GunSlingerAUS

Thanks so much - this worked, but will it have a performance impact? I only tried it while flying formations, not on a crowded carrier deck.

I'll be doing the same with my 3090,if and when it ever arrives. I'm thinking we should come up with a standardised test using the free DCS assets (Frogfoot and map) so everyone can get a good idea of performance?

ED just admitted that Vulkan ain't gonna be a thing for a looooong time - which means 2-3 years at the very least. This was kinda obvious for anyone who has been following this studio for the better part of 20 years and also knows how hard it is to convert software to multithreading, but to see them saying it in a video should finally put the Vulkan talk to bed. It ain't gonna happen in DCS or MAC.

Which means DCS is GPU limited at 4K Max detail settings. So even on a flatscreen, I'll see a performance uplift going from the 2080 Ti to the 3090. Now I just gotta get my hands on one - going to be over AU$3000 here in Australia (equates to about US$2130). Ouch. And yes, I've seen the new leaked 3090 benches that show it's only 10 to 20% faster than the 3080. So be it. :(

I honestly don't think we'll see Vulkan in DCS - not unless they do a v3.0. As for DLSS 2.0, I'm not sure if the engine is suited for it - it's drawing a lot of long distance scenes, and from what I've seen in games like Control, that's where DLSS doesn't do so well.

So you're running all settings maxed, including mirrors? I have a very similar system, but I do run either 2xMSAA or 4xMSAA as I get jaggies on the 55" screen at 4K, and I'm running the Tomcat with 512 mirrors. I get drops to around 45fps in some scenarios. I can't imagine your CPU is giving you a 30% boost over mine?

So, just to be clear - will a 3090 deliver better FPS for me in the Reverb compared to my 2080 Ti? I thought it would, as I was under the impression that my GPU is the bottleneck in VR?

Yeah, I saw he mentioned that, but other reviews didn't point this out. They also showed a very small performance increase in MSFS. Did you get the feeling that Gamers Nexus focused a bit too much on their new tests (cold plate pressure, and hints of that fancy air flow chamber thingy), rather than the performance tests we care about? I think hardware unboxed is the best review I've seen so far - Jay Z 2 Cents wasn't bad either, which is a surprise to me as I usually find his videos junk.

Ooh, can't wait! The 3080 does deliver upwards of 30% performance increase over the 2080 Ti in a very small number of games, so who knows, DCS could surprise. It is interesting to see that MSFS 2020 doesn't see much of a performance increase though.

The benchmarks are now live. It seems we can expect the 3080 to deliver around a 20% to 30% performance increase in traditionally rendered gfx engines. Ie, non-Rtx. over the 2080 Ti. This probably isn't enough to push my Reverb to 60fps minimum, which is required to remove the need for asynchronous reprojection. However, it may be enough for other HMDs, though I'm not familiar with their baseline performance. So have the reviews sobered your expectations for VR and the 3000 series? Personally, as a 2080 Ti owner it's not quite the leap I need. I think the 3090 *might* just scrape in with 60fps. But at US$1500, it's a big *might*. On the other hand, if you are running a 1080 or 2080 (non-Ti, non-Super), then the performance increase should be substantial across most games - whether that is true of DCS is yet to be seen. Can't wait to see what VR results you folks get with the 3080. Would be fantastic to see some kind of standardised tests. Pssst, ED, a simple benchmark tool would once again be welcome!

As far as I can recall, this problem has been acknowledged for many, many years. I remember posting about it around 4 or 5 years ago, and is the reason my milsim squad (which uses NATOPS procedures as much as possible) uses dot labels during Ops. It's a major disappointment for me with DCS, as I love the fidelity of 4K on my screen. Apparently ED has tried several times to code a fix, to no avail. I should note that it's not the longer range that is the problem for many - it's this weird intermediate distance of around 1nm where the way the model is drawn is problematic.

If you're blind in one eye, can I ask why you don't just run a monitor with Track IR? You'll get better image quality than VR? Or is it the in-cockpit feeling of VR that you like, even though you can't experience the 3D effect?

When running DCS it's easy - if you decrease the resolution (which relies on the GPU) and don't get a performance increase, you're probably CPU bound. If you decrease object draw distance and shadows (which rely on the CPU) and don't get a performance increase, you're probably GPU bound. It's really not as complicated as many make it out to be.

This. DCS is CPU dependent... to a point. Run a Reverb with its incredibly high resolution and refresh rate, and the GPU becomes the bottleneck. It's pretty easy to test this - simply lowering the resolution or MSAA on the Reverb and you get better fps. But this comes at a huge trade-off, with poor image quality. This is why I'm hoping the 3090 has the horsepower to power my Reverb at 60fps without making massive image quality drops.

Hey folks, thanks for your thoughts on this subject. Please don't turn this into a flame thread - I simply wanted to have a clear guide on the topic for newcomers to VR. To be very clear, AR is fine for some people, I'm not criticizing them at all. However, I was told for months that I'd configured my PC incorrectly, which caused me much stress and wasted time; if I can save just one other VR user from having this poor experience, it was worth the time posting this information. It'd be awesome if it could be stickied, as I think a LOT of new Reverb G2 owners will find this post relevant.

