Pikey

Thank you for your suggestions I tried them. I feel like changing my Facebook status with OpenXR to "it's complicated" The settings at 80% you provided did get MP into the 40's and workable, so that's a thing. Frames are enough, even tested on Blue Flag Syria with an Apache and its high 30's (but not into 20's) which with OpenXR is better behaved than SteamVR by a large amount. So this is a bottom line, with the tracking enhancement I'm keeping it with the following issues. - Water to medium is a hard sell - it makes sense to look at Baileys startup shortcut maker to have that launch on demand. It's a big loss for me for when I look at it. - Cockpit Global illumination I really miss and it also causes some oddities in the game - I've had bright lights - that's not VR's fault, but its a less tested config and I'm going to be busy reporting some horrid lighting bugs for a while, including "Bright white ocean", and "Cockpit under a spotlight" - There's a small distant shimmer, especially with complex clouds at Medium visibility. I need more pixels on screen to resolve that, it's where this is. I'm using NIS and unlocked frames. - pixel detail is bang on minimum acceptable for cockpit reading, but I get high textures, so its give and take. - Chimney and clutter - somewhat irrelevant once completely gone, so a good saving. - AA is x16, I always thought that was fairly cheap, but ED's stock VR has it reduced to 4. Any lower and the runway end turns to mush FPS comparisons for OpenVR and SteamVR are like trying to convert GBP to EUR. A little less OpenXR is equivalent to higher SteaVR FPS, but the tracking and comfort are light years apart. My brain has become so tolerant of all the jerky shaking and stuck moments in WMR ,that putting OpenXR on was like a bandage to the senses. It reminds me of Occulus Rift tracking actually. The Rift had this ability to handle low frames way better than the HPO Reverb with its stations. WithRrift you could have 10FPS but your head still tracked. I don't understand the technical detail, but the tracking for Reverb is dogsh1p - you build up this tolerance to discomfort and just put up with it because you have to. Developing an immunity is no substitute for a gentle experience. I might never reach the pixel density on this card to give me a good horizon for spotting, but it should make me less exhausted. I hope this project continues, if its me, I could trade some of that tracking for some CPU and the end result of more pixels, as it is, I think after a very long battle with settings I will keep OpenXR. Thanks.

sorry to interupt, but I get dead events for the instant kills and slow burners. Is there a more concrete type testing to perform to accept this as officially fixed?

Yeah its close... I think my conclusion is that if you have the system to support it, its superior, with the lack of steamVR items like fpsVR included in that statement, but for MP and lower specs it's ending up losing out on the grunt to carry it up to minimum FPS play. It definitely tackles texture related nonsense, i've even found it more stable with an old bug/thread where swapping cockpits would eventually break SteamVR and cause a black screen. It's really killing me, but the actual result when trying to play on Blueflag is that the overcompensation on smoothness ends up looking like your HMD turned into an eel swimming through jelly. Super sad about this because it tackles nearly all the basic issues with SteamVR. The irony is that if you take down the quality settings it doesn't really affect OpenXR like it does SteamVR. OpenVR remains very stable on frames, both up and down. For example I cant get it over 55/60 FPS and Steam would hit 90 in game, and going lower, SteamVR recovers quickly from troughs of poor performances, where OpenXR seems stuck in the mud. It's like SteamVR took a chill pill. Long wait until 3xxx series GPU are acceptable price for me. But people should really ignore the noise in the thread and just follow the instructions - it's one quick default install and swap three files into your DCS directory, how that can be screwed up I just don't know, also the OP didn't do any favours with the conversation, people don't read that, apparently over 4 lines and you are passive-aggressive these days...

I'm so torn by OpenXR. It gives and takes and I think its just having not enough hardware that's letting me down. i5 10600K and 2080S. I reinstalled things and spent more time using it than benchmarking it and I warmed to it. The way I have found it is that the tracking is the most amazing improvement. Soft in turns, doesn't stop if turning your head fast or at extreme angles like looking down and to the side, the actual tracking issues solved from SteamVR, never have to recentre at the start. But the loss is pure frames and CPU frametime. I cannot explain it. This means that going into MP I get the frames descend into the danger zone and things start to go wrong from then on. Offline, I think its better than SteamVR and in MP it wont get up high enough to guarantee 30FPS all the time. A lot of 3xxx series owners claimed this is just scoring points needlessly as if frames don't matter, but when you are clear of the low frames the benefit of the tracking without loss becomes way more obvious. Pros: Tracking is awesome, smoother turning, get rid of Steam, can change settings with Dev tools on the fly, better offline compraritive performance, easier to play with, starts faster, popular community effort. I think (and its impossible to benchmark) it behaves much better for low texture memory usage. I found the F14 which was notoriously heavy and juddery was smoother and mor eenjoyable to fly. Cons: FPS is lower, it struggles worse at low FPS than SteamVR at the same FPS, can't use fpsVR, its a mod and resets with updates, appears to be more CPU bound and performed badly in MP, some of the settings in the Dev tools dont seem to work well or at all Whatr I do know is that you cannot compare them well, I thought I could, but I cannot.

