Fraunhofer, Europe’s largest application-oriented R&D organization, recently unveiled a new VR headset prototype at the Augmented World Expo (AWE) in Munich this past week that aims to reduce the size and weight of VR headsets by using high-resolution microdisplays.

German VR publication MIXED (German) got a chance to try out the unnamed prototype headset, reporting that it incorporates four OLED microdisplays, each of them offering a resolution of 1,920 × 1,200 pixels (2,300 ppi) at a frame rate of up to 120 Hz.

Image courtesy MIXED, Tobias Kammann

The headset uses two OLED/CMOS WUXGA microdisplays per eye, giving each eye a hypothetical max resolution of 4,800 × 1,920 pixels. MIXED reports that the displays were demonstrated at AWE running at 60Hz though, something that was likely done to mitigate the massive GPU compute requirement needed to push such a high-resolution and frame rate.

The optical system (cross-section seen below) comes from the project partner Limbak, an IP company developing extreme-performance optics for virtual and augmented Reality.

Image courtesy MIXED, Tobias Kammann

The headset’s fused optics are said to present the user with a seamless view, and boasts around a 100 degree horizontal field of view (FOV), and 75 degree vertical FOV—about the same horizontally as an Oculus Rift or HTC Vive, and around 25% smaller vertically than the two. What’s more, the prototype headset is reportedly a quarter of the weight of standard consumer headsets, which hover around 500g (~1 lbs).

SEE ALSO
Hands-on: Panasonic's 220 Degree VR Headset Uses Crazy Fused Lenses and 4 Screens

MIXED’s Tobias Kammann remarks that Fraunhofer’s prototype headset is “outstandingly sharp: individual pixels are barely recognizable. If you focus on the center of the image, the image impression is clean, detailed and better than, for example, Vive Pro or Oculus Go.”

Image courtesy MIXED, Tobias Kammann

Kammann however reports the prototype includes noticeable optical distortion and some blurriness, giving him the impression of looking at the OLED displays through “the bottom of a [glass] bottle.”

These are still early days though, as the 3D-printed prototype was only fully assembled about four weeks ago—not nearly enough time to perfect the in-software distortion correction required for a viewing experience that truly does the content justice. According to Kammann, the visuals used in the demo were also less than impressive, offering AWE-goers a look at a static scene in a virtual living room and a high-resolution 360-degree photo of a forest.

Still, Fraunhofer’s headset offers a glimpse at a display pathway that some manufacturers may opt-in for in the future; consumers are bound to expect slimmer devices that are more akin to glasses, and less like the big black ‘face boxes’ of today’s generation. That said, it’s still too early to say when a microdisplay VR headset will come to consumers, although Fraunhofer is currently looking for manufacturing partners who will further support the product’s development.

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. See here for more information.


  • Engineer_92

    Something new everyday! Excited to see what we have in 10 years

    • Peter Hansen

      I would like to skip those years though. :)

      • Rex Thorne

        A coma is the best way to do that, but it has some downsides.

        • Peter Hansen

          Yes, I heard of. ;)

    • Rex Thorne

      That’s about when this will be consumer-ready to run with a Geforce 2980ti.

  • nasprin

    Well, the Headset-Resolution nowadays seems to be held back by the GPU computing capability, not the absence of display technology, but the size of the headset is nevertheless impressive.
    Still, i think pretty much every RtVR reader will agree that a FOV less than on modern headsets ist not something we want.

    • kontis

      Incorrect. It’s 100% held back by display tech and the recent slowdown of PPI improvements (Samsung originally planned 4K galaxy Note in 2015). GPU power is almost completely irrelevant for the screens manufacturers make. Oculus GO is 20x slower than high end gaming GPU and has a screen with 20% more subpixels than Vive PRO.
      Oculus has 3 HMDs now and guess what – the one with the LOWEST resolution (the Rift) is powered by the most powerful GPUs and CPUs (PC).

      • Engineer_92

        Not to say you’re wrong, but you’re wrong. Claiming GPU power is completely irrelevant is absurd. There have already been demos of micro-displays with high pixel density. PPI is not an issue here. The industry is moving so fast when it comes to technical advancements and that’s one of the reasons why the Oculus Quest and GO have higher resolutions. Because they’re NEWER. With Quest closing out gen 1, the Rift 2 is obviously the next move. That headset will bring all the R&D from the past few years into one amazing piece of tech

        • bkydcmpr

          Rift 2 has been cancelled. I don’t know when we would get a decent high resolution headsets at this point, sad.

