St4rgun

Actually ED shared progress on this topic at the very moment they announced it can be expected in Q3 2021. Right now the "we are not ready to share any news yet" standrard reply template IS the biggest news: it means it's not ready, because Q3 is over. The community never wanted to write the ETA for the devs, it would be insane. They're just asking for rough estimates. Sometimes ED gives estimates, sometimes they misses them. But it's always much better to correct yourself and give a new estimate according to your best actual knowledge on the topic. Not an exact date, but a ballpark. Why? Because it's the polite way if the community asks for it. I'm sure, that responsible people like us who are working as developers themselves on different projects in the world won't be upset if a deadline is missed and should be corrected. But the radio silence and lack of communication is never a good path.

Well, at least we can hopefully erase the last option from my previous list. Unfortunatelly not too many of other options got harmed during BN-s answer... It is still mysterious why there is no development reports on this topic like for the modules (i.e. AH-64D) about the progress. Strange. This could be some form NDA or something. Anyway, if the team is "not ready to share any news yet" then we only left with the hope for the next big update. Then the next one. And so on.

It's not only about the fps increase. It's about the possibility to refactor the whole codebase, possible to change back to forward rendering which could solve the AA bugs of DCS in VR. Eliminating the jaggies and shimmering, the lighting errors, the rendering bugs of transparent textures AND fps increase. The Vulkan migration means completely rethink and modernize a decade old engine which is not built from the ground up with VR in mind. It WAS a very important decision for ED in the past to go the Vulkan way while also working on the multithreading optimization in parallel. I'm pretty sure, that when it will be ready this will change the VR experience in DCS a LOT. Surely, it's az enormous task, so it's impossible to estimate the correct ETA at the beginning. But they are working on this migration process for more than TWO YEARS. During that time they had the possibility to fine tune the ETA. They had to think about something when they set Q3 2021 in the January newsletter. It's not fair to say "not raise the expectations too high" when the newsletter says this: "Vulkan API and Multi-threading With the integration of these libraries, DCS will benefit from improved performance throughout, including in virtual reality. This has been a substantial task to develop and integrate into our codebase, but we are nearing the end of this journey, and expect it to be available in the 3rd quarter of this year." All right, we are expecting. Now what? How much more time should we expect? about a week about a month maybe a quarter possibly half a year about a year years no one knows right now, but maybe never can be done because of unforseen issues The devs MUST be able to choose one of the previous options. I'm sure they can. The second decision is how and when to tell it to us. Please.

Absolutely. As I recall all previous new things HAD some preliminary news before deployment regardless being new features / modules. I mean like the sneak peak videos now with the AH-64D, the pictures and videos of the new clouds before 2.7 and even news about the lighting changes before 2.5.6 back then. Now I feel that we got only silence. Not a single demo video, not an image, not a predicted framerate data or any other new feature list. Nothing about Vulkan specific, which can be judged and compared to the previous engine performance / quality. But why? Meanwhile the last predicted ("planned") ETA of Q3 2021 silently passed... IMHO it would be unprecedenced for ED to give the community the engine migration as a well kept secret gift without preliminary hints. What's in the background? I'm a grown up person (nearly 50), but I'm waiting for Vulkan as a child waits for a christmas present. My expectations are enormous.

The current situation with GPU market implies that we'll be unable to buy any strong GPU for a reasonable price in the next year or more. This is really sad, but at the same time it's the BEST moment right now to help the users with next gen VR headsets (like G2) to be able to squeeze all the available horsepower from current hardware. Optimization is a MUST these days. If only the long journey of Vulkan reach to the end right now then it would be heaven for DCS users. Let the VR users also experience the "OMG" factor of this extraordinary game in its full glory.

Just a funny coincidence: what will be on October 13-14? https://www.khronos.org/events/vulkanised-2021 FUNNY.

It was mentioned by the devs that Vulkan development has good progress. It was back in 2018... Since then in the 2021 january newsletter it was mentioned to be PLANNED to release in Q3. Now Q3 2021 practically ends. Not a single hint or trace about Vulkan yet. At the beginning of 2020(!) this interesting conversation surfaced on reddit: Q: "Miss Perederko, Will DCS continue to use Deferred Shading with the full migration or go back to forward rendering?" A: "I will not give you a direct answer and keep it for internal use. However, Vulkan is not linked to a particular way of rendering. We'll take advantage of that freedom to optimize the performance and deliver a better picture. Kate Perederko, COO ED" I understand that for every project to give feasible timeline estimation it's advised to apply Murphy's law: get the predicted development time, multiply it by 2 and switch it to one step bigger scale (i.e. 2 weeksTM job will need 4 months). Now we are talking about YEARS of development. In january they had to be some kind of "certain" that implementing Vulkan should be finished in 1-2 months to be able to "PLAN" Q3 for deployment and write it down in a newsletter. But we still only got silence, which induces frustration. I'd also be the happiest to know, that the next possible BIG update (somewhere in October) will include core tech migration as well including Vulkan, but if the news is "bad" like "forget 2021, we should wait at least for 2022", then let it be. Just say us a word. Please.

