Microsoft pulled the plug on support for its entire WMR platform on Windows 11 last year, putting an official end to the company’s foray into PC VR headsets. Now, an unofficial SteamVR driver hopes to bring it back.

Microsoft deprecated the Mixed Reality Portal app, Windows Mixed Reality for SteamVR, and Steam VR Beta when Windows 11 24H2 update rolled out last October, making a fleet of PC VR headsets from Acer, Asus, Dell, Lenovo, HP, and Samsung essentially expensive paperweights.

Granted, if you haven’t updated to Windows 11 24H2, or are still on Windows 10, Microsoft says you’ll be able to play SteamVR content through November 2026 before the plug is pulled for good. Still, that’s a bitter pill to swallow for users of WMR’s most modern headset, HP Reverb G2, which released less than five years ago.

Now, Reddit user ‘mbucchia’ claims an unofficial SteamVR driver is in the works, which aims to bring all WMR headsets back into the fold sometime this Fall. Below you can see the first HP Reverb (2019) in action:

Dubbed the ‘Oasis’ driver, mbucchia says the SteamVR driver “does not need the Mixed Reality Portal,” which was deprecated in Windows 11 24H2 last year alongside SteamVR beta support.

“This means it can work on Windows 11 24H2 and newer. It supports full 6DoF tracking along with motion controllers,” mbucchia says.

“As mentioned on another post, I don’t have all WMR headsets to test with. Though I can tell you that it [also] works on the original Acer AH100. It should in theory work on any brand/model,” mbucchia says.

The first wave of WMR headsets launched in 2017 | Image courtesy Microsoft

Work on the driver is remaining behind closed doors, mbucchia says, noting that Oasis will be restricted to Nvidia GPUs due to the way SteamVR interfaces with the GPU drivers.

Most interesting of all, though, is Oasis isn’t being undertaken by just anyone. Mbucchia claims they are currently an employee at Microsoft who has previously worked in the company’s Mixed Reality division.

“I am bound by NDAs and other obligations. I want to be clear that I have taken much care to NOT BREACH any of these agreements while working on this project. In particular, I am leveraging SteamVR for a lot of heavy lifting and I am not borrowing any Microsoft intellectual property,” mbucchia says.

SEE ALSO
Meta Teases Oakley Partnership for Sportier Smart Glasses, Reportedly Releasing This Year

For these reasons, Oasis won’t feature a beta, or similar early access, before its release in Fall 2025. It also won’t be open source.

“Much of the code is the result of deep reverse-engineer. Reverse-engineering that if shared, could be construed as exposing internals of programs like SteamVR or the Nvidia GPU drivers,” mbucchia explains. “Not[e] that here again, I am NOT BREACHING any proprietary/intellectual property. Having respect for both Valve and Nvidia, I will not divulge any of the code that they do not consider public.”

Mbucchia says they’ll reveal more about the project in the Windows Mixed Reality subreddit leading up to its Fall 2025 release.

Newsletter graphic

This article may contain affiliate links. If you click an affiliate link and buy a product we may receive a small commission which helps support the publication. More information.

Scott Hayden
Well before the first modern XR products hit the market, Scott recognized the potential of the technology and set out to understand and document its growth. He has been professionally reporting on the space for nearly a decade as Editor at Road to VR, authoring more than 4,000 articles on the topic. Scott brings that seasoned insight to his reporting from major industry events across the globe.
  • Christian Schildwaechter

    About a year ago mbucchia/Matthieu Bucchianeri posted a rather detailed explanation why a replacement driver like Oasis wouldn't be possible due to how WMR was integrated, relying on parts inside Windows that couldn't be easily replaced, or at all. One reason is that Windows automatically takes over all connected HDMI devices that identify themselves with a unique EDID number. Working around that would require special hardware so expensive that buying a new HMD would be cheaper.

    I first suspected that this new driver was a port of the Monado project, which implemented WMR support for Linux, but it is apparently completely different, and a lot more complicated. From the post explaining why it wouldn't be possible:

    The OS Components are a black box, so it's not like someone can just go and replace it with their own. Re-implementing the Public APIs would be unnecessarily complex and tedious vs just re-writing something or using Monado (assuming the EDID limitation can be bypassed one day). And there is no public documentation of the driver APIs AFAICT. By the time all of this is tied back together, your cables will have died anyway.

