VR experiences IL-2 Korea

[FuriousMeow]

On 7/12/2026 at 6:28 PM, Aapje said:

We've had polls on what headset people use, and the Quest was by far the most popular.

You are making the classic mistake of thinking that everyone has the high-end gear, just because those people are most vocal.

Which is a rather useless statistic since the survey only counts a headset if you have it connected to the PC at the time of the survey. I've done several surveys and they never counted my headset.

I'm not making any mistake. The Quest has a few tiers. It wasn't long ago an x80 Ti GPU, the second to top tier GPU, was $650 - around one of the Quest models. A 570 Ti is still a high end GPU and costs about the same as a top Quest.

As far as the Steam surveys, they can still detect the device. Just unplugging it doesn't make it go away, open up device manager and select "hidden devices" while unplugged - still there.

Anyway don't care - we can double Steam's survey to 4%. Still a very small amount, and I can guarantee VR isn't more than 10% for GB, or will be for Korea - and that's extremely  generous.

Edited July 14 by FuriousMeow

[Panzerlang]

Classic Catch-22...low number of good VR games, less reason to invest in VR. Low VR take-up, less reason for devs to make good VR games. Ouberos or something.

None the less, Korea has VR and there's hope it can yet be optimised.

[FuriousMeow]

53 minutes ago, Panzerlang said:

None the less, Korea has VR and there's hope it can yet be optimised.

Agreed, any performance improvements are a plus for all.

[YoYo]

3 hours ago, FuriousMeow said:

Still a very small amount, and I can guarantee VR isn't more than 10% for GB, or will be for Korea - and that's extremely  generous.

That's still 10%, and it's not a small group at all. Is it better to have $900,000 or $1 million? In the case of combat simulators, this group can be larger than in the case of civilian simulators (maybe 15-20% or more). The IL-2 GB reportedly has around 800-900k users, which translates to close to 100,000 VR users. It's definitely worth pursuing such a large market. For some developers, selling only 100,000 worth of game copies is a pipe dream. As for Steam results, I wouldn't rely on them at all, perhaps just a rough estimate to get a point across. Simulator enthusiasts largely choose to own simulators outside of typical channels. Directly from the manufacturer (MSFS > XBox, DCS > directly, IL-2 directly). I'll admit that my friends and I rarely buy aviation-related titles through Steam. Instead, they buy them directly from the developers – they get faster support for the title (patches, content, etc.), and this is how they show support for the developer by ensuring the continued growth of the simulator market (Steam imposes a very heavy financial burden). Personally, I don't own any simulators through Steam.

Some survey results. MSFS around 10% https://static.navigraph.com/survey/FlightSim_Community_Survey 2026.pdf

DCS - 24%https://dcsworld.pl/ankieta-polskiej-spolecznosci-dcs-world-wyniki/10972/

Of course, this is still just a preview, but one thing can be said for sure: it's still in the 10-20% range (I bet it's closer to this 20% in the case of combat simulators), which can be a considerable sum for developers, not to mention loyalty and future DLC. This is as important a part of the simulator market as support for a specific GPU, sound system, etc. Let us also add that this market, although slowly, is still growing (for example, Pimax is reportedly selling 30-40% more headsets than a year ago, new ones are also coming out, such as Steam Frame, Meta is also working on something, etc.). To ignore this was to shoot yourself in the foot especially in flight simulators or simracing, where it is also a very large market.

Edited July 14 by YoYo

[AEthelraedUnraed]

40 minutes ago, YoYo said:

That's still 10%, and it's not a small group at all. Is it better to have $900,000 or $1 million? In the case of combat simulators, this group can be larger than in the case of civilian simulators (maybe 15-20% or more). The IL-2 GB reportedly has around 800-900k users, which translates to close to 100,000 VR users. It's definitely worth pursuing such a large market. For some developers, selling only 100,000 worth of game copies is a pipe dream. As for Steam results, I wouldn't rely on them at all, perhaps just a rough estimate to get a point across. Simulator enthusiasts largely choose to own simulators outside of typical channels. Directly from the manufacturer (MSFS > XBox, DCS > directly, IL-2 directly). I'll admit that my friends and I rarely buy aviation-related titles through Steam. Instead, they buy them directly from the developers – they get faster support for the title (patches, content, etc.), and this is how they show support for the developer by ensuring the continued growth of the simulator market (Steam imposes a very heavy financial burden). Personally, I don't own any simulators through Steam.