Unfortunately I still can't confirm this observation. With my G2 and MSAA 2x I tried every possible roll angle with my head left and right with small increments but nothing happens with the shimmering. Of course when the majority of the horizontal/vertical lines became diagonal at around 45° then the pixelation (jaggieness) of the lines will be smaller due to pixel arrangement, but that's natural phenomenon of the aliasing effect. What is interesting that in the world not EVERYTHING is aliasing (shimmering), like the buildings (which are part of the map), roads, cars on the roads, rivers, shoreline, etc are FINE without shimmering. But every object which seems as transparent textures like electricity pylons, lighting poles, fences and distant objects like airplanes on the ground or objects/buildings which are NOT part of the original map, but placed in the editor (like the subbase complex in the Oilfield campaign of the Mi8 module) shimmering badly from the distance. At least for me the new clouds became MUCH better since introducing them in 2.7, but there's still some aliasing at the distant clouds at the horizon and every objects aliased which are seen BEFORE the cloud layer. But for me the heavy shimmering and jittering is gone (at least with some presets). I observed that the transparent AA settings in nVidia panel CAN reduce shimmering but only for some objects: the contours of the trees and the painted yellow lines at the airports are NOT shimmering with that setting. Q3 is near its end, so I'm eagerly waiting for any news about Vulkan as always...

Looking at the new Mosquito trailer I became excited but very sad at the same time. If only I could see this quality in my G2... It actually seems a rendered high quality animation instead an in-game footage compared to the VR picture quality of DCS right now. Not even a small trace of shimmering, everything is fluid and perfect. Simply breathtaking. ED, at the very moment when your graphic engine will be able to deliver THIS quality in VR, then the battle is over with "other simulators" once for all. Please do your best to make this dream come true! Thank you.

There were high hopes in Vulkan implementation as a cure for performance issues in DCS here in the forums for years indeed. But I think about it as a great possibility to refactor parts of the rendering engine (which I'm pretty sure done before several times in the background, at least when 2.5 then 2.7 came out). As ED wrote about the aliasing of the distant clouds with 2.7's new cloud system as "known bug" which they are trying to eliminate, but did not done yet gave me the hope that they are rebuilding some parts of the rendering core completely while implementing Vulkan to handle the AA issue once for all. Right now the last official words from ED is a planned deadline of Q3 2021 for Vulkan to introduce, but without any further details. That's why I was curious about some of the PLANNED new functionalities, but I know we won't get any details unfortunately.

I disagree, it is possible to have some meaningful milestones which are understandable for the general public (i.e. lighting is tested and working, AA technology under testing etc.), or for at least for those players who takes care about the underlying engine's performance / feature list. It's frustrating that at the future sometime we will get the word "it is done, 3.0.0 is here" or something, but without any previous hints about how far we are from that point at least approximately.

It would be so nice to have some hints about the development status of the Vulkan engine rewrite. Nowadays the newsletters contain detailed info during the development of the modules (like the AH-64D), showing WIP screenshots and sharing useful and interesting descriptions. But the engine rewrite / improvement process is a big black hole. It is always at the "it will be ready, whet it will be done", and the "if will have any new information to share, then we will" state. The very futuristic approach could be some progress indicator bar with 100% meaning the introduction of the new Vulkan engine with all the new features, so we could see that currently the process is at 10%, 50% or 90%. Or maybe 1%, but it's unlikely because earlier they mentioned Q3 2021 for the estimated deadline. I can't imagine that the dev team working without milestones, it would be interesting to read about these as well.

I've found an interesting technique here (I hope it is allowed to link another company's tech here): https://www.livenda.com It seems it can eliminate the biggest drawback of TAA: blurring the distance. So There should be some modern AA technique which is compatible with VR and deferred shading. Another solution would be VRSS which is nVidia tech, so vendor dependent (https://developer.nvidia.com/blog/nvidia-vrss-a-zero-effort-way-to-improve-your-vr-image-quality/). The ingame SSAA would be really nice but at least for me it simple does not work: if I switch off MSAA and turn SSAA to 2x then the GPU load will be higher without visible antialiasing effects. I have really high hopes in the dev team. I'm sure they will sort it out somehow.