    And now, only a year later, he apparently has done the impossible, for which he deserves a ton of praise. And he was probably the only one who could have done it. mbucchia was part of Microsoft's WMR division, had access to the required information and the capabilities to reverse engineer and reimplement parts without breaking any rules of his still employer. He is also the one behind the open source OpenXR-Toolkit, so the Oasis WMR driver not being open sourced mostly has to do with confidentiality agreements he has signed.

    As to why he went and invested what were probably ungodly amounts of time to revive a family of headsets that Microsoft has killed multiple times in many ways, another reddit quote:

    I dont care lol, I have some unfinished business with this headset XD

    This is one Microsoft engineer that apparently was even more pissed by Microsoft dropping WMR support than the users that bought the headsets. There is always the question what happens when a large company like Microsoft or Meta loses its strategic interest in VR/XR and stops pouring in billions, but as long as there are people who absolutely refuse to let VR die, I am still very hopeful for the future.

    • Andrew Jakobs

      There are always people who love their old hardware even if the rest if the community has moved on, and these people see it as a hobby/challenge to keep their hardware running. And luckily for WMR headset owners he is making it possible for them… But if you're still on the first batch of WMR headsets, you're better off just buying a second hand quest 2 or Pico 3/4. But if you have the HP G2 it still is technically an excellent headset (if you don't mind the cable).

      • Christian Schildwaechter

        TL;DR: keeping old hardware alive is one aspect, but just as important is becoming less dependent on the whim of companies.

        While people do amazing things with old, unsupported hardware, it is not only about keeping things running even though there are newer/better alternatives. It's a lot about not depending on the policy of one company that introduces a single point of failure.

        Projects like Monado not only enable using WMR and other HMDs on Linux, they provide a complete template that others can base their work on. Monado created the only open source implementation of the full OpenXR stack, which could allow for example much smaller manufacturers to come out with experimental headsets.

        And Valve just released the first version of the Linux based SteamOS 3 also targeting 3rd party alternatives to the Steam Deck, with official support for Lenovo, but others working too, and better than with Windows. This happening shortly before Microsoft stops support for Windows 10 is probably not a coincidence, and a lot of people may be tempted to switch to Linux/Steam OS instead of Windows 11 that often requires buying new hardware

        Proton fixed most issues with gaming on Linux, but so far only Valve supported VR on Linux, everything else requires Windows, incl. streaming solutions like VRDesktop. And again open source came to the rescue, with the ALVR client/server first allowing to stream from Linux to pretty much anything, and then within days of the release of AVP making it possible to stream to the first available 3.5K HMD. It's not just about preserving your investment, it's about empowering the users.

        • Thud

          Christian, I always appreciate your insights

        • Andrew Jakobs

          Yep, but a lot of times it just starts as a hobby to try to support these devices/API's.

      • kraeuterbutter

        i love my Quest3..
        love the lenses compared to all the WMR-Headsets i had before..

        BUT…
        for example Kajak-Mirage: the HP Reverb G2 has a much better image, looks much more detailed and "photorealistic"
        for most games i prefere the Quest3 by far..

        but on that game, the Reverb G2 looks much better
        (hate other things on that headset… unstable with WMR, sweetspot too small, …)

        hope to enter the high-resolution-club soon..
        but can´t deside…
        bigscreen 2, Pimax Super, Pimax Air ?, PImax Air SE, meganX or play for dream?

        every headset would be a step up in graphic fidelity by a huge step..
        BUT: every headsets has its downsides too…
        do i still want a cable ?
        do i still want to use my old base-stations, the Index-controllers i meanwhile do not like that much anymore ?
        play-for-dream sounds good.. but: standalone without possibility to use all my quest-store-games?
        bigscreen 2 -> cool, but: can only use in my working room were the gaming-pc is located.. do i want to be limited to that, now i know the benefits of standalone…

        hmmm..
        hope a improved quest4 comes next year !!

        • Andrew Jakobs

          Don't count on a Quest 4 next year according to some news articles last week.

