Microdisplay Tradeoffs

But the benefits of using a microdisplay don’t come without tradeoffs.

A 2-inch display (for a single eye) is the absolute ideal for VR, says Kopin. Traditional VR displays—like those found today in the Rift and Vive—are limited by how small they can be made. Shrinking them down to 2-inches while keeping them packed full of pixels is the challenge. Meanwhile, microdisplays benefit from having incredibly tiny and dense pixels, but when it comes to VR, the challenge is—counterintuitively—actually making them larger.

Photo courtesy Kopin

The microdisplay manufacturing process is similar to how computer processors are created, which involves etching microscopic structures onto silicon wafers. This process is highly specialized, requires phenomenally expensive equipment, and is only profitable at scale. Because most facilities in the world capable of this sort of manufacturing are set up to make the smallest components possible, it’s non-trivial to manufacture a larger microdisplay without a massive capital investment.

SEE ALSO
Samsung's New VR Display Has Nearly 3.5x More Pixels Than Rift & Vive

Kopin told me that, because of the diminutive size of the microdisplay, achieving an immersive field of view is one of the challenges that face this display technology; the image needs heavy magnification in order to achieve a wide field of view. They claim their current Lightning display is capable of delivering up to a 110 degree diagonal field of view using a dual-element Fresnel lens. They are also working in conjunction with 3M to develop an even smaller ‘pancake’ optic which they say can achieve a 100 degree field of view.

Photo by Road to VR

However, having not seen through the headset yet, it isn’t clear to me how much of a toll this optical transformation has on the resultant image quality, nor the extent or the effectiveness of the counter-distortions necessary to counteract the optics; my impression is that achieving that wide field of view from a microdisplay is not trivial. Kopin seemed to indicate the same, saying that larger displays would likely have an edge in the field of view department compared to microdisplays, at least until microdisplays can be made larger.

To that end, Kopin and partners have made an initial $150 million investment to create a new facility capable of manufacturing a larger and higher resolution version of their VR microdisplay, the company says. This is many years further out into the future, likely after the completion of the new facility, but the company’s roadmap indicates they are aiming for a 1.3-inch 3,072 x 3,072 display followed by a 1.37-inch 4,096 x 4069 display, in due course.

Photo by Road to VR

As the displays get larger, however, they succumb to the same optical physics that necessitate the bulk of today’s headsets: an increasing focal length that requires the lens to be further from the display in order for it to focus correctly. Even so, at that size the microdisplay should retain its advantage in pixel density.

So on one hand, you have traditional OLED display makers like Samsung working to shrink their displays, and microdisplay makers like Kopin trying to enlarge their displays. Who will hit the sweet spot first remains to be seen.

1
2

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.


  • Lucio Lima

    Only “110 degrees” ? NO, thank you!
    We need more FOV!

    • TheTruth

      Well 110 is the same as vive/rift/psvr so you’ll be getting it for its form factor and resolution. This is an acceptable upgrade till gen 3 with hopefully wider fov.

      I’m assuming/hoping they’ll figure out integrated wireless before gen 3

    • Joan Villora Jofré

      Actually 100 degrees.

    • REP

      I don’t see a future for this lenses. FOV is too low and foveate rendering will be the norm moving forward, resolution will take a secondary seat.

      • sntxrrr

        You still need a panel with such a high pixel count for the part that’s in focus and looked at.

  • Sch@dows

    Those displays are great, but unusable without eye tracking to make fovated rendering and lessen the burden of the GPU.
    And the headset is so small, I have a hard time believing the tracking is as accurate as others.

    • navika

      It’s just a display. Can a PS2 run on a 4kTV? Of course it can run. Display resolution has nothing to do with the power of the device.

      Power of the device is needed if you want to run content made in 4k.

      So, if you upgrade your monitor to 4k, you will not be able to play any games anymore until you upgrade your PC as well???

      • Sch@dows

        No, PS2 isn’t capable of 4k rendering (1080i max). And there is no need to have a 4k display if you don’t have any 4k sources.

