HTC Vive Is Getting A $220 Plug-And-Play Eye Tracking Peripheral Next Month

by Joe Durbin • April 27th, 2017

Eye tracking is one feature that could could benefit the performance and affordability of high-end virtual reality headsets. The HTC Vive should become the first mainstream headset to put that theory to the test.

A Chinese startup known as 7invensun (pronounced seven-in-ven-sun) is announcing it will be releasing  a new eye tracking module for the Vive next month. The module is called the aGlass and it will be available for “limited pre-order sales” next month, according to HTC. The company is referring to this first roll-out as a developer kit, but pre-orders are open to anyone.

According to HTC, the system will cost about $220 USD and will release first in China next month before rolling out towards Q3 in the west.

Unlike other eye tracking solutions that require hardware to be installed at the manufacturer level, the 7invensun devices are modular in nature. The thin plastic overlays can be placed manually inside the Vive headset by the average VR user, according to the company. The eye trackers are designed to be wired directly to the headset over USB. Two separate USB chords are connected to each of the aGlass devices. The two chords are then joined by a USB combiner and fed into the Vive’s single port.

The aGlass consists of two separate trackers built specifically to fit alongside the lenses of the Vive. Each tracker has a halo of IR lights combined with sensors that can track the movements of each of your eyes and eyelids. It is said to support customized lenses depending on the specific vision concerns of the individual customer.

This type of tech can have a variety of use cases but the most immediate is foveated rendering.

Foveated rendering is a process that combines eye tracking and software to adjust the way a VR experience is rendered in real time. With foveated rendering, the PC running your Vive only has to render the greatest detail in the small area on which your eyes are directly focused. This dramatically lowers the cost of the hardware required to successfully show a convincing VR experience. According to 7invensun spokespeople, this tech could allow Vive to run on older generation graphics hardware.

7invensun_renderingCurrently, VR demands graphics cards and CPUs that are among the most powerful that the various manufacturers can provide. With foveated rendering, however, users can lower the workload demanded by their Vives and run VR on older, cheaper hardware from NVIDIA, AMD, Intel, etc.

The aGlass comes with custom software allowing you to manually apply foveated rendering to any HTC Vive experience and the amount of the effect being applied. In a demonstration, we saw the device running with NVIDIA’s VR Funhouse experience with a performance jump from 45 frames-