        • Christian Schildwaechter

          What's your time frame and budget? The only way to keep using your Quest games is buying a Horizon OS HMD, with the Quest 4 apparently postponed to after 2026, and the promised Asus/Lenovo headsets still nowhere to be seen. Just for future safety I'd avoid anything that doesn't come with eye tracking, which kicks out the MeganeX and means Beyond 2e instead of Beyond 2, which in your case adds EUR 180.

          My current strategy is "wait and see what Valve is cooking", which is of course a very risky strategy, despite recent rumors of Deckard coming in (late) 2025/2026. I'm pretty sure it will offer an excellent cost/benefit ratio due to Valve paying for the development with Steam money, will be less restricted than a Quest 4 that still has to target a more budget sensitive audience, and thanks to the Steam Deck and a constant flow of improvements for it over the last three years, I trust Valve a lot with providing excellent support for Deckard. If it ever comes out.

          • kraeuterbutter

            Timeframe: after this summer
            budget: depends: a "cheap" Crystal Light + a Xreal One Pro ?
            or only a VR-Headset for 2000 ?
            depends…
            not more than 2000
            my considerations:

          • Christian Schildwaechter

            Your timeframe is somewhat unfortunate, as we are on the verge of getting a new class of hires HMDs, most of which aren't available yet, and all of them priced rather high, taking advantage of enthusiasts longing for a new generation. Most of them will be based on similar hardware with high margins, at least until several alternatives are readily available and start to compete by price.

            But you won't get either a Quest 3 successor nor Deckard anytime soon, and no other standalones will offer a VR gaming library worth mentioning, with Samsung probably the only one with a chance to get at least some games/ports thanks to Android XR. So for gaming you'll need your PC, which reduces the advantage of more mobile inside-out tracking.

            Pimax Super is expensive, doesn't require lighthouses, and will probably be shipping after summer. Crystal light is much cheaper with about 56% of the Super's pixels, but doesn't have eye tracking, which according to @XRC provides about 40% performance improvement, so you will need a similar GPU for both. The Air (SE) may or may not release in 2025 and offer microOLED compared to QLED on Crystal, with similar resolutions to Super/Light for roughly 1/3rd more money, but much smaller FoV.

            Everything else also has a smaller FoV than the Cystal, and none of the VR/passthrough HMDs are particularly useful for productivity use yet. Crystal Super and Beyond 2 have very good lenses and the displays get excellent reviews, something we don't know about any of the not yet available ones. The MeganeX has some minor warping issues, still excellent, but for driving 3.8K without ETFR you'll need a much beefier GPU. It also requires lighthouses, and is priced similar to what is expected for the Samsung Moohan standalone that would run flat Android apps, making it the best option for non-VR gaming use.

            Whatever you pick now will be a compromise, as each of the near term solutions comes with significant disadvantages, and none of them are future-proof. In many ways the Pimax Crystal Light plus Xreal One Pro would be the most pragmatic solution with an established set of functions that you know will work. It simply too early to pick one HMD that is useful for both gaming and productivity use, either tethered or standalone, as the only hires standalone already available is AVP, which is not a good choice for gaming.

            For someone only looking for an upgrade for VR gaming the choice would be much easier, boiling down mostly to whether a tether is acceptable, with some already well tested solutions available. If you separate VR and non-VR use, the Beyond 2 lacking passthrough may again be an option. For wireless streaming or standalone and non-VR use, waiting longer until at least two or three options are available and tested would be much better. If you really want something soon, it might be best to go with the cheapest hires tethered solution for VR gaming, and expect to buy another one in two to three years, when more wireless/standalone options have become available at lower prices.

            As always, the answer is "it depends" with no trivial answer, and varying and constantly changing dependencies. I will mostly ignore whatever releases in the next 1-2 years, instead just continue to live with the resolution of Quest 3, and made my number 1 requirement "must run SteamOS standalone". And I won't reconsider my requirements list before 2027, which makes the selection process significantly easier and saves me a lot of time that I can then waste on posting about VR instead of using it.

            https://uploads.disquscdn.com/images/a47d2231680b3e38fdace9a6a471508ace48861010337388f8f0dbc481034346.png