I'm surprised you get it on the Rift S - I thought it was low enough resolution that you could run at the same framerate as the HMD's refresh rate?

Thanks for your great feedback on my post. With the Reverb G2 on the way, a lot of players are about to encounter Async Reprojection for the first time. I thought it would be good for them to have an informed idea of what to expect, spelt out clearly and accurately.

There is a troubling issue in modern flight sims with regards to VR that rarely receives the attention it should, and resulted in me misspending AU$1200 on a HP Reverb. Asynchronous Reprojection (hereafter referred to as AR), also known as synchronous timewarp, asynchronous spacewarp and SteamVR Motion Smoothing, is a technology used to “mimic” extra frames when the user’s system cannot send rendered frames at the refresh rate of the headset. Many VR users refer to the issues caused by AR as ‘ghosting’, ‘judder’, ‘stuttering’, or ‘smoothness’. I’m sure there are some other names I’m not familiar with, but the vague way in which it is described, not to mention the huge variety of “fixes” to deal with it, are part of the reason I’m writing this. The impending launch of the Reverb G2 is also another reason. When I encountered the issue, I was told I hadn't configured my PC correctly, so spent weeks trying to solve this problem. Turns out there isn't a solution. So I thought I’d make a post about it to help clear up some confusion, and potentially save some simmers from spending a great deal of money chasing a VR experience that currently does not exist. Note that I’m not a VR specialist, but I was a technology journalist focused on enthusiast PC gaming hardware for 20 years with a passion for VR. I have a well-configured 9700K w/ 2080 Ti and all the trimmings, and I can’t run the Reverb without this issue. I apologize if I make any inaccuracies in this post, please feel free to clear them up with your feedback. A huge thanks to Aurelius, who is far more knowledgeable in this field than I am, for reading over this post and correcting the many inaccuracies! This guy knows his stuff! Please, please note that I’m not disparaging you if you are a VR user and enjoy it – lucky you! I wish I didn’t notice this problem, as VR is an incredible experience; however, I think if you’ve played other VR titles where AR is not used at all, it’s very hard not to notice the AR side-effects. Having said that, as evidenced by this forum, there are DCS VR users who do not notice AR, even in DCS. Whether you’ll be able to notice it or not can only be tested when you do put on your chosen HMD. I do believe it will become more of an issue with the Reverb G2, though the RTX 3090 (and possibly 3080) may have the requisite horsepower to overcome this problem. AR is generally not visible on “through-the-lens” videos because of the manner in which cameras differ from the human eye. Because modern graphics cards lack the ability to send the requisite rendered frames to higher resolution HMD’s like the Reverb and Valve Index, AR becomes much more prevalent. You will know if DCS is running AR if the in-game fps counter does not match the HMD's selected refresh rate (which is 90Hz for the Reverb Pro and G2, though this can be lowered to 60Hz). AR automatically kicks in when the PC supplying the frames can not match a framerate equal to the refresh rate of the HMD. If you’re running a lower resolution HMD than the Reverb, such as the Rift or Vive, and can actually match your PC’s outputted framerate to the HMD’s refresh rate, AR is not used and will not appear as an issue at all. AR was conceived to be used in rare instances when the PC “hiccups”, and can’t match the refresh rate – it absolutely was not designed to be used 100% of the time, as is often seen in flight sims. This is why every other VR game is designed to ensure that the frame rate matches the refresh rate. So, what actually is AR? It is a mathematical method for using coordinate transforms in the form of matrix computations to deliver a class of motion interpolation technologies aimed ensuring a virtual reality headset's responsiveness to user motion even when the GPU isn't able to keep up with the headset's target framerate, and to reduce perceived input lag at all times regardless of framerate. Reprojection involves the headset's driver taking one or multiple previously rendered frames and using newer motion information from the headset's sensors to "reproject" or "warp" the previous frame into a prediction of what a normally rendered frame should now look like to account for the change in all six degrees of freedom (6DOF). "Asynchronous" refers to this process being continuously performed in parallel with rendering, allowing reprojected frames to be displayed without delay in case a regular frame isn't rendered in time, and is used in all frames by default to reduce perceived latency.” If your PC is outputting 45fps to a 90Hz HMD, the HMD needs to display 90 frames regardless. (VR requires higher Hertz because of the interleaving of two displays to fool your occipital lobe into believing depth is being seen.) Therefore, it uses a clever algorithm to match every frame from your PC with a fake second frame. This means that rotational image display (3DOF) remains smooth, even if the objects within the scene do not. Unfortunately, artificially creating 50% of the frames is not a miracle software hack that works without any issues. If it did, we’d see it in every game engine on the planet, 2D or 3D! Imagine being able to double your performance without needing the hardware to do so! So then, what do AR’s side effects, or artifacts, actually look like? For me, the thing most noticeable was the ground “jumping” or “stuttering” the closer I got. Even worse was behavior in dogfights or formation flying; other aircraft have a very weird “shadow” image layered over the true 3D model which lags behind them by about a meter, yet also hovers over the top of them. It’s rather difficult to describe, and unfortunately I can’t find any illustrative diagrams. Please post some if you can find them! As mentioned, some people don’t appear to notice AR’s artifacts. It also seems to be minimized if you’re looking at objects that aren’t moving across the scene quickly – such as the ground from high altitude or in slow-moving aircraft. The best way to remove its issues is to not use it – but to do this, your PC must output frames that match the HMD’s refresh rate. This can sometimes be achieved by lowering the graphical detail or resolution, but often it cannot. You can also forcefully disable AR by config tweaks in most platforms, but this usually results in the overall experience feeling less smooth when you rotate your head. (Any time the relative 6DOF velocity (dy/dx) is high, the distortion grows because the reprojection algorithm must account for the larger difference, which results in the viewer seeing things as “warped” or “torn”. That is why an object close to your aircraft (a tree, .. another plane in formation, etc…) is noticeably degraded while something in the vast distance (a mountain) appears to be unaffected. The direct mathematical reason for this is the coordinate transform projection.) I hope this description helps those who are struggling to understand how AR affects a smooth user experience. Also, if you’re looking to make the splash into VR, this is a rather large issue that you should be aware of. I implore you to road-test your chosen HMD in person before outlaying any monies, I certainly wish I had! Note that watching a through the lens video or hearing other people swear that they don’t notice it doesn’t mean that you will not (lucky you if you don’t!). I can’t speak to which version of AR is superior, be it Oculus’ ASW, or SteamVR’s Motion Smoothing, as I’ve only seen it thru the Reverb Pro, but I would love to hear the opinions of those who can. (Up until about 2018, the code from Oculus was superior to Valve. Gradually, Valve reworked their algorithms (some say, reverse engineered Oculus’) and they are now generally seen as identical.) If you notice any erroneous statements, please feel free to correct me. And don’t take it personally if you love the VR experience even with AR on; I’m not attacking you at all. I actually envy you, and hope to join you for a great VR experience when the 3090 releases.