Some survey results. MSFS around 10% https://static.navigraph.com/survey/FlightSim_Community_Survey 2026.pdf

DCS - 24%https://dcsworld.pl/ankieta-polskiej-spolecznosci-dcs-world-wyniki/10972/

Of course, this is still just a preview, but one thing can be said for sure: it's still in the 10-20% range (I bet it's closer to this 20% in the case of combat simulators), which can be a considerable sum for developers, not to mention loyalty and future DLC. This is as important a part of the simulator market as support for a specific GPU, sound system, etc. Let us also add that this market, although slowly, is still growing (for example, Pimax is reportedly selling 30-40% more headsets than a year ago, new ones are also coming out, such as Steam Frame, Meta is also working on something, etc.). To ignore this was to shoot yourself in the foot especially in flight simulators or simracing, where it is also a very large market.

The Navigraph survey is heavily geared towards flight sim enthusiasts rather than towards casual gamers. I don't think it's much more reliable than the Steam Survey, except perhaps for providing an upper bound.

The same goes for your statement that "Simulator enthusiasts largely choose to own simulators outside of typical channels" and "buy [their games] directly from the developers". I very much doubt that IL2 has sold more copies through their own site than through Steam.

No matter what statistics and surveys you take, they're all skewed. Yours are just skewed in a different manner.

[Qcumber]

@LukeFF Do the devs have any VR stats that they would be willing to share? I've accidentally started a wildfire here. Only you can put this out. 

[GiftGruen]

I'd maybe add the hypothesis that VR users who obviously are willing + able to invest quite a lot of money in their hobby, buy more DLC content and not always wait for sales actions, so the yearly amount of $ they leave at IL2 is bigger in average.   For a company it's not only the number of users, but also ( and even more ) the number of $ a user is investing.

So 20% of VR users maybe mean 40% of cash.

Disclaimer: Of course I have no statistics, it's just an assumption.

Edited July 14 by GiftGruen

[Zakson85]

Let me offer my own take to back up your point. I own all the expansions for IL-2 GB/DCS, as well as the premium version of IL-2 Korea. I fly exclusively in VR; without it, combat flight sims might as well not exist for me. I’ve spent a fortune on this "crazy" hobby of mine, and I’ll keep doing it-even though I’m part of that small minority who fly in VR. 

[Talisman]

1 hour ago, Zakson85 said:

Let me offer my own take to back up your point. I own all the expansions for IL-2 GB/DCS, as well as the premium version of IL-2 Korea. I fly exclusively in VR; without it, combat flight sims might as well not exist for me. I’ve spent a fortune on this "crazy" hobby of mine, and I’ll keep doing it-even though I’m part of that small minority who fly in VR. 

Very much agree.  No VR would mean no flight simulation hobby for me.  Need force feedback too!  Flight simulation feels like a dead aircraft experience to me without VR and force feedback, including haptic. 

Happy landings,

Talisman

Edited July 14 by Talisman

[Aapje]

13 hours ago, FuriousMeow said:

I'm not making any mistake. The Quest has a few tiers. It wasn't long ago an x80 Ti GPU, the second to top tier GPU, was $650 - around one of the Quest models.

It's hard to take you seriously when you make claims like this. The last time a x80 Ti was $650, was the 980 Ti over 10 years ago. That CPU is now way too weak for VR.

The GPU market changed a lot between then and now, and comparing the GPU pricing of that period to pricing of VR headsets today, seems like an attempt to cherry pick data points to support your position, not to make an accurate analysis. 

Quote

A 5070 Ti is still a high end GPU and costs about the same as a top Quest.

Yes, $900 is about the same as $619. 🤔

[Dash,Polder]

When is the last time you saw one of the conventions they attend set up their booth without it?

What do you see as a major selling point when you scroll midway down the main site page?

https://il2-korea.com/en/

Yeah, it's a very important selling point.  In their own words, "brings the immersion to the highest levels", they know, they'll tune, it takes time.

[belphegor]

55 minutes ago, Aapje said:

It's hard to take you seriously when you make claims like this. The last time a x80 Ti was $650, was the 980 Ti over 10 years ago. That CPU is now way too weak for VR.

The GPU market changed a lot between then and now, and comparing the GPU pricing of that period to pricing of VR headsets today, seems like an attempt to cherry pick data points to support your position, not to make an accurate analysis. 

Yes, $900 is about the same as $619. 🤔

If you adjust for inflation, the GTX 980 Ti for $650 in 2015 is equivalent to $915 today.

[YoYo]

⚠️ Today's DD was dedicated to us! 

https://media.il2-korea.com/news/dd_81

[GiftGruen]

Interesting Devblog ! Thx !

[DecabrisT]

Первые впечатления от Pimax DreamAir в Корее:

 

[Antonovomatic]

Quote

We even created a prototype for this mode throughout almost the entire development cycle, but the effect amounted to no more than a 5% performance boost.