I made some interesting observations during my investigation of this AA issue. With 2x MSAA it seems that every edge which is part of an object is antialiased AGAINST THE SAME OBJECT perfectly, but NOT ANTIALIASED against other objects or the background. I have no idea how the rendering engine works but it seems that the antialiasing is done on object level before all the objects are assembled in the final picture. But why is this happening only in VR? Example: in the Mi8 cockpit try to examine one of the cyclics joystick, or any of the instruments' bezel. The bezel's outer edge is not antialiased against the left console blue body, but the instruments scales are perfectly antialiased (maybe the are textures, so that's why they are filtered perfectly). The extreme example is the engineers' console: the small white text at the switches aliases very badly when changing viewing distance (unreadable from the copilot seat). Maybe these small white letters are separate objects not simple texture. Meanwhile the MIN-MAX lettering of the throttle handle on the collective the text are perfectly antialiased regardless of the viewing distance. There are some reflective materials where the outer edge is absolutely bad: like the window heaters cables in the Mi8 cockpit frame: the outer edges are LIGHT BLUE (like they are antialiased against the blue sky), while having a dark background. Weird. I have not seen appearing and disappearing antialiasing effect at all. Some edges are always antialiased but others are not and they are shimmering constantly.

The headset resolution is only half part of the problem. Aliasing can and will occur at high resolution also, but this time the problem is not only aliasing of the edges, but some kind of flickering aliasing which shimmers a LOT. Without AA it's unevitable that some parts of objects on the scene will be smaller than the pixel size (subpixel resolution). In that case if the movement of the viewer or the object causes this size to be bigger than one pixel then this pixel is shown, but is hidden otherwise. This phenomemon called shimmering, flickering of object pixels. For the eyes it's really bad to focus on the important parts of the scene while a LOT of unimportant moving noise can be seen at the not-antialiased shimmering edges. If you don't use AA, then the filtering is done by your brain which induce fatigue. If you don't use AA but set the resolutoin much higher then the supersampling done the job of the AA while downsampling to the native resolution of the HMD. What is really strange for me in DCS, that 2x MSAA quality looks pretty much perfect for me using my G2 at only 50% resolution. Where it works, it works perfectly. But somehow some parts of the whole image is not antialiased AT ALL. Thats the problem. There are always edges here and there whose resolution is just horrible. Not too much, but enough to ruin the whole experience. If only we could get some 2x MSAA level of antialiasing done right for the whole scene every time, then I'd be happy.

All the choppers I own (UH1, Mi8, Mi24) renders the main and tail rotors differently for the left and right eyes at certain rotor RPM values. This phenomenon can be reproduced every time by stopping the engine and watch the rotors to slow down. The effect is same in cockpit and outside views. Main rotor approximate RPM values: 100% - 75% RPM -> both main and tail rotors are rendering fine, both eyes are in sync 75% - 50-48% -> rendering of the main rotor in the LEFT eye changes to show discreet blades instead of blurred ones, it is out of sync with the right eye; the tail rotor is still fine 50-48% - 25% - main rotor slowed down to switch the RIGHT eye also to discreet blades, so now it's back in sync with the left eye; tail rotor is still fast enough to be fine 25% - 18% - main rotor is fine, tail rotor slowed down, so the again the LEFT eye showing discreet blades of the tail rotors while the right eye is showing blurred ones (eyes out of sync) under 18% - both rotors are rendered fine, in sync for both eyes (showing discreet blades). This effect is really distracting, please correct as soon as possible.

This binding problem already surfaced in the last update, so please find a fix! 1 and 2 keyboard commands properly switch between pilot and RIO, but the hotas or joystick bindig does not work at all. I'm using VR (Reverb G2) if it is important. Never mind, I made the same mistake as sobe did: I forgot to bind the keys in the RIO section...

I have high hopes that the heavy aliasing of the new clouds on the horizon is not solved by ED yet because they already try to figure out the AA concept as a whole anyway. They won't allocate resources on solving the clouds aliasing alone when maybe a modernized engine is close with proper AA. I will be really disappointed if the present situation stay with us more than half a year...

First I was using my trusty Rift CV1 with my 2080 in DCS for more than a year and I get use to "ignore" the screendoor-effect of that HMD. I was pre-ordered Reverb G2 back in July 2020 and patiently waited to get any performance measurements whether I have to change my 2080 to a 3080 or not. But I was on the unlucky side: my 2080 died just a week before I got my G2 in the middle of December 2020. I immediately switched to the "spare" 1070 but try to imagine the comprimises I had to make to get it work with the G2 with reasonable framerate. DCS was a stuttering mess then. Because of the holidays it took more than a month to get my money back for the 2080 under the warranty at the end of January 2021. But during that period the VGA crysis already escalated with insane prices, so now I stuck with my 1070. The good news, that the latest 2.7 version DCS is pretty much fine with the G2 even with the 1070. You can see my settings above, the framerate is mostly 30ish which can be handled perfectly by SteamVR's motion reprojection (G2 set to 90 Hz). The single best setting was the forced DX11 mode in SteamVR, I think this helps a lot in my case, when the GPU is much weaker than my CPU (10700k). With all these settings I can play DCS with the G2 for hours, but only offline missions. The Mi-24 module is the only one with significantly lower framerate (20ish), so any optimization for that is highly welcome. Anyhow the immersion in VR is a blast with G2. I hope that in the future I will be able to reach a rock solid 45fps which is pretty perfect for me with my HMD. But the image quality difference between the pancake and VR version is night and day. All is caused mainly by the lack of proper AA. The "other simulator" uses TAA which is taxing on my 1070 but the AA is wastly superior. Unfortunately the clarity can suffer because of the blurring in the distance, so DCS is better in this field. Sometimes I switch my SteamVR and OpenXR to set my G2 to 100% res and see what to expect in picture quality. This is a slideshow, but "the other sim" is simple breathtaking, while DCS is really cool, but plagued by shimmering. I think if ED can finally implement proper AA then DCS will be paramount in VR. All we can hope that the arrival of the Vulkan is closer than we think.

