RealDCSpilot

Then it's time to go out on the streets and cry "Make WMR Great Again" ...

@Buzzer1977 Goodness... The topic is WMR getting depricated and i'm trying to tell you why the concept failed. And i'm looking at the whole history with Microsoft trying to bind VR HMD manufacturers to their system. I'm into VR since 2012 and owned VR headsets since 2014. WMR - the naming alone is already misleading, the runtime has the name WMR and the native API is also named WMR!!! But no one really developed or integrated the WMR API into their games. Before VR it was only used for Hololens projects and after 2016 only a handful of small developers tried it and put their VR games on the only store with WMR API VR support - the Microsoft app store. OpenXR was added a bit later to be supported by the WMR runtime, now look how long it took for OpenXR to mature and at the dates how many OpenXR games are out there and when were they released? The whole WMR thing was a dead horse from the beginning and the WMR for SteamVR wrapper was the afterthought solution trying to keep the bad idea alive. On many occasions the WMR for SteamVR plugin simply broke, they were always running behind SteamVR's progress. @Baldrick33 Acer, Dell, HP, Lenovo and Asus HMD basically failed instantly. I would call that a desaster. Samsung and HP tried once more with a more sophisticated HMDs, but Samsung also gave up in the end. The tragedy of the HP Reverb is that HP still kept staying with WMR. They were on the right track with the most parts of their hardware, but should have made themselves independent from Microsoft in all those years. I also still remember how many people just bought the Reverb in 2019 without having the hardware to drive those highres panels. Even the best consumer GPUs in 2019 weren't really able to deliver the resolutions the Reverb demanded. Because nobody listened to HP, they never intended to sell the headset to normal consumers. The Reverb of 2019 with those early highres panels was intended for engineers and professional designers in CAD workflow integrations, on highend GPUs to be able to see fine lines in 3D data and text. That's one reason why the Reverb was not planned to be produced in large numbers in the first place. But then the simmers discovered it for their niche market. It's good for seated VR, mostly without the need of those crappy controllers. But the performance with demanding sims was abysmal, so programmers from the community stepped in with creating free tools to get access to tuning options. But this only extended the life support for the dead horse. It's a good thing that some bad ideas die out naturally. And when WMR has vanished, alternatives will be available.

You are still stuck only looking in your own small niche. I still remember the flood of new VR users seeking help in 2017 when WMR headsets were released and i remember the second flood in 2019 when the Reverb was released. You also still don't get what i meant with WMR being the third leg. Simply ask ED why they never added WMR API support after implementing LibOVR and quickly after that OpenVR back in the days.

The only thing everyone was interested in were the Reverb's first highres panels in an affordable VR headset. Everything around it was cheap as hell (the controllers... ouch, the tracking...), but those panels were perfect for sims with cockpits with instruments etc. If guys like mbucchia didn't came to the rescue with OpenXR toolkit or OpenComposite as a direct wrapper from OpenVR to OpenXR, no WMR user would had a good experience with VR games. That's the only reason why WMR itself survived this long. It's death was quite foreseeable 6 years ago.

No, that's the other problem of WMR. It's a prison created by Microsoft. And HP's problem was to enter it. They hopped into bed with Microsoft and all the dependencies that come with it. It might have been a good idea on paper first, lowering the cost for developing efforts by simply following Microsoft's specifications and requirements. The only solution for current WMR users would be a kind soul of programmer who codes an open "jailbreak" to maintain and keep WMR dependent devices alive. An own driver, runtime and connection to OpenXR would be better. However, since HP stopped their VR endeavour at the moment, there will be no progress in hardware capabilities anyway and WMR devices are doomed to die out eventually. No one will care about it in a couple of years anyway, there will be other and better HMDs to choose from. Just to clarify: WMR was also an API like OpenVR, LibOVR (Oculus) and OpenXR (later). WMR as a runtime (WMR portal) wanted to see VR apps made for WMR the API. But at this time (2017) LibOVR and OpenVR were the most common APIs already, so no serious developer wanted to go the extra mile to integrate or build their app for the third much less accepted VR API. As a cumbersome afterthought, WMR for SteamVR was created, a wrapper to translate WMR to OpenVR to give WMR HMD owners access to Steam's VR library. (There was no official OpenXR game available until MS 2020 flight sim was released). What a mess. With the introduction of the HP Reverb, HP actually tried to avoid the VR gaming market first and provide their headset to professional users only who develop their own WMR apps for business use cases.

You mix up the device with an API and a runtime that no one needed. WMR was an API and a runtime. If the Reverb never had OpenXR support you wouldn't even know that it exists. WMR was one of those fails of Microsoft, completely misunderstanding a new and slowly growing market. Check how much of the other WMR headsets failed so badly in 2017.

Guy Godin from Virtual Desktop announced that he will code his client streamer app for the Vision Pro as soon as he gets one in his hands (and Apple allows to publish it). From there it might be the best possible but also very expensive PCVR headset without controllers for a relatively short period of time. Good for any kind of sim like flying and racing.

It's a good thing. WMR was always the third leg no one needed. If you look back in time, on how much trouble it caused within the VR community... It will need years alone to diminish the number of users which confuse the OpenXR API with OpenXR toolkit and OpenComposite etc.

I can confirm that just having Ultra Leap tracking software installed prevents DCS from working in OpenXR. I actually don't want to use hand tracking in DCS, but i need Ultra Leap installed.