Can you see anything in the mirror or is this all derived from in headset comparisons? I don't see the difference. What is actually changing? To be precise I see the same thing but with the OpenXR I get up to 30 frames less than with SteamVR and I can literally see the stutter in the mirror. I don't understand how I seem to be th eonly person who gets worse performance?

it's literally an OVGME double click on/off once the OpenXR dev tools are installed. It backs out and in perfectly. I also tested DCS repair, it removes to the original also and takes less than 45 seconds to do one on NVMe. Simplest thing to add, but obvbiously you have to readd it each DCS update. Classic gotchas are running SteamVR, or a repair afterinstalling. Will cause a Black screen. That's it. I still don't understand the benefit yet, in the HMD i'm not seeing the difference. I wonder if the placebo is caused by the DCS repair?

PD and SteamVR Super Sampling are not the same code but they are the same thing, more pixels/resolution. Community has swung on when is more performant, the only consensus I recall is lower PD and use higher SSS. I use 1.0 and I can go to 130% on Steam, depending on mission. SInce DCS is in open development this can change. The only way you can get a juddery mess in the menu system is by wrecking something outside of DCS. In the menu system most people will get 90 fps until maps and such get loaded in. If you meant something completely different with the word "menus" then please be more detailed, since you can get issues joining servers in MP if they are struggling under certain circumstances. Target frames sub 30ms. outside of MP you should have a lower CPU frametime, inside a MP mission the CPU will begin to overtake and thats when performance bottlenecks elsewhere. You target what you play in consistently. The main change is what you play, not so much your settings unless you do something crazy to the supersampling. Aim to keep CPU and GPU frametimes together. Often to rescue VR performance restart the PC, fo some reason, it can sometimes go into a mess for no reason. Can be affected by latency to the server even, I've seen people mass join servers, server starts to drop frames, players start to drop frames and its nowt to do with the graphics. One of the worst offenders people moan about is on patch day when their shadres recompile and there's a juddery mess. Can watch that in the dcs.log The simple solution to that is allow your PC to run some heavy unit missions offline in 2D until it stops ebfore trying to run that in VR.

Sure, except I benchmarked a track file offline. With the Steam VR I could measure framteimes in the baseline and the Cpu was all green from fpsVR. Am I supposed to see an improvement in frames, I felt from the OP I should? I can say I saw less peaks when I drew out the graph. Has anyone in this thread done before and after benchmarks?

Hi thanks for your response. The loss of 20 FPS when you can dip lower than 45 is damaging to me since the screen is a juddery mess at the 22.5 fps mark which can be reached quite often in big MP servers. For everything not data related It will remain my opinion which is unaffected by your own. VR with DCS is about headroom. MP requires it. I spend most of my time sub 45 frames. The benchmark I used for the tests is offline so the headroom illustration is important. Offline players have less to concern themselves with as they tend to stick to optimised playing configurations. We aren't talking about negligble performance gains, I am citing a recorded straight up loss of 33% against SteamVR. WHere I had 90, it was 60. That theme continued when frames dipped to 30 and I got 20. If this doesn't bother anyone then my post shouldnt be of interest and can be safely ignored.

I spent a great deal of time benchmarking this and I can find no benefit and only loss of FPS. For sure it is working, and I dont have stability issues or black screens. SteamVR remains off and the game actually loads a lot faster. (in one test, 45s faster) I've used two builds of DCS to test and finally I get consistent results with them, using PD 1.0 and the VR preset. Only FPS is recorded, it's taken directly from the game programmatically in the same way for each test. As well as testing on two builds, I have reverted changes backwards and forwards on the same build and obtained the exact same figures before, and after removing the changes and then after again. It's entirely consistent. The scene I use to test is a track file, so there is almost un affected GPU frametime by CPU bottleneck. The scene is 5 minutes and covers some cockpit, external, city and woodland shots across the Caucasus. Frames are reecorded via script from the game. The Headset remains motionless on the desk throughout. I was quite disappointed since I wanted this to be a magic fix. I am very happy someone is working on this technique. I don't know the developers and have no reason to contradict them. I have not done this work in any official cpactiy as a tester, just as a curious individual. I remain open to tweaking and looking at this in the future if ther eare substantial changes, however at this point I find the changes quite damaging to DCS performance and will not be using them. It's possible that something isn't correctly configured, however the files replaced clearly do something to get it working without SteamVR, good luck with you own findings!

8GB RAM will make the game chunter a bit these days, not sure about the graphics card but the minimum specs are below As said, its a free download and there are free planes, so the word free, should make it clear you have nothing to lose except time And by that I mean most of us, as in, the last decade I have lost to DCS Start here: https://www.digitalcombatsimulator.com/en/support/faq/ Minimum system requirements (LOW graphics settings): OS 64-bit Windows 10; DirectX11; CPU: Intel Core i3 at 2.8 GHz or AMD FX; RAM: 8 GB (16 GB for heavy missions); Free hard disk space: 120 GB; Discrete video card NVIDIA GeForce GTX 760 / AMD R9 280X or better; requires internet activation. Offline play has very basic pilot and roster with awards. Some people built addons for that. You can play with keyboard and mouse badly, and better the more you spend on peripherals For anything with a propellor it will be very hard without a twist stick at least, but for jets like the Viper you are good for Z and X ont he taxiway. Documentaiton is included for the product and for each module and Training missions are supplied and begin quite basic during Early access and can become quite in depth once the module matures.

