Woona

One more run, this time across McCarran in Nevada where I usually test ground performance and anisotropic filtering. Definite preference to the theoretically applied ReBAR. Note the flipped order as compared to last post. ReBAR is a relatively new tech and we're only starting to learn about it, but we already know it's system-dependent. Since you only mention the appropriate flags, did you check that your hardware is compatible? - Install relevant BIOS and VBIOS updates, then update your GPU driver and enable ReBAR in your UEFI? Also note that with VR headsets we're often locked to a reprojection framerate or full framerate, so a bump from 70-80 in an average situation means you'll still only see 45fps/whichever rate is your reprojection.

Shimmering will always be more noticeable with sharpening, so I'd put my money on that being down to CAS (from your 'universal shader'). I've run quite a lot of tests with the VR shaders and can't say I've ever seen a degradation in shimmering/AA from it, nor do I know that there'd be a logical explanation to that. CAS however will most definitely make shimmering more noticeable, although cleaned up heavily with MSAA (read: I use CAS and MSAA 2x myself and shimmering is negligible, DCS PD 1.4 + Index HMD). Let me know how DDU and a DCS wipe works out for you.

Hi guys, really excellent work. Blown away at how quick someone found out about the profile flags. Here's my performance measurement on a single set of benchmarks on my regular Batumi run. While I don't see increases in maximum FPS, I see a pull in averages (94 -> 99) and minimums (60 -> 64) which are luckily both the metrics I want to see lifted with any performance benchmarks. The 99th percentile numbers are well within margin of error so no recorded differences there so far, unfortunately. I haven't given it the empirical test as I'm heading out the door soon, but I'll sit down tonight and add empirical data. Microstutters, dropped frames thru FrameView, yada yada.

I experienced the same issue for an extended period but regained my MSAA settings' efficacy after giving DCS a complete wipe from my drives and DDU'ing my drivers for my 3080, then reinstalling the latest driver from Nvidia. I'm not sure which of the two made the difference but one of them made MSAA return to DCS for me. DDU: https://www.guru3d.com/files-details/display-driver-uninstaller-download.html As for wiping DCS and getting a clean install, I uninstalled it as normal, wiped the DCS folder where it was installed and removed the DCS folder in my savedgames, then reinstalled anew. A somewhat tedious process but it's there. NVCP AA settings do not impact DCS. See https://youtu.be/dd6tsxmCosQ?t=2231 for measurements. Don't mean to harp on you, just trying to kill that misconception wherever it pops up.

This post just kinda popped into my head here tonight. I tried dragging my res down a tack, from 5332 x 5924 to 4768 x 5300. I use MSAA x2 anyway, so my supersampling is exclusively for spotting targets and making out distant details. With CAS that's obviously a lot better and more efficient, so I dropped my res from my usual beyond-diminishing-returns point to something more fair. The following is empirical, but I can spot trees and buildings precisely the same as before and can't say I feel like there's a difference between the pictures produced between the two resolutions. Can deffo recommend trying to drop your supersampling if you're like me and usually run at a pretty high SS. With this lower setting + CAS, I can increase my tree vis setting to 100% and still see a 5FPS increase to avg and 90th percentile, or leave things as-is and see an uptick all the way to 104fps avg: I'll poke around with it a bit more later, but if anyone else here who usually supersamples pretty high wants to try the same thing, I'd love to hear your impressions.

For now, there's no worthwhile anti-aliasing through Reshade that'll decrease shimmering on buildings or water without taking up the same amount of performance as DCS' own MSAA. So while you can get SMAA and FXAA to work, their performance toll (at a reasonable sharpness) will be higher and the image quality lower than just using 2x MSAA. I've tried so many combinations but it's just not as good as MSAA in DCS. In short: I'm not entirely sure what you mean by "new shaders in conjunction with this." Most other .fx files from normal Reshade builds work fine, and I'm using the VR Shaders mod obviously, works perfectly fine.

Yes, that's the correct link and file hierarchy. This should be the recommended download for anyone wanting to use CAS and color edits with DCS along with Performance mode by principle. No problemo. I'll try to add these complete performance tables wherever it's relevant, hopefully it won't get too annoying or in-your-face. Just want to make sure we're starting off on a good and objective note for what's surely going to be a standard VR recommendation moving forward. Fholger's work is fantastic and deserves to be treated right, I feel.

You can see my summary table from FV on my Caucasus standardized benchmark with Performance mode disabled and then enabled here: https://docs.google.com/spreadsheets/d/1ojDB6q4DHzFXC9plOsesBCqHYYnIGxJkKkw3CIwbVHc/edit?usp=sharing This can be caused by a ton of variables. Component overhead across our systems and usecases not the least of them. This has been replicable 3 separate times for me. To make sure I wasn't going crazy, I recorded these numbers one more time just now and uploaded the csv to Sheets so I could share them. That's the doc up there. Clean run-through to clear any unusual microstutters on first run, reload, benchmark, enable performance mode, reload, benchmark. In this run, I have several areas mostly sky-filled where I turn upside-down and immediately fill the viewport with lots of buildings. In these scenarios, the CAS consistently causes a quick stutter (see 15.04 Min FPS above). If the consensus is that it either brings no difference or a performance benefit, would it not be preferable that Performance mode is recommended generally? It seems odd that we shouldn't recommend it as a rule of thumb until we know more. I say this because I fear lurkers on this thread make the assumption that it makes no difference, inevitably leaving some passer-by worse off. Oh, and a side-note: I see some screenshots with folks running the entire suite of .fx, albeit disabled. The new build on Nexus from fholger contains an optimized CAS+color/contrast .fx and no other jazz, specifically made to be lightweight. It should show only CAS.fx in the reshade menu. Sorry if this has already been mentioned or whatever else, it's just notably more performant so, hey, worth mentioning. As for FXAA, it compiles fine for me but the image is blurry as all hell. Cockpit gauges in the F-16 become literally unreadable. That and SMAA combined made for an interestingly sharp and less aliased image, however upon testing I found the combination to have the same or worse performance than 2x MSAA. The latter produces a far better image anyway. You can see a quick through-the-lens example of a clean picture versus FXAA in my ReShade from my shitty stream here (33:19 vs. 35:00):