      • Andrew Jakobs

        Incorrect. It’s 100% held back by GPU power…. Hell even the 2080ti isn’t capable of running a dual 4K screen at 90 fps and with everything on high/max/ultra in an AAA game..
        You might be willing to spend $1200+ on a GPU, mainstream isn’t, mainstream isn’t even willing to spend $260+ on the current 1060/6GB which is required for ‘lowend’ PCVR..

        • Who’s this?

          GPU power is not holding it back. GPU power is insufficient but the display tech is not up to standard.

          • Muzufuzo

            GPU speed IS holding VR back. It’s not only about the number of pixels in the screen and refresh rates. Games (and other software) could become more advanced overall, if not for contemporary slow hardware. 2018 cards are only twice as fast for the same price than cards in 2013. Compare that to orders of magnitude that we were used to before. For 600$ you got: 60 gflops in 2003 –> 600 in 2008 –> 6000 in 2013 –> 12000 in 2018. Unless you count ray-tracing performance, then 600$ RTX 2070 has got 45000.

      • Peter Hansen

        Smart. On the point.

        Every smart TV is able to do acceptable upscaling to 4k.

        • Engineer_92

          Right. And smart TVs have to produce twice the amount of frames in stereoscopic 3D??

          The answer to that is no, if you didn’t know lol

          • Peter Hansen

            That is not the point. If the upscaling is done in hardware, this cannot be that time consuming. And beyond that point it is just a matter of single-frame image quality.

          • Engineer_92

            Then what is the point? Because obviously TVs aren’t HMDs. So break it down for me

          • Peter Hansen

            My point is that it would be nice to have higher resolution screens in VR headsets. And that it seems feasible to have a good (as opposed to Pimax) hardware upscaler which relieves the GPU of part of its burden to deliver actual 4k. It does not even sound unrealistic to have an optional upscaler which can be set to pass-through once the GPU power is there or someone in his/her madness has bought two 2080 Ti’s.

            My point is that having higher resolution screens in VR headsets, and less SDE, is not limited by GPU power, as @kontis pointed out. And the Oculus Go, or Gear VR for that matter, is a very good example.

            Was that really that complicated now?

          • Engineer_92

            Not complicated at all. Just wanted to hear it from the horses mouth. I understand what you’re saying but to support the claim that GPU is “100 percent irrelevant” is a bit ridiculous

          • Peter Hansen

            I hope you are not one of those claiming that smartphone graphics power nowadays already is nearly on par with PC graphics power. Because THAT is actually ridiculous.

            The horse

          • Engineer_92

            How you arrived to that conclusion is beyond me. Nowhere did I say they were on par. If you actually read the comment you’ll see that I literally say you can’t equate the two. Like it’s right there lol. Come on man, do better

          • Peter Hansen

            I did not remember. I read a lot all day. And I just said “I hope”. So my hopes came true. Such a great day.

          • Muzufuzo

            Users have different expectations from mobile and PC VR. In mobile you accept lower quality and even worse tracking but in PC you want the best possible for the cost of loosing mobility outside of the “VR room”.

          • Sven Viking

            Yup, expectations are the most relevant thing here, and it’s also the reason why GPU advances may not actually make as much difference as people are thinking.

            If you went back a number of years people would be saying you’d need Quake-level graphics in VR before it could become popular. Now that GPUs can manage many times that quality at much higher resolutions, people don’t think it’s good enough because their standards for game graphics have increased. Once GPUs allow 2018’s AAA-quality graphics in VR, people will see them as outdated because they still don’t match the quality of contemporary flat games.

            Increasing resolution would bring resolution closer to people’s expectations but simultaneously widen the gap between expectated graphics quality and reality.

            The circuit-breaker is likely to be eye-tracked foveated rendering, which should theoretically allow VR graphics to outmatch those of flat games running on the same hardware.

          • Muzufuzo

            In VR everything seems closer to you than in “flat games” and therefore needs to be of higher quality, in higher res and higher frequency. It’s supposed to be virtual REALITY.