This continues to sound like a seriously flawed experiment being used to justify preconceived/desired conclusions about QV.  I struggle to understand how such a significant reduction in rendered pixel count does not equate to direct performance uplifts - especially when such uplifts can be clearly demonstrated by lowering render resolution in both flatscreen and VR... as this very blog post suggests as a solution for the use of the absolute highest resolution HMDs!

Quote

We also added an option to lower the resolution in VR mode below the headset’s settings. This proved necessary when using an 8K device. Even high-end graphics cards can’t handle this resolution: video memory usage and shader load are too high.

Then there is just a complete disconnect in why QV is desirable vs upscaling technologies - the former doesn't reduce central vision fidelity and clarity.  That alone makes foveated rendering (including fixed, without eye-tracking) stand head-and-shoulders above upscaling for VR simming IMHO.  
 

Quote

the performance boost from Quadview was the same as from DLSS, and using both together did not provide any noticeable additional performance.

Here is is clearly stated that there is performance parity between the two technologies, so with that being equal I am confused as to why fidelity and clarity would go unmentioned since that is where the methods most diverge, and I don't think I'm speaking out of turn to say that is a massive consideration for all VR simmers.

Edited July 14 by Antonovomatic

[YoYo]

Yes, I read the whole thing from DD, but I have somewhat mixed feelings. On the one hand, it's great, and thanks a lot for sharing the information. It's also clear (which is commendable) that it's stated that things aren't always perfect, and I thank you for that (no more painting it pink). On the other hand, I have some concerns about what will happen in the future, as some of this text suggests that things have hit a wall and the only solution is the next generation of top-tier graphics cards, which doesn't sound very optimistic. At least that's how I interpreted it. I'm not really keen on flying with DLSS because it has many flaws, and in my opinion, it won't work in combat simulators (it can cause more frustration than enjoyment). Also, Landscape Detail x2 looks very poor in terms of textures (they're largely blurred). Perhaps it should be divided into objects from x1 to x4, trees from x1 and x4, and ground textures from x1 to x4? So basically, it's similar to MSFS. This way, object rendering would be separate from ground textures. DCS also has a LOD factor that can be lowered. This causes pop-ups, but it's a workaround to maintain high gameplay smoothness. Maybe this is also a solution? I hope they're not giving up, and that this isn't some kind of excuse or confirmation that nothing positive will happen. The content states that they're continuing to work on improving the title and optimization, but oh well, time will tell, as I'm waiting for the next patches. It would be nice if there were better optimization for VR and a compromise between graphics and performance.

Oh, I don't know why IPD+/- was mentioned, since it doesn't work (OpenXR). I don't know if this bug was overlooked from devs ?

[Antonovomatic]

 

Quote

When evaluating VR system performance, it's important to remember that VR headsets operate with discrete FPS values.

 

This is also simply incorrect for at least the widest adopted brand of PCVR HMD, and likely others.  I can't speak for Pimax firmware, but Meta has had an option to completely disable their "Asynchronous Spacewarp" (1/2 refresh rate, 2x frame interpolation/generation).  This is a very common configuration for simmers as the interpolation introduces artifacts and latency, while running at native framerate of the game (automatically capped to HMD refresh rate) doesn't usually result in a significant reduction in final framerate over 1/2 refresh-interpolated but eliminates the aforementioned downsides. 

There is also a load band where the HMD is rapidly switching between native-frames at HMD refresh and 1/2-interpolated, which results in extremely obvious stuttering and changes in latency.  The sacrifice of maybe 10-15% "final" framerate is well worth those being natural frames in combat flight sims.  Fast/small objects (like tracers) also still look awful when interpolation is engaged, this artifacting/smearing/blurring/ghosting is only worsened when an upscaler is also a factor.

Edited July 14 by Antonovomatic

[YoYo]

7 minutes ago, Antonovomatic said:

This is also simply incorrect for at least the widest adopted brand of PCVR HMD, and likely others.  I can't speak for Pimax firmware, but Meta has had an option to completely disable their "Asynchronous Spacewarp" (1/2 refresh rate, 2x frame interpolation/generation).  This is a very common configuration for simmers as the interpolation introduces artifacts and latency, while running at native framerate of the game (automatically capped to HMD refresh rate) doesn't usually result in a significant reduction in final framerate over 1/2 refresh-interpolated but eliminates the aforementioned downsides.

I don't use Asynchronous Spacewarp / Motion Reprojection in any simulator. I have it disabled by default in the Oculus Debug Tool. It ruins the image for me and causes additional artifacts. It was good, but four years ago. As long as your headset can maintain the frame rate of refresh rate of monitors in VR, the image is super smooth and very sharp. It works flawlessly in IL-2 BoX. Actually, it works for me in Korea too, except in cities and larger villages, where I have performance issues.