One more vote for F5E (for me the Mirage is very illogical and complicated instruments-vise, the F5 is much more easier). Easy to fly, and a bit oldish, but clean, easy to understand systems. Do you fly in VR or on monitor?

I managed to play with all the nVidia control panel parameters just to find out that the followind settings REALLY DOES HELP with the AA issue at least with my Reverb G2: Antialiasing – Settings - ENHANCED Texture filtering – anisotropic sample optimization - OFF Texture filtering – Quality - HIGH QUALITY Texture Filtering – Negative LOD Bias - CLAMP Multi-Frame Sampled AA - OFF (causes graphical elements jittering if switched ON) Antialiasing Transparency - 8x SSAA I use ingame MSAA 2x which can be handled well even by my poor GTX 1070 (Reverb G2 resolution set to 50% in SteamVR, 1.0 PD ingame). The ingame SSAA 1.5x or 2.0x does NOT do anything on the picture in VR (at least for me), but it's taxing on the framerate, so it should be held OFF. With all these settings the yellow lines on the airports' concrete no longer shimmer at all. the contours of the trees are also perfect. Overall the image is much more pleasant than before, so it worth a try to check out. Unfortunatelly I still can't resolve the aliasing of the buildings' shadows, they are not anti-aliased at all. Also the edges of some objects are problematic (edges of some of the elements of the cockpit, like the HUD etc).

The problem is not the quality difference between TAA and MSAA, but the issue, that some of the objects on the picture is NOT anti-aliased AT ALL in VR, even with MSAA 2x and SSAA. It can be related to some transparent textures and can have correlation with the heavy aliasing of the new clouds at the distance. Somehow the horizon is what aliasing more but only in certain circumstances (the open water at the horizon anti-aliased perfectly, while the mountain ridge line aliased much more). I'm trying to classify which objects or parts are not anti-aliased properly.

Nowadays I'm waiting for the next "holy grail" open beta update week-by-week to see something magical in the development notes... With no luck. It's really sad to see a seriously very long list of every itty-bitty small feature's polishing (like the AIM-120 missile guidance fine tuning, or changing the fuse time of a specific weapon) every week while the FUNDAMENTALS of the visuals is not resolved in VR. I feel somehow that the VR branch is somehow neglected by the devs. I would really like to see the STELLAR image quality of this game in 2D to be able to get in VR as well. Devs, any comments on this topic? Can we have any small hint from the future development roadmap of the VR-specific development? Will the Vulkan API help us? Do we have ANY hope in the next months or years to change this situation? I'm using DCS in VR on a daily basis, but getting tired of this unresolved issue.

Right now unfortunately I have to use my spare GTX 1070 only (10700k, 32 GB) with the Reverb G2, so I have to use 50% resolution set in SteamVR. PD is 1.0 in DCS to be able to play with usable framerates. But in my case the framerate is not the main issure, I know the heavy GPU bottleneck right now, but I'm solely talking about picture quality. It does not matter if I set MSAA 4x with SSAA on and even 100% resolution in SteamVR so to watch a slideshow on my GTX 1070, but the picutre still shimmers heavily. Actually it seems like that some parts of the picture not antialiased AT ALL: like edges of the the projectors of the HUD, the contours of the trees, the painted lines on the airports's concrete, the horizon, the low clouds (this is a known phenomenon, so not count here), the shadows of the planes, some parts of the contours of the planes, the water shoreline etc. Shimmering EVERYWHERE, but in VR ONLY. Very annoying indeed. On 2D it's totally different, so I think something somehow does NOT work properly with the AA in VR. That's why I hope that somewhere during the edge engine change to Vulkan maybe the rendering process will be refined also to eliminate this issue.

Any news on the VR AA topic? Tons of new features emerge week-by-week but this fundamental feature seems quite neglected in VR. Am I doing something wrong?