It's the current maximum bitrate you can choose for HEVC @10-bit codec. Keep in mind that this is not 1:1 comparable to the bitrate of the h.264 codec. HEVC is a much newer and more efficient codec, plus 10-bit means color depth of billions of colors (vs millions in h.264). Just a note: Virtual Desktop also makes use of the 2 encoder chips of your 4090, so that one encoder works on the stream for one eye (sliced encoding). The good thing about the 150 bitrate is also that you can get away with a Wifi 5 router directly connected to your PC in your room, it's more than enough to run best quality for wireless VR in your situation.

Sorry, can't tell anything about Facebook headsets. I think i read somewhere about a recent Oculus update that ruined some of the functionalities. But with a Pico 4 you can basically connect the headset in three ways to Virtual Desktop on PC, wireless, USB 3 cable or Ethernet adapter. Some people use an extra Wifi-router dedicated for VR in the same room. @Derrick006 With your hardware you should get performance similar to this (DCS graphics settings maxed, screen recording is 30fps only):

I don't think it's DCS, have you tried Virtual Desktop with VDXR? Get rid of all the Oculus and OpenXR tinkering before using it so it can't interfere. You should get very good results with DCS_mt.exe, 90 Hz, DCS pixel density setting 1.2, VD at godlike - HEVC 10-bit at 150 Mbps. Latest Nvidia driver with CUDA - Sysmem Fallback Policy -> Prefer No Sysmem Fallback.

There is a completely new ground crew menu. The only problem with it on my side is that it's only clickable with the mouse. I'm in VR with mouse buttons bound to HOTAS and centered mouse cursor in the VR headset, i can't interact with the menu with this configuration.

For me the issue seems to be completely gone now. Sadly i can't put my finger directly on it, to much parallel changes were going on in the latest weeks (switched from SteamVR to VDXR with Virtual Desktop). But my suspect #1 is the new Nvidia driver feature. I can now play for hours in huge multiplayer servers, switch modules and so on, no freezes - no stuttering.

Just a hint: Process Lasso was very helpful for years tuning DCS performance on a 10th gen Intel CPU and Win10. When i switched to a 13900K and a 4090 on Win11 it became the opposite, Process Lasso was the culprit for horrible stuttering. So i completely removed it and i'm enjoying smooth performance since i did that. Consider taking it out of the equation when testing the mt.lua. The only thing i have running is unparked cores. The mt.lua made no difference, everything runs fine. All in VR, DCS mt exe, OpenXR, VDXR.

Sorry but no. This has nothing to do with Pico's announcement. VDXR is direct OpenXR API access for the Virtual Desktop Streamer app on your Windows PC. It's an idea, concept and execution by mbucchia and Guy Godin. It removes the detour for OpenXR going through SteamVR. It would have been possible at any time, completely independent of Pico. And it's a good thing for all mobile VR headsets that work with Virtual Desktop.

I do everything HOT3 with the throttle. The little analog stick for moving the Viviane/Periscope, a hat for zoom and the other features and one button to fire. The rest is just depending on the situation, most of the time i keep moving anyway because the outdated tactics with being a stationary target hanging somewhere in the air doesn't end well in most cases. If you have to go so close with 4300 meters max range it's better to do "hit and run" like a Mi-24. Check at ~25 minutes:

Alright, having the mirror running fullscreen at 4K is definitely a bad idea unless you need it for special purposes like screen recording. Every window on desktop needs VRAM, so the bigger the window is the more resources it will eat. I don't know how Facebook manages it's VR runtime with the mirror, if it works the way you described it's not the best way to do it. However, if you have control over it - good for you.

Just for info, i accidently found CPU frame time of my 13900K. It's between 2 ms to 3 ms. So yeah, not much benefit from a "3D" in the name. Choose whatever best possible CPU and GPU you can get for your budget. Intel or AMD - both deliver very good results with the best Nvidia GPU.

Usually the desktop mirror doesn't cost any extra performance. It's just a passthrough of what is rendered in one of the eyes and mapped onto the 2D face of a window at the specified resolution. And i guess on Windows, a 3D app can't run "windowless".

@mbucchia Runs really well. Thanks!

VDXR integrated in Virtual Desktop is officially released now! You can now run DCS with the Pico 4 without SteamVR but in native OpenXR. This should also come with a little performance increase for everyone. Simply choose VDXR in the options of VD streamer on desktop and then launch the DCS.exe (multi-threading version).

- Simply tick "use DCS system resolution" in the VR settings tab - close DCS - go edit the options.lua in your savegames folder - look for ["graphics"] = { make sure that ["fullScreen"] = false, and edit the values behind ["height"] = and ["width"] = to your liking This way you can let the VR mirror take any rectangle shape you want. Need a square? No problem. (A square would be the closest shape to what is rendered for each eye in your headset) The VR mirror will always be placed in the dead center of your desktop. After this, the resolution selection tab in system settings will show your edited resolution, never touch it again. I recommend a height value that is always a couple hundred pixels below your monitor's desktop resolution, something like 1200 on a 1440p monitor. This prevents "slipping" of the VR mirror under the taskbar and tapping "out of window" mode if the mouse cursor hits the taskbar.

You missed the memo. Do not use the "Autohover" which is WIP atm.

Please close this thread, i was waiting for 7 years to finally enjoy the Gazelle with a flight model that matches the standard of DCS. I think jef32 will be able to show a little patience to wait a bit for a feature that is easily to work around with proper trained trimming of the cyclic. @jef32 The Gazelle module was grounded in my shelter for 7 years, can you beat that?