Edited July 14 by YoYo

[Dash,Polder]

Yeah, in a Pimax smoothing is what they call it, and it's not even really bad, in props is where you notice it most, thing is they've re-written and redesigned smoothing and we should see it soon in an update.  With smoothing on I can really crank up the settings overall.  One area I question is shadows, off is such a stark contrast in everything compared when any level is turned on.  Easy to see in the hanger when you switch them off, it's as if someone turned the hanger lights on and cleared the air.  I see no real discernible difference between the shadow levels, they all look the same to me.  The world looks better with them off and that's a few FPS added to the bank for other things.  Shadows seemed to do more for me in GB's and I could see each levels difference.  My thoughts are they aren't done yet. 

[GiftGruen]

Yes, motion reprojection doesn‘t fit to a combat flight sim. It‘s annoying. Better stick to stable 70 instead of 90/45 and maybe slightly reduce resolutions.

Nevertheless there are some hints in the Blog which features use most power. I will try out. 

in summary the Blog sounds a little like ‚it is what it is‘. Hmm. We‘ll see. Taking into account that I habe the luxurious situation to own a 5090 - and nevertheless have okayish but not great results - I see quite some frustration potential for the community.

btw - from my experience in IL2 GB ( and also from VR benchmarks with HiRes ) the jump from 4090 to 5090 is rather bigger than the 10-20% the Devs stated in the Blog.

i also dont understand why reducing pixles by fvr only gives 5% in Korea while reducing pixels in general gives quite huge boosts. Makes no sense for me and really sounds more like a excuse. 

In the end, the more modern engine - while looking better - runs slower than the old one at the moment. So I need to reduce eye candy to have the same performance. Is it still looking better then? Hmmm….

At least I like that they communicate. That‘s good. And maybe it simply needs some patience.

Edited July 14 by GiftGruen

[Qcumber]

The DD said there was only a 5% improvement with QVFR. I would take that as a starting point. They have not said what headset they used for tests or what settings they used. In my experience you need to tweak the QVFR settings carefully. Other flight sims use DX12 and still get good performance boosts from QVFR. 

[Picchio]

We really need to start collecting data like it was done for GB: https://forum.il2sturmovik.com/topic/66924-syn_vander-benchmark-v6-to-measure-il-2-performance-in-monitor-vr/

[Antonovomatic]

12 hours ago, Qcumber said:

The DD said there was only a 5% improvement with QVFR. I would take that as a starting point. They have not said what headset they used for tests or what settings they used. In my experience you need to tweak the QVFR settings carefully. Other flight sims use DX12 and still get good performance boosts from QVFR. 

Yeah one gut speculation I'd throw out would be that these testing scenarios were also highly CPU-loading, which I suppose makes sense since "busy" situations are likely to be some of the most GPU-taxing... but not sure that reflects what makes up the majority of user experience, since a healthy amount of time is spent in travel/transition rather than combat.  And even when in combat it won't always be high unit counts/featuring the most taxing unit types or environments to render and simulate.

Edit: I want to add I too do really appreciate the blog post, I think there just might need to be some more internal advocacy for the experiences players are voicing on this subject, merely from the public responses I have seen.  I'm sure they wanted to get the writeup out in a timely matter, but it really read like there is a serious disconnect in why players are actually interested in foveated rendering, and what improvements the most reasonable of us are enticed by from it, including versus the existing upscaling tech. 

I really am still baffled by that bit about the VR res scaling options being such an important lever for improving performance... but being followed almost immediately by the statement that wha should be an equal, if not more, significant reduction in rendered pixel count via/QV incongruously had little to no impact.

TL;DR
Central/focus clarity needs to be a major factor under consideration, and for the moment/from what has been said by the studio, that is not the impression I came away with.

Edited July 15 by Antonovomatic

[GiftGruen]

Today I spended 2 hours to find an acceptable balance between quality and performance in VR.

Well. Frustrating. Tried the settings that we‘ve been told to use in the latest DevBlog, yes.

Either lower resolution down to Quest2 ‚blockYness‘ or switch off almost each and every eye candy thing. Orherwise it dips into low fps as soon ‚something is happening’

With somewhat higher settings it‘s fluid only flying high and alone.

That‘s not amusing with a CrystalLight driven by a 5090 and I am quite frustrated right now. 

IL2 BoX with MSAAx4 and higher Resolution in the end runs better and looks better - despite of some really cool visual effects in Korea and the beautyful MIG cockpit. But the performance pricetag for the beauty is too high right now, at least for me.

Did I say already that I am frustrated?

Well. I think I‘ll drop it for a while and try again after maybe some techmagic is done.