Just to be clear, with Performance mode, I see 0 performance hit with no added load beyond a 1-2% margin of error load measurement on my FrameView runs in my usual Batumi benchmark. It's only when Perf mode is off that I see anything. Really impressive stuff. ED does not like me so my inputs are null, understandably so, but for any others here it might be worth suggesting to ED directly that CAS be integrated into the game natively. It does not fall under the GPU-specific technology category despite being an AMD technology, as CAS is completely open source. Would be a huge benefactor to the average DCS VR Joe and the technology is right there and evidently very effective. Back to work I go.

I've got a video coming on it but it's on hold due to my work schedule. I don't have time to sift through the thread so TLDR is that the performance hit on first run is -3fps (68.930fps avg -> 65.080 fps avg) while minimums see a similar decrease (51fps avg -> 49fps avg). Maximums are more affected (85 fps avg -> 79 fps avg). More notably, I can see on my FrameView tables that there are numerous stutters and drops with the sharpening enabled, and I can feel them too. With performance mode enabled my FrameView tables are basically 1:1 with my DCS otherwise. Can't tell any visual difference between non-performance and performance. Really impressed with CAS. Also the updated version on Nexus has less overhead as measured from my normal DCS bench run. Measurably better.

@Goffik Thanks, definitely agree on annoying caps lock titles. I didn't think it could translate into titles just pointing to it being outdated, but I guess it just ticks the right wrongs, even just for version titles. I think I'll just throw them off my channel and live with it. @speed-of-heat I'll make OVGME/JSGME recommended in the next video after 2.7 comes out - at the assumption that you or kegetys are able to make these modifications work after 2.7 surely breaks them. Hopefully this will amount to less repetitive errors here.

@speed-of-heat Yeah, me either. I use JSGME every time, never failed me. Just can't say the same for some others. I can always experiment with recommending JSGME/OVGME in the next one and see if the number of not-working-comments increases/decreases.

Ahoy, so the reason the videos recommend as such is because I found some buddies (read: people who help with testing my routines externally before I upload a video) had issues with installing the mod with any effect using JSGME or Ovgme. I checked their directories myself to no avail. If it could happen in my relatively small group, the chance of it happening between several thousands of others is pretty large. Asking them to simply overwrite, the fail% was 0. I can't logically say why this occurs in some small margin of users, but there you are. It's been a while, so I'll give it another test externally using only JSGME or Ovgme and see if it works with the current versions of DCS. I can't say I've ever experienced direct installation causing any errors. Conflicting mods (or installing into Saved Games lmao) have been the issue in 100% of my troubleshooting cases so far. Hey, so asking out of genuine, unsolicited curiosity: What made you follow this video and not look on my channel for the updated guide? I'm trying (and I'm not trying to be aggressive, although my tone indicates otherwise) to wrap my head around how people aren't following a giant caps lock title and description. I keep those up because YouTube's algorithm absolutely digs my updated guides 6 feet under unless I reference the old videos over to the new ones, but I might just nuke the old ones all-together at this point and just live with it.

Thanks a lot. Again, if you have a list of corrections or any additions to the guide, please let me know at any time.

Hi. I'm the guy who created the DCS: VR Optimizations guides on YouTube. I really apologize for any incorrect guidelines given in the video. My schedule at the time was - mildly speaking - overfilled as-is. I made the guide despite this due to countless people in-game complaining about performance, especially after 2.5.6 released. I made the guides to raise awareness to this thread and its benefits, so more people could make use of it. Simultaneously, my hopes were that ED would recognize the activity of this thread and its topic, a way of signalling just how eager VR users are for anything at all to make DCS more reasonably playable for them, at least performance-wise. This is why I've been trying to redirect as many people as humanly possible directly into this thread when they've shown feedback of any kind towards the shader mod, or even the guide in its entirety. I'm trying all I can, both in the video itself, the video's description, the pinned comment and in each positive response, to ensure Speed's and Kegety's credit for this work as I'm obviously nothing but a conveyer. I hope that part's being carried out sufficiently. I'm currently on my summer break and won't be back in the office nor doing research for another 3 weeks. If anyone wants to give me a list of corrections on the latest video (already including terrain file removal on that list), I'd be very grateful. I want to continue improving the guide with everything the community has to offer into a single package that anyone can easily refer their friend or acquaintance to. I know you aren't taking much credit for your work Speed, but the truth of the matter is that this mod would most likely have gone for the history books if it hadn't been for your voluntary contribution. Thanks, man. You deserve it. I hope you're okay with me continually trying to raise awareness towards the thread and mod here. I usually do not post on the forums, so I apologize for the late involvement. I can't promise I'll post anything after this, but trust that I follow this thread closely regardless.