you're stuck between MSAAx2 and MSAAx4 you could try that Steam SuperDooperSampling or whatever they call it. It's OK I guess. Oddly DCS SSAA seems to be one of those things that uses GPU cycles and doesn't do anything. The combination of supersampling with a post processing filter might be interesting. Can't mix MSAA with SMAA but you should be able to run your PD up a little and polish it up after the fact. ReShade's SMAA filter is pretty decent, plus there are tools to adjust the color palette etc. Also some trippy stuff if you're in the mood. You've got a beast of a GPU. If you can hold 45fps with everything maxed... I dunno man. I wouldn't kick it to the curb just yet. Good silicon is... well, good!

Well, certainly maxing out the settings will amplify any differences between one GPU and another, so I can see the utility for benchmarking. Not sure of the relevance for game play. I don't think there's any GPU that will consistently run DCS at 90hz even at low settings. Beyond that, there are certain settings that carry a GPU cost, with no discernible effect on visuals. I question the value of increasing PD >1 on modern high res headsets, for much the same reason. As you've seen, FPS only matters if you can cross some threshold (30, 45, 60, 90). VR is different.

There's the CPU side, and the GPU side. They are related, to the extent that DCS has a lot of objects with a lot of detail running around at any given time. Once you get into multiplayer you'll get real sensitive to the CPU side of things. So far we've been more or less keeping the CPU out of it. In single player, especially on the Caucasus map, your performance is mostly about how many pixels you can render. Driving a G2 is challenging; it has a LOT of pixels. In order to run it at native resolution you have to render about 9.3 million pixels. Only three GPU's on the market can do that, maybe four if you count the 6800XT. Textures may be a different matter, interesting comments on that here. I've always thought TMU count had a correlation with DCS performance, and the 2080 doesn't have a ton of those relatively speaking. I'm not totally sure that lowering your pixel count is going to get you there as easily as lowering your texture settings, as you are doing. Let me know what you decide, I'd be interested to know. As for AF, we all kind of have the idea its "free" although nothing is free in life. I've seen estimates that it can take up to 20% of your GPU power to max that out. AF affects how textures look at a tangent, you can see the effect most easily by looking at runways. You'll have to figure out how much of that you want to live with. Net of everything, I think you were right to begin with, at this point probably best to dial textures and AF down until you can get it running how you want it. Getting back to the pixel count. There's an easy way and a hard way to regulate the number of pixels you're rendering. The easy way is with the Steam Global slider. The hard way as you've seen is with FSR, which is a little more complicated than it needs to be. Now you have at least four knobs to twiddle! Easy to get yourself twisted up. If FSR is frustrating you, it's OK to set it aside. Among other things, if you can find a Steam resolution you like, we can calculate the equivalent ScaleFactor if you want to do a simple A/B test. I don't think DCS does a bad job of throwing pixels on a screen, there's only so much you can do with 9 million pixels. The biggest headache in DCS is on the CPU side, where everybody wishes we could use more than a couple of threads at a time. That does affect VR performance -- your poor CPU has to do the math twice for each frame -- but that's another rant for another day. Good news is, once you get your frame size (total pixels rendered), texture quality and AF squared away in single player, those settings will carry over into multi-player. In MP you may have to adjust vis range, trees, shadows, water, cloud settings. May have to upgrade your CPU or RAM. But at least your GPU won't go out of tune just because you jumped into a server. You're pretty much there, really. If the low textures look OK to you, you're in business. ETA: I think reshade 5.0 was the first to have native VR support. There were some mods for earlier versions but not much point in that, 5.0.2 is pretty slick.

I can't get it to work either

Anti-aliasing smackdown -- continued. Today I took her for a spin to compare a couple of anti-aliasing techniques in flight, both over the city and over the desert. I also set out some Russian trucks in Alamo Bravo to see how these things affect spotting. Again, DLAA is a really interesting option, it results in a very soft image and really crushes the shimmering. Unfortunately, it crushes my framerates too so once again I abandoned that one. Also, I don't like it. You might; it has the same look as the TAA, but it actually works in DCS. SMAA is growing on me. Just gotta set it up right. First use the "Display Depth" widget to see if ReShade has proper access to depth information. Note you have to be in a mission, in a 3D environment for this to work, it won't work on the opening screen. Then, turn that off and go into the SMAA widget and "Enable Predicated Thresholding" (see pic). That's the secret sauce; SMAA looks pretty darn good, at a very modest cost if you do that. NOTE: access to the depth buffer is disabled in multiplayer. I tested SMAA alone, using stock settings and predicated thresholding, against my "Big Three" faves (SMAA, Curves, Lumasharpen) and running native resolution, 100/100/1.0 with no anti-aliasing at all. I have the "Big Three" looking a little garish right now, but it works. Lumasharpen might work a little too well for some people. Makes your MFD's and HUD's look awesome, but some of your shimmering comes back. If you really hate shimmering, you might want to leave that one off. Also I took VR Perf kit out for a ride. Since we last spoke, I adjusted my anti-aliasing settings. Currently using Mode 1 (FXAA), subpix 0.75, edge detec 0.16, darkness thresh 0.8. Works pretty well. But note, of the three main options (SMAA, FXAA, and native resolution without AA) only FXAA had a noticeable effect on spotting distance. With SMAA, the images are a little blurry and squirmy but you can see em. Plain vanilla native is still the best option for spotting. Thoughts: Honestly I don't know that I plan to use any of these in routine missions. Maybe Lumasharpen for MFD clarity. As for the others, once you get outside the city, it really doesn't matter much. If you're out sightseeing, yeah pile em on. In a misssion? Eeeeehhhhhh.... By the same token, I've kind of lost interest in supersampling. Not for a high resolution headset like the G2. To the extent MSAA works at all in DCS, it is, essentially, supersampling. And it does a pretty darn good job if you ask me. So in other words, if you're inclined to supersample to begin with, just use MSAA. If you don't have the headroom to run MSAA, and you want some anti-aliasing anyway, I can recommend VRTK if you want all the effects at minimum cost, and can live with the rather crude antialiasing FXAA gives ya. Otherwise you owe it to yourself to give SMAA a try. It's pretty good stuff.

It's really the next gen headsets you want to watch, I think. Suspect the G2 is gonna be the last of the "brute force" hmd's that actually work, and that the pimax 8k will prove to be the bridge too far. We are almost there, our early adapters seem to be loving the varjo

Three years of college French and I'm still such a baby! If I understand correctly, I think you need to uninstall FSR, repair your DCS install and then re-install FSR. We go through that drama every time we upgrade. Mod managers make that process a bit less painful

Check TED's thread on his 6900xt, it'll all be relevant. Especially the parts on tuning. These things aren't quite tuned for DCS right out of the box. But, you can cheerfully ignore all the stuff about FSR mods, as of yesterday that's included in adrenaline

Developers have a choice of scaling some things, and rendering others at native resolution. Which I'll admit is pretty cool. Rest of us pretty much have to scale the whole scene after the fact. Which is fine, for DCS anyway. AMD isn't keeping it a big secret, it's the same protocol fholger uses in vrperfkit and other mods.

Nice. Those things (6800xt) are hard to come by. A lot of us are very happy with AMD + G2 you're gonna love it

Hm. Define "maxed out" in VR terms... Answer depends on the headset-- watcha got?