        But with eye tracking, fovated rendering become possible, and with it you can benefit from 4k pixel density where you’re looking at, with just a portion of the computing power required from full 4k rendering.

        There is no use for a display which cannot be used by the most powerful hardware available to the public.

        • JMB

          While I share some of your sentiments it is simply not true, that a super high res screen wouldn’t afford a considerably improved experience even if that res cannot currently be rendered at a native level and/or be supported by foveated rendering. The main issue with the current generation of HMDs regarding visual fidelity is their screendoor effect, which a higher res screen would reduce considerably EVEN if the displayed resolution were the same as it is now.

          • Sch@dows

            Indeed, I won’t deny (since it is the first thing that jump to my face the 1st time I put the Vive on. But it can also be improved by other mean, like the tech use for the screen (ie: PSVR who has almost no screen door effect despite its lower resolution).

            But, VR also lack image quality. This can be related to lack of details/complexity because the hardware feeding the headset isn’t powerful enough, but also because of the image definition which results in an image way more aliased or blurry than normally.
            And both of these issues can only be improved by either lowering the stress put on the hardware, or increasing the rendering resolution.
            And if that’s not done yet, it is because current top-end PC lack that kind of power, so there is a absolute need for eye-tracking and fovated rendering to reach those improvements.

            The fact that I put image details & complexity above the screen door effect, is a personal opinion obviously, based on the fact that after 5min of immersion, I completely forget SDE (while I still notice the lack of image quality).
            Some people might think the opposite, or even think that increasing the FOV is the number one priority above everything else mentioned above.

    • Sam Illingworth

      Eye tracking and foveated rendering already exist and work and will only get better with time, so there’s no problem there is there?
      And I get the impression this is to show off the display tech, it’s not a product they’re trying to sell you. It doesn’t matter what tracking it has.

      • Sch@dows

        The display tech is great, for sure. But the article is also (and mostly) about how that tech was used in a compact headset, way smaller than the Vive and Rift). But if that headset doesn’t provide at least the same level of functionalities (tracking) or doesn’t let you use that tech properly (eyetracking must be in the headset so ou need room for it) … the comparison is flawed.

        • Sam Illingworth

          I don’t think eye tracking will add much weight – the addon for Vive is tiny, can’t weight much. Same goes for tracking sensors, I reckon.

          I think perhaps the adjustability adds quite a bit of bulk to the Vive maybe?

    • Stefan Küppers

      I wonder as well if the statement is to broad.
      In my mind for example you would need signifcantly less processing power if you only want to display 4k stereo video compared to a complex game.
      Sure you could run into bandwith and decode problems but overall it should be manageable.

      • Sch@dows

        Indeed, it might but better suited to whatch movies, but in that case, it shouldn’t be compared to VR headset, and more with headset like Sony HMD series and the likes.

        • craylon

          yup
          I wonder if we need subcategories in the future like gaming headset, video headset, work headset.

          If the device is a lot lighter then the vive and can improve upon a 2-3 monitor installation at every day work tasks like spreadsheets or web developing I would be interested in buying such a device even besides my vive.

    • sntxrrr

      Tracking shouldn’t be a problem since the FOV is still equal to what’s currently on the market. The smaller hardware components might actually help placing the IR lights and cameras in the HMD

    • Get Schwifty!

      Good point about the tracking-

    • Brent

      LOL how would the tracking suffer from a smaller size?? and who cares about eye tracking just get a gtx 1080 or dowwnsample

      • Sch@dows

        Tracking requires hardware, be it external ones like the cameras and IR marker of the Rift or the Base station and the internal sensors and chipsets to compute positional data of the Vive.
        Since the headset didn’t feature any of those, it only leave us with inside-out positional tracking (the kind of google is pushing for) but that also requires additional hardware in the headset too (cam and chipsets).
        So unless the headset is planing on mimicking the atrocious tracking, imprecision and unresponsive behavior of the gear vr, it needs space in the headset to put things in … hence too small = is there even positional tracking in this headset.