          • kraeuterbutter

            # Beyond 2e: downside:
            its Lighthouse
            i have lighthouse (index and controllers), BUT:
            my new gaming-room has big mirrors.. so i would have to cover them every time i use it..
            no problem with Quest oder Pico with the mirrors
            another downside: index-Controller..
            loved them (once)
            but: thumbstick-drift
            and compared to Quest3 and Pico4- controller they are less ergonomic for MY hands
            AND: cable..
            it would only be possible to use the beyond in this one room
            another Problem: Bigscreen is small.. so tracking the headset with lighthouse is harder..
            it would be better to have 3 or 4 Basestations –> not possible: i have gen 1 Basestations, so only 2 possible

          • kraeuterbutter

            maybe better: Pimax Air (SE):
            maybe same displays as bigscreen beyond2
            but more fov (maybe less binocular overlap)
            SLAM-Tracking – should be no problem with my mirrors..
            still a cable
            but maybe in future option to use it as a video-google for wathcing movies over Smartphone (??)
            there was something..
            if possible and good: could replace xreal one pro i wana buy aswell

          • kraeuterbutter

            pimax Super
            (or much cheaper: pimax light):

            maybe the last chance to wear such a headset !
            in the moment iam still used to 500g+ headsets..
            in the future: i think its a nogo
            so: its the last chance for such a headset to wander on my head

          • kraeuterbutter

            most versatile headset: would be a quest3-succesor, wireless, standalone (with Quest bibliothek)
            and: best for using for work…
            i can not use any of my other headsets, because my companies Notebook is not powerful enough..

            but with a standalone-Headset like Quest/Pico: the VR-hocuspokus is done by the headset itself..
            so even a weak notebook can send its (virtual) screens to the headsets, works fine (immersed)
            so thats a point agains all cable-headsets as well (Sofar – maybe valves deckard ads something completley new to that game.. splitrendering etc. etc.)

    • Mbucchia is awesome

  • XRC

    Mbucchia deserves praise for their work on openXR tools and Quad views implementation, from the community thank you!!

  • Christian Schildwaechter

    I hope so, but sometimes the light at the end of the tunnel is not the way out of the dark, but the train that will run you over.

    • XRC

      Cela prend un peu plus de temps que prévu, mais Bellevue a un autre banger

      • Christian Schildwaechter

        I still have a 1997 Logitech 6DoF CyberMan from my first round of DIY VR, and still treasure Bellevue's 2015 Steam Controller, waiting for its return as a stand-alone device (Ibex?) since Valve told The Verge in 2022 that they wanted to release a successor. So I'm very familiar with the little bit of extra time required, have shown a lot of patience and acceptance for Valve time, and am pretty sure that whatever comes will be great just by extrapolation how insanely great (and mostly underused) Steam Input already is on the Deck.

        But for the last decade VR hardware has been rather limited for what I'd like to do with it, and now we finally get devices with high enough resolution and passthrough that using them as a replacement for desktop machines with added "spatial" features actually seems feasible, allowing completely new interfaces. So while I spent a lot of time knowing that the hardware wasn't ready yet, acceptable because evolving technology simply takes some/a lot of time, I'm now getting somewhat nervous.

        AVP is already pretty close, but way too heavy/uncomfortable, as I am very sensitive to pressure on my face and head, and way too expensive. If my technical expectations for Deckard are correct, this could be the first device that really would allow for some things I've been waiting for for literally decades. So there is a new kind of pressure: up to now what I wanted was simply not feasible, but soon it may be. And one danger is that what I always considered a great use case turns out to be completely unpractical in reality, and all, or at least some of my dreams of the future get shattered to pieces. I trust Valve to deliver a banger, but it will come with a built-in reality check of my ideas, and that may turn out to be more of a train to crush me than the end of a 30 year long tunnel.
        https://uploads.disquscdn.com/images/e984aef9a917daa65ad51b72de7fa8cf826e5307ad64297a37c739c41716c9d5.jpg

  • DeeW

    Will it work with the Samsung Odyssey + ?

    • XRC

      Yes all wmr use same drivers

      • DeeW

        GTK, Thanks.

  • Steven Williamson

    Reportedly Microsoft cut a deal with Meta to support their headsets, many suspect this is why they intentionally killed off WMR, so I will be surprised if Microsoft allows this driver to be released publicly.

  • Mbucchia is doing god's work for the XR community