twistking

This is really a non-issue. If it's esotericism, it's the least harmful kind of it. No need to discuss this endlessly. If @Art-J is correct, then this shows that the devs generally agree that it might be a good idea to delete. If their implementation indeed got broken again, my gut tells me, that we might not get an official response though But really, this one is not that important.

I agree for general technical issues of the DCS engine. For this particular issue about validity of shader cache deletion, i feel confident enough to say that it should be done by everyone as a first step of troubleshooting performance issues or visual bugs. It's in the same category as "do a repair/file verfification" and "disable your fu**ing mods". If you do encounter a problem, it's just good practice to do so and contrary to some other suggestions that get thrown around in these forums, it does not come with any risk of adverse effects, apart from the slightly longer loading times on first flight.

that would be great. hopefully we'll also see speech recognition to manage communications with AI agents.

Can someone check if external aircraft lights are visible over greater distances now? On 2.8 an aircraft in full disco mode would completely vanish into the darkness after less than a mile or so... With refinments to LOD and spotting, maybe the devs have finally revisited this longstanding problem?!

That's interesting. But i assume that the Nvidia Shader Cache has nothing to do with the precompiled data library in your DCS user folder. I would guess that it's more about VRAM usage for such data?! I'm out of my depth here... I'm with DCS for over a decade now and i think i only once had issues fixed by deleting fxo/metashaders. I think it's still decent advice to do it when troubleshooting, simply because there are absolutely no adverse effects and it's easy to do even for beginners. I do it after every major update... but i'm one of those that does a steam game file verification from time to time just for peace of mind... That's kinda true. I think the advice is not "bad", but there is already a lot of guessing and confusion in this community about technical issues that those semi-correct takes muddy the water even further. It's a bit of a pet peeve of mine: It's the combination of an audience that is less "gamer" (so maybe slightly less hardware knowledgeable), an application that is notorious for having bugs and performance issues and - most importantly - devs that don't communicate about the technical details and challenges they face. As an example the ARMA series of games has always been difficult ton run well and was plagued by many issues (performance eventually got "solved" by modern processors just bruteforcing through it), but the devs were always very engaged with the community about those issues. Not only would they actively try to give suggestion in the forums, more importantly they would communicate - often quite extensively - their challenges with the engine and would provide the technical info for users to better understand the potential issues they'd encounter. This was very helpful to grow a community that has a good understanding of the technical foundation. In comparison the DCS community seems to be more occupied with superstition, occult rituals, placebos and half-truths...

Well, i agree, in that i also never saw performance issues caused by shader compilation, but it's very obvious that DCS loads longer when it does recompile the shaders. Also deleting fxo/metashaders was from time to time suggested by the devs/mods AND it's obvious that shaders NEED to be recompiled after certain changes in the renderer, because outdated code could plausibly cause issues. It's not DCS only: All modern PC Games have this potential problem (console games often have their shaders precompiled afaik, because hardware is a known). Normally it's just a non-issue, because games would just trigger a recompile themselfes, if a patch or user settings requires it. I can think of a few other games though, that in patchnotes often have the suggestion to delete old shader files either as a precaution after each update, or as a common solution to visual and performance issues.

i think we do not loose much here: setting up more than two mavs in a single approach seems more like circus-trick to me. you'd probably risk flying right into enemy AAA fire for being heads down too long before having set it all up...

Don't know if i get what you're saying. Why would there be no shader compilation after a fresh install. How else would the neccessary shader code be compiled? I only wonder why DCS cannot simply delete and rebuild shaders by itself after each update. I thought this would have been a feature since 2.7 or 2.8 ...

i'm pretty sure that the video's author is wrong about DCS compiling shaders during flight. On deletion of fxo/metashaders the shaders get compiled during loading: I never experienced any form of shader-compilation-stutter during gameplay and i'm on a low-end system. The idea of DCS needing many sessions over multiple days to rebuild the shader cache is bonkers, honestly. Also i would recommend deleting the folders (fxo/metashaders) instead of the files. It's not super important, but just good practice - to make sure you also get hidden files, in case there are some. DCS does not need to "repair" itself upon deletion of those folders: They will simply be created when the new files get written.

Check your RAM usage with windows task manager during mission. If DCS does not fill up your RAM completely after or during a more complex mission, you will not see any benefits from more RAM. If DCS creeps towards ~30 GB, but the whole system keeps being responsive when alt-tabbing and/or directly after closing DCS, you're probably still fine, but might see some minor improvments in certain cases. If however DCS takes ~30 GB and your whole computer gets less responsive when alt-tabbing, or when your computer takes a few seconds to be responsive again after DCS closes, you will most certainly see benefits with more ram. Another indicator that you "need" more ram, would be that DCS performance gets worse, the longer you play. If you however see DCS perform badly in certain situation, while DCS RAM usage sits well below 30 GB then again RAM size is NOT the problem! I would guess (but that's really just a guess) that for Singleplayer you should mostly be fine with 32 GB. It depends mostly on the complexity od the scenario. Only way to know for sure, is to check RAM usage yourself.