        As for the computing power, even 1080 Ti struggles to get good performance/framerate with an image quality that is still inferior to what you can get with normal displays, to due to the lack of resolution, Cranking up the rendering resolution to fit a better display, won’t certainly change that fact (it will be even worse), 1080Ti or not (and even if it was possible, that would limit VR to only a handful of users).

  • remosito

    > I got to handle and wear a functional prototype at E3 2017, but
    unfortunately I didn’t actually get to see VR content through it since,
    according to the company, the only computer the company had on hand that
    was cooperating with the demands of driving a custom 4,096 x 2,048
    resolution across both displays at 120Hz had to be shipped off to CES
    Asia (another conference which is also running this week).

    that is the lamest cheap ass excuse I have ever heard. And just stinks of their on-site prototype not being functional (something broke?). There is a truck-ton of VR content. Some of which is a lot less demanding than others. Worst case just show a static 3D image to showcase the massively improved pixel densitiy per arc degree….

    All this apart from the total planning and logisitcs failure on the part of the people involved. If you have two tradeshows. You buy/build 4 goddamn systems months ahead of time (a backup for each site and ship all of them seperately). And what kinda people do they have over in asia that can’t even buy/build computer rigs! This is all soo laughable.

    • Sam Illingworth

      Calm down man. Perhaps they’re a very small company? Perhaps it doesn’t matter as it’s the technology that’s interesting, not the people presenting it.

      • remosito

        Yes the technology would sure be interesting if one could actually see it. Or our VR journalists in this case.

      • Get Schwifty!

        The problem here is a medium and tech which is visual simply “needs to be seen to be believed”. Now from a culture which thinks nothing of hijacking technology wherever possible like the Borg, I am more dubious about the viability of claims and want to see the results to back it all up before getting too excited.

        Don’t get me wrong, I am _very_ excited at the possibilities here, especially the vastly reduced physical footprint itself of the HMD, but talk is cheap. If you have killer tech, I think I would want the world to see it. OTOH, it is probably a random, ad hoc encounter with someone from the business and not a planned review. I am guardedly optimistic until I see a real review of the visuals. Love also to know just how much this might cost, etc. but since we are talking “many years” for a factory to even produce the displays, might as well forget about it for now.

    • Gabe_b

      Also, you can just provide a lower resolution input, it’ll still look better than on an existing display due to less screen dooring

  • Rigelleo

    Since optical aberrations are inversely proportional to focal length, such a system probably suffers from poor image quality. It does not matter if you have millions of pixels if they are not visible because they are blurry

  • I’d rather it be the same size as the oculus if it meant ditching the Fresnel lenses. After using the vive and owning the rift, Fresnel lenses are an absolute dealbreaker for me on future headsets. The glare from these lenses makes VR desktop and movie environments like bigscreen unusable for more than short periods of time due to eye strain. Hooking an HDMI input to the PSVR control box and watching a movie is night-and-day more comfortable than the rift, despite the limitations and lower resolution, entirely because of the lack of Fresnel lenses.

    Virtual screen environments will be the first broadly usable technology to benefit from high res displays, particularly with laptop based VR with the MAX-Q form factor coming shortly, and Windows Mixed Reality bringing desktop apps to VR as a base-line for everyone. It’s time manufacturers realize these lenses don’t cut it, and come up with more creative solutions.

    • Get Schwifty!

      I agree – Fresnel lenses are needed currently but we need to get past them ASAP.

      • RFC_VR

        Fresnel in Vive very noticeable, tried Rift recently impressed with hybrid Fresnel lens. Then tried PSVR (Ps4 pro) with Farpoint very impressed with lens and display panel.

        • Andrew Jakobs

          And to think the resolution on the PSVR is even less than both the Vive and Rift, and yet still delivers a very good image without any distraction SDE (I was expecting a lot of SDE hearing from people about it, as I have the DK2 which has pretty much SDE, but then I had a PSVR on my head a couple of weeks ago and I was like, WTF, there is hardly any real distracting SDE, people are just bitching like hell)..

          • RFC_VR

            Yes I was very impressed with the Sony RGB display panel in the PSVR, if you read up on their tech (sub pixel density) it’s optically impressive and a real contrast to the Samsung PenTile panels in Vive and Oculus​.