          • Sven Viking

            Yup, and my point is that will never happen with just linear advancement since flat games are necessarily easier to render for (one viewport, smaller FOV, lower acceptable frame rate). They’ll always have better graphics quality on the same hardware unless something (e.g. eye-tracked foveated rendering) fundamentally changes the situation.

          • Muzufuzo

            The problem is that a 1000$ PC in 2018 can’t achieve in Vive Pro VR level of graphics that was possible with a 1000$ PC and Crysis 3 in 2013.

          • Sven Viking

            Yeah, but if we were currently in 2013 people would be saying “the problem is that a 1000$ PC in 2013 can’t achieve in Vive Pro VR level of graphics that was possible with a 1000$ PC and Half-Life 2 in 2008”, and five years from now people will be saying “the problem is that a 1000$ PC in 2023 can’t achieve in Vive Pro VR level of graphics that was possible with a 1000$ PC and Battlefield V in 2018”.

            If not for foveated rendering, that is.

          • Muzufuzo

            Wrong. In 2013 with Radeon 280X or GeForce 770 you could easily render HL2 level graphics in Vive Pro at 90 fps. There’s lack of new faster hardware at reasonable price level (200-350$). Most people can at most afford 1060 6GB or 580 which are maybe twice as fast as what I mentioned from 2013. This has to change for VR to flourish. What we need is 1080 Ti priced like 1050 Ti.

          • Sven Viking

            By taking issue with only the details of the example I take it that you accept my larger point.

          • Muzufuzo

            Point is that we need VR which finally looses screen door effect and is clear so text is 100% readable. Without FR GPUs can’t achieve that atm.

          • Yeah, I see the comparison you are trying to make, but they really aren’t the same. Crysis 3 was and is a pretty graphically intense (cherry picked if you will) example. VR isn’t quite the same medium as a 2D screen. There’s quite a few more “interaction” vectors etc. VR is a NEW medium that does need to come into “focus” and will, but shouldn’t be directly compared to computer gaming flat screen graphics; it’s more than just games.

          • Muzufuzo

            ok but more than 5 years has passed since then, you could play Crysis 3 perfectly well in mid 2013 with a 1000$ PC using newest GeForce 780 which is still a decent card, not far from 1060 3GB edition

  • daveinpublic

    Exciting to see, but I wouldn’t be surprised if other companies are already testing these sorts of things, just not going to trade shows and making press releases because they’re trying to surprise their competitors.

  • Exciting stuff! Now if we could just make some progress on Foveated Rendering so GPU’s could actually handle all of these lovely pixels!

  • gothicvillas

    Even smaller FOV… Lol is this a joke?

    • kontis

      Pretty much the standard way HMD industry always thought what consumers want – good clarity, sharp resolution, small form factor, but not wider FOV because it was already “basically IMAX”, better than your TV, why would you want even more? ;)

      This is why a teenager in a garage could disrupt this industry – they were this much delusional.

      After the release of the Oculus DK1 there were many experts and veterans mocking it for being an absurdly pointless and inferior design and then it sold better than any HMD that existed before…

  • impurekind

    More good stuff.

    I’m really liking all the experimentation and pretty quick progress I’m seeing happening in the VR space.

    The future of VR is bright indeed if this keeps up.

  • Andrew Jakobs

    Microdisplays aren’t anything new, that’s what most older vr headsets used.. slamming 2 together for one eye is what’s new.

  • mfx

    That’s so great news. Looking forward those VR techs that doesn’t feel like a fridge on your eyes.

    • Tomas Sandven

      If only they felt like fridges… c:

      (It gets quite hot inside the headset after a couple of games of Holopoint)

  • Lucidfeuer

    The cross-optical system is so smart, but why conceive such a system around a narrow obsolete (and intolerable) FOV we’ve been stuck for almost 5 years?

    • Andrew Jakobs

      I’d rather have the current FOV with perfect opticals (no godray’s or other distortions) than have a higher FOV with godrays’ and other distortions.
      It’s just not easy to have good optics without distortion and wide FOV..

      But you’re right, smaller FOV’s than we’ve currently have is certainly not the way forward..

  • Remosito

    Wouldn’t two wuxga displays per eye give 2400×1920 per eye?

  • Very interesting…