Did anyone notice any improvments to the distance to which aircraft external lights are rendered? It's my biggest pet peeve (i'd argue, it's more than a peeve actually, since it does have a significantly detremental effect on gameplay and immersion for flying at night, especially with older aircraft) and i cannot check for myself, since i'm a very stable guy. Anyhow, have you noticed improvments? Ext lights rendererd out further generally? Hornet's ext lights being not being huge glowing orbs anymore perhaps? Improvments to spotting of carrier meatball etc. ? ...

I see. If you say that 72fps with ASW (so basically DLSS working with 36fps) does not make artifacting notably worse than true 90, then this means that my point holds no value in practice. That mean that those artifacts come from lack of spatial information instead of temporal. Not very intuitive, but i have no reason to doubt you. I'm still quite confident that low FPS will introduce additional artifacting on image details that change details faster than can temporally be resolved on lower FPS, meaning no effective motion vector can be build from it. F.e. mechnical rotary counters changing digits... In these situations it seems logical that DLSS could even make details worse compared to the (low) base resolution without upscaling. I would be interested in seeing how DLAA and DLSS compare visually, if you set your high pixel density for use with DLSS and compare to DLAA not with native res, but with a resolution that gives similar fps to the DLSS setup. Or in other words: DLSS with high pixel vs DLAA with pixel density slightly higher than native to matching fps.

Don't know my frametime (and can't test right now), but i exclusively fly hand-made SP and Coop scenarios optimized for not punishing the CPU too hard. Also im on a monitor currently (no VR) and while i can hold 60fps mostly even on decent settings, i'm sure that i'm CPU bound 90% of the time. I would guess that this would mean that my cpu frame time is roughly 1/60 of a second? Or do i have my math wrong? I mean my CPU is over 10 years old. I find it astounding that DCS runs so well honestly while looking good (but again, no VR, 1080p, only smaller scale missions).

Missing my point you do, young padawan! DLSS creates a high resolution image (and is technically able to create perceived higher res than native in scenes with predictable or constantly slow movment). This is done by subpixel jittering an temporal accumulation of information, so the image gets refined "over time" - or more precisely over a sequence of sequential frames. Therefore in the same scene with moderate movment, double fps should mean more clarity, because a good estiamtion can be made in half the time. Additionally on higher fps, changes in the scenes are more gradual, so the guessing is easy and the algorythm should have an easier time "understanding" motion of the scene. Motion vectors only help little, when you have massive difference between two frames, because of fast relative motion. Also whenever the algorythm does a happy accident (which happens everywhere, all the time... to some degree) the resulting artifacts should be smaller when relative motion is slower, or when the difference between sequential images is lower, which is the case - again - on higher fps. Now imagine what happen if you have low FPS - so more guesswork for DLSS and potentially more visible artifacts to begin with - and now introduce another algorythm that seperately from DLSS interpolates between sequential frames to create a synthetic frame from two sequential frames that are already potentially impaired by motion vector guesswork. It's an unholy alliance of two "AI" systems playing a game of telephone. This is all a purely theoretical take on DLSS. I can't try it for myself unfortunately. I'm also not implying that your observations and personal experiences are wrong. Maybe the effects i describe are not that relevant in practice, but they must be technically true. Maybe keep an open eye for those aspects and keep an open mind when doing testing and tweaking. FPS must have an effect on the quality of DLSS (maybe not a relevant one though). I still think it's worth keeping in mind. I also strongly suspect that users reporting very bad image quality with DLSS do not actually stumbled on the limitations of DLSS, but encountered implementation bugs, run bad settings or similar. Also some issues reported would clearly be solved by doubling FPS: For example the mechanical digits in analogie gauges being blurry when changing the displayed numbers. On 45fps there is simply not enough temporal resolution to resolve that motion - or for the "AI" to "udnerstand" it. Again. I'm not saying that DLSS is good and your observations are bad. But completely denying the effects of FPS is not helpful for troubleshooting and understanding artifacting - even if the actual differences between low and high FPS are mabye indeed smaller than i would expect from a purely theoretical standpoint. Hope that makes sense. Thanks!

Thanks for the reply... although i still strongly disagree with the chart... or with the validity of the chart regarding DCS. That doesn't mean that i don't believe you getting very good eprformance from your CPU. But there is not neccessarily direct causality between the chart and your experience. By the way - and i'm not trying to be mean - your post sums up the "problematic" nature of anecdotal evidence in this forum regarding performance of different pc parts: You have a great experience with your CPU, but that prooves nothing for those who want to truly understand how different parts perform in comparison to each other. MS Flight Simulator is not a good stand-in for DCS. 1080p flatscreen is not a good stand-in for HQ VR. Your feedback is still valuable though!

i did not post that chart, so don't have the power to remove it. would also be rude;) thanks for your reply. with my new pc i would want to aim at a similar experience to what you describe. i'll go cheaper on the GPU, because i can easily upgrade later and will use lower pixel density or more aggressive DLSS until i can afford a more powerful GPU. But that's the only real compromise i'm willing to make. I'm planning to use the CPU for many years (i'm still on a highly overclocked second gen intel i7-860... well, yeah...), so it should have some headroom ideally. My big concern is: Are you CPU limited or GPU limited with you setup and settings? I would like to go big on CPU and RAM speed to get a system that is potentially capable of high quality, 90fps VR (disregard GPU limitation, where i'm willing to make a compromise). It's of course difficult to see how CPU requirements will change with the dynamic campaign (I'm strictly COOP PvE and SinglePlayer, so i'm more concerned about that than those super complex public MP missions), but i would hope that ED can put AI on different threads before the dyn campaign arrives and even creates some sort of AI instancing to help with those big, dynamic scenarios. TLDNR: Did MT and latest gen CPUs "solve" DCS HQ 90hz VR?

known issue. i think it's the combination of SSS and DLSS/DLAA. try disabling one of them.