            Perhaps we should not be surprised lest we forget Sony’s decades of class leading consumer electronics and display manufacturing…

            Minimal SDE and non fresnel lenses gave Farpoint a very clear line of sight with no noticeable distortion.

            Perhaps just some mura correction needed (software patch?)

    • Totally agree.

    • sfmike

      One more person who hates the fesnel lenses. The GearVR is superior for the fact they are not used. I’ll take some color aberrations on the edges over the god rays that make it seem you have cataracts at an early age. Fesnel lenses have to go.

    • kool

      I was going to say I’ve had no problem with extended sessions on the psvr. Thanks for explaining why.

  • Interesting technology. Given the consideration of page 2, I wonder how much this tech will help in next gen VR headsets, but it is cool nonetheless

  • I don’t really care about the headsets staying the same basic size as they are now as long as both the resolution and field of view numbers continue to increase and they eventually go wireless too. Making the overall headset smaller is properly the least important thing to me at this point in time. And anyone who actually has a clue about VR should be thinking similarly; there’s zero point is obsessing over making it smaller until we can actually nail that other stuff in at least the current form factor. Not that I mind anyone also trying to reduce the size too–but there’s zero chance I’d buy a smaller headset if it means having to go backwards in any of the actually important areas in making VR a truly immersive and great experience.

    • VR deluxe

      I would happily wear a race helmet size VR device right now if only it had double the resolution of the rift and vive. SDE and god rays make the current experience worthless

    • Master E

      I’m hoping these little companies creating pixel dense micro displays just means HTC and Oculus are brewing up something sweet.

      And agree I’d rather see resolution and FoV improve, but… one of the problems with VR for the masses (non enthusiasts) is the disconnection it creates from the real world and people around you. Why I think AR/VR HMDs and smaller more practical gear IS something they all have to have in mind.

  • Foreign Devil

    What is the original purpose of making microdisplays before VR? I can’t think of any. In which case it’s surprising this couldn’t be optimized in size for VR.

    • benz145

      Lots of high-end usage: think military HUD displays, professional camera viewfinders, etc.

    • projectors

      • Foreign Devil

        Ah now that is the only use case that makes sense to me. . because you will need to magnify a tiny postage size 4K screen throught he projector. . the other stuff like HUD and camera viewfinder doesn’t makes sense since the small size and pixel density if overkill for those.

        • The Home Cinema Projector market is always bouncing along. Companies like JVC with their DLA models and Sony with their VPL models have high end 4K Projectors costing between $5,000 and $10,000. They use a single 4K panel too (although they are not exactly LCD but based on a more advanced version of it). The good news is that they also use advanced optics so hopefully it will be a cross-compatible market and costs will drop all around. Many other projectors use standard LCD panels. It is a big market and will certainly help VR.

  • El_MUERkO

    Wider FOV is the biggest future selling point for me, it’s the limited FOV which snaps my immersion most, I’d love a duel card setup, one running each 4kx4k screen, with a nice 120 degree FOV, that’d be heaven :)

    • Harmen

      I just think that I am wearing skiing glasses, that stops me being bothered bij 110 fov.

    • Robert England

      120°? That’s nothing. You can get that in Gear VR simply by modding. I did it, now I can see the entire ‘image’ with none wasted / unseen around edges.
      The PIMAX headsets, (5k and 8k) have 200° FOV
      Wearality 150°

  • psuedonymous

    The big sticking point with microdisplays is the eyebox & exit pupil. As you grow the FoV, those shrink. If you go too far, you end up with a display that if you look directly forward you can fill the entire visual field, but if you look to the side your pupil exits the area within which the display is visible and you see nothing. This is a fundamental optical problem that needs to be solved before microdisplays are viable for wide FoVs without large (and heavy and expensive due to using many solid precision elements) optical assemblies.

  • Kopin-u-late

    You’ve got a bad typo in there. It should be “nonfunctional prototype” instead, ha ha

  • Keopsys

    Yeeeaaw, that’s coming. I personally do not care too much about tracking so if they could make also a ‘just video version’ a bit less expensive that would be great.