Thanks again, I'm going to get the shadows rendered just the way I want with your help! As mentioned, my issue with the shadows is that when I'm flying in close formation, at a distance of around 20 feet between Tomcats, the lower-res shadow keeps swapping back and forth to a higher-res shadow. Note that I'm only using the internal view, so don't care about external. So, I'm thinking the best solution would be to make it so that hard shadows don't draw on other aircraft at all. How would I change the values for high shadows, as seen below, to do this? I'm still a little unclear on the values. Here's the default high setting: { --high 4096, {0.02, 25.0, 100.0, 400.0, 1500.0, 7000}, {0.02, 2.5, 25.0, 250.0, 1500.0, 7000} }, The first line is for "split distances outside", the second is for "split distances inside cockpit". What does that actually mean? Is it that the second line is used to determine where the shadows change when looking out from the cockpit? If so, the first value is a mere 2cm away? Then the next is 2.5m - which value is it that would determine the shadows on the other aircraft? The 25? Thanks again for your help - would be lovely if the developers interacted with us lol!

Thank you for the information - that's extremely helpful. I kind of was thinking this was how it worked - but this leads to a couple of new questions for clarity: 1.) What units are these distances measured in? 2.) Why are there two sections for shadows - one being Nevada, and the second being?

I'd like to modify the shadows on other aircraft, as when I'm flying in close formation with other Tomcats, the shadow is constantly swapping between the soft and hard/detailed shadow. I've figured out how to draw the more detailed shadow at greater distance, but my tweak absolutely destroys performance when there are a few aircraft in close range, such as on the SC deck. So I'd like to remove these detailed shadows as much as possible, without compromising my cockpit shadows. Does anyone know what the various values in the shadows.lua file represent to help me get this done?

Hey guys, I want to change one shadow issue - the swapping of the softer shadow to the more defined shadow on other aircraft? When I'm flying close formation, it's constantly swapping between the two on my lead, and looks horrible. Is there a way to make it only render the softer shadow? I figured out how to make the harder shadow draw at longer distance, but it absolutely kills framerates on the SC, where there are lots of aircraft in a close range.

pretty easy really - to lower my view I have get in an uncomfy real world position. It's vastly easier to just drop the seat, refuel, then return it to normal.

Add it to the list...

I've been practicing AAR with the KC135 in the F-14 instant action mission, and had no luck for hours. Then I did two things which saw me hook up and make a successful AAR straight away. 1.) I lowered my seat height so that I could see the fuel pod at all times. Without doing this, if I was looking forward all I had for reference was the fuel line, which is a nightmare to stay in formation with once joined. I bound seat height to a throttle hat-switch. 2.) I tweaked my joystick curves to have +10 for both X and Y (Virpil MongoosT with 200mm extension). I wish I'd tried these tweaks earlier, as it made refuelling SO much easier. Oh yeah, I also used Bailey's awesome voice attack profile for the Tomcat so that I could communicate with the tanker hands-free. Hope this helps, as the seat height, for me, was the key to staying connected.