DLAA cannot be on when DLSS is on. That's mostly semantics. It's not that DLSS utilizes DLAA: They are basically the same thing! Technically it would be most correct to have "DLAA" as the highest quality setting of DLSS: DLSS without upscaling. I think the way it is layed out in the DCS UI currently is good though. The way i described would be technically more correct, but probably less intuitive for most users. No. See above.

Ok. Thanks. Regarding the 90fps: I think it should work with your 4090 (yeah, maybe not with 200% res). My hypothesis is, that 90fps will make DLSS much more usable, which would give you the needed boost to performance in return. Even if you decide that 45fps and DLAA feels better in the end, i would be interested if you feel that DLSS is less blurry on 90 than it's on 45. For Science!!!

Nice! That sounds promising. What are "TAZ, Barthek? Mods i assume? Did you also fly in larger, complex missions? Do i understand it correctly that with 5.1ghz you only have 5ms cpu frametime, so theoretically enough headroom to easily go for 90fps (disregarding GPU)? Have you tried aiming for 90fps VR and using more aggressive DLSS to get there? You should try, if you haven't. I'm very confident that DLSS will look way better on 90fps, because it basically feeds off temporal information, which 90fps gives plenty. On 45fps the difference between sequential frames is rather large, so more guessing involved. On 90 it should have much less ghosting and better perceived resolution and clarity. In theory: I'm on a potato pc and can't test for myself;)

well technically yes*. i'm really not an expert on this, but from my understanding DCS is a bit inefficient in its stereo rendering, because the CPU prepares two scenes completely seperately, one for each eye (obviously), while more efficient approaches would allow the pipeline to share much work needed for each eye, so that stereoscopy would not double the workload for the CPU (one scene with two views, instead of two scenes). Don't quote me on that though, i'm just an interested layperson. What i wanted to express is that DCS's VR pipeline is not the most efficient and that that is most probably the main reason for the difficulties of getting optimal VR results. *There is technically a way to make stereographic 3d from a single viewport render, by using the z-buffer and doing some post processing with it to create a stereo image. That gives you an image with depth, but it's a little bit fake and does not look as convincing as "true" stereo rendering. I think it's dirt cheap to compute, but there must be a reason why it never really caught on.

it's a public forum and i did not know, those were not mine to answer. your sixths question is a rhetorical question, is it not? honest question (english is not my first language). ps: i begin to suspect that most of your questions were rhetorical.

1. I think not possible yet, but ED stated that an interface for external apps is in the works. I assume this will allows what you are asking for. 2. I guess you are not suppose to change those settings. The options might be visible however, because you can build your own "virtual" radio. Therefore it might be handy to see how the other are configured. That's my guess. I was surprised to see those settigns exposed in Matt wagner's video about it. I'm not on 2.9 yet, so can only guess really... 3. Don't know. What settings would you want to change? 4. I don't know how SRS does radios for FC3, but i think you are supposed to use those "virtual" radios (see #2). Either copy one of the existing ones, or create a completely new one with custom parameters. 5. I think it's not possible at the moment. I just close the interface and use the cockpit-displays... works for organized coop... or at least for me... 6. Won't try to answer rhetorical questions.

Thanks for the reply. I would want to point out however, that (Microsoft) Flight Simulator is not necessarily comparable to DCS. VR in DCS is a bit of a diva concerning performance, not because virtual aircrafts are involved, but because the application bruteforces inefficient stereo rendering over a thinly multithreaded rendering thread carefully balanced ontop of a heap of vintage spaghetti code. The only valid similarity between the two - i would argue - is that they are both VR compatible games that are easily CPU bottlenecked.

Hello, anecdotical evidence from reading forum signatures suggests, that AMD Zen4 CPUs are quite popular with DCS players these days. I know that current gen AMD CPUs are generally considered very good CPUs for gaming and often beat intel 13th (and probably even 14th gen) in gaming focused benchmarks. However, in some games Intel is stronger and generally speaking Intel CPUs seem to have slightly better multicore performance at a similar price point. I'm currently deciding with which to go for my next build, which i want to specifically build around DCS VR. Are there any hard facts which CPU might be better? I've read many adulations from new AMD customers, but they of course will always only compare the new AMDs to an Intel chip that is several generations older, so no comparison really. Do we (the DCS community) have any reliable comaprisons on how those different chips compare to each other? Thanks.