    When looking at the picture with the device open, I feel there is so much empty space that they could do a version even thinner with 2 rounds were the displays are. That would be a cool look too, who said the front had to be flat like oculus ?

    • Master E

      Or how about some of the curved flexible lense/screen technology

      Imagine a flexible or curved screen magnifying a micro display the size of a racing helmets visor… 180 degree fov

      That’d be nice

  • ConceptVBS

    Vaporware. Mark. My. Word.

  • Kevin CCIE

    Has no one noticed that there are NO ELECTRONICS in the so called light and smaller Kopin VR headset. No wonder why they gave no demos to anyone and gave the fake – the PC required to run it has been shipped – lie/excuse.

    • benz145

      See the updated article : )

  • impurekind

    I honestly think I’d rather have more filed of view than resolution for now. But obviously I want both in time.

  • Raphael

    I too have a 2k per eye system I’m developing with fellow engineers. Displays are made for vr with up to 140hz. We use a mechanical gyroscope for tracking together with ultrasound emitters and detectors for positional. Field of view is our big selling point at 170 degrees. Design uses a new type of as yet un-made lens so we’re seeking $20,000 to produce. Have raised $80 in one day.

  • Henree

    what graphics card do you need to run that? a 1080 will not be enough….

  • Lucio Lima

    NO, thanks!
    I only buy when the FOV is greater than 100 degree!

  • Ed

    As many others have expressed too, FOV improvements are most important to me. More resolution is nice but it won’t be hugely important to me until increased FOV demands a higher pixel density. I also want to see lighthouse tracking support on new headsets. There’s no reason for every HMD manufacturer to reinvent the wheel with tracking when we have such a solid and open tracking platform that anyone can use.

  • ZenInsight

    At this point if it isn’t wireless… I’m less excited.

  • Joe Black

    Accurate 6DOF+RoomScale or bust.

  • NooYawker

    Prototypes are always awesome, because they’re prototypes. Now getting that to large scale manufacturing at a marketable price is a whole different story. There’s lot’s of great experiments going on, let’s see how many actually make it onto gen 2 devices.

  • What I am waiting for is 2k x 2k, a field of view that fills my full vision, no artifacts, oled contrasts, bright display and 120hz.

  • traschcanman

    “4K OLED ‘Lightning’ microdisplay ”

    2K x 2K is 4MP not 4K

    4K is 4K x 2K = 8 MP

  • Surykaty

    I’m a proffessional and I can easily find budget for even 4 quadro cards and a 5 thousand dollar VR headset (the quadro cards should in theory have no problem with such a headset while not playing games). My question is then… why aren’t there any PRO headsets with higher resolution? Why aren’t there any professional VR tools to work with? The whole VR developement seems sooo spazzy.. first big advancements then slumber.. I think everything is waiting for big players to release version 2 headsets with a considerable display upgrade.

    • Robert England

      PIMAX. 8K. 200° FOV ;”)

      • Surykaty

        I have very little confidence in chinese products. Plus the reviews are not that stellar.

        • The FOV has blown every single reviewer away as far as I have seen and that is on early prototypes. Time will tell if the final product is as good as everybody hopes though.

          I have good feelings about it myself. Besides, there is nothing better coming out yet so until somebody announces something else and I assume here, this is still the most anticipated VR headset for 2018.

          As to Quadro cards, I think in the future NVidia may need bring out a pure VR card as things shift from monitors to VR/AR. Quadro are all about support, drivers, memory etc and the recent P series can pack a punch. But it also targets a different technology. Instead of precise viewport rendering and the ability to handle masses of geometry in vram they would also need to focus on hardware for VR like foveated rendering at the driver level, huge resolutions with single board but with one gpu per eye rendering and hardware acceleration of lens transforms. This would then split into two groups again, the pro market and the consumer versions so we have legacy cards like Quadro/GTX and a new breed of cards like QuadroVR / GTXVR cards. All speculation of course :)