There's way too much info already out there, makign a guide will bury it in this weeks videos but feel free. 1. Launch your VR via steam 2. Launch DCS. 3. Go to settings (cog) 4. On SYSTEM tab, click VR preset 5. Go to VR tab 6. settings, Click enable Virtual Reality Headset. (This setting will be persistent, you can use shortcuts or a launcher to switch directly in the future) System restarts in VR (Put facemask on). Play and adjust for your own FPS comfort level. What I know is that people are not the same and they will choose to give up more for FPS or want more for this setting or that, but that's entirely up to you, there is some guides out there for each one, but there is no fixed method or certainty that X feature removes Y FPS, it depends mostly on the mission, then the resolution/supersampling (Steam Super Sampling in Steam VR) and Pixel Density (PD) in game. Remember you do not use MSAA or SSAO. Start with the VR setting and a quickstart mission. For 2 bucks get "fpsVR" which will give you access to performance information which can be used to tune further. You can't escape tuning because there is a simple fact. You are not allowed to have all of the frames, you have to give up some in exchange for the pretties. Pretties vs frames is the endless choice. With fpsVR tune based on multiplayer since the CPU usage is higher compared to the offline which you will bottle neck the GPU first. Use a baseline that makes sense. Typically there is an element of degradation when the frametime reaches red. Frametime can go from 12 to 30 to get from green to red. Adding settings usually adds one or two milliseconds frametime per feature. Anything else is past beginner level.

I saw this a couple of times. The most interesting was one of my own servers. I updated it for Apache and changed a lot of things, but it's sub 1 FPS on joining. I noted that a local player didn't have the issue, in VR. WHen I had it running locally, it wasn't the same behaviour either. This means something is network related as that is the only change, but it doesn't really give much data. There is nothign in the logs, when I added a FPS count it only showed the server had OK FPS and I didn't. I don't use warehousing.

Thanks ED, this was a new benchmark in EA quality.

Hold my beer...

it is slower. Yet its the only helicopter module I've managed to fly with KB and mouse due to that nice stable movement. What you can't do is move the stick fast to recover something you didn't react to in the first place. It could probably do with a bit more sensitivity, esp the collective.

I reproduced this only by double binding the cursor to two different input devices - my left throttle on VIRPIL mongoose and an Xbox controller. When I did that the cursor would seize after a few presses. Not sure if that is a bug or even the same as reported here.

When exiting a mission you have the opportunity to save what happened as a track file. This is a copy of the actions you took during the mission and allows someone to review what you did. If you still have the issue then attaching the track file removes all doubt. If you can reproduce that and add it you might either be able to have someone advise or be creating a bug report, I coudlnt really tell or help from the description though.

Well it's inconsistent, if you started without caring for the hangar, why care for it later? Imagine going back in, just the same way you imagined getting in the aircraft. The doors work like was said, same as the runway lighting and the rest of the sim, not everything has a reason or a switch. There's tricks to get around it, but better to either use the hangar you started from or park it in an open apron.

I need a 1TB for DCS as I have a couple of versions. I haven't installed 2 terrains and the size of each install is exactly 250GB, so for one install 500GB is minimum. If you want to do anything else, get a 1TB M.2

The term "labels" redirects here.

42VFG is flying twice weekly; Warbird Wednesday which is low key WW2 shootem-up, and Sundays which cater for any of the 80's+ modules and whatever people want. We would still like to see a couple more people and get to know you slowly. We don't do mass invitations as we are primarily a social group. Reminder: we still plan flights, don't use F10 map and use settings tuned towards realism and we all play in VR with HOTAS. UK server, events started from Discord start at 1900 GMT, access to resources and hookups for casual flights any other time. If you are suited to us drop me a PM.

VR does require extra effort. The thing is, that its a fairly small minority that can't enjoy it, usually due to hardware/expense. Also when giving up the frames and settings of 2D a lot of people can't be happy with that and end up in an endless cycle of trying to improve which leads to disatisfaction. I have used VR for years and had three headsets. I even gave away my Trackir. There's plenty of things that are suboptimal with VR in general but none of it is relating to enjoyment and immersion. A wireless mouse on a little side stand and good peripherals solve 90% of issues, the rest of stuff is solved by mods or just worked through.

Not only in wartime was basic crashing during landing and takeoff a typically common event, but experienced pilots wreck millions of pounds of machinery every year today, which is heartbreaking on many levels. One of the reasons I like the warbird flying experience is that combat with other planes is merely a side dish, you have to do combat with the aircraft and then the ground before that!

Usually the tower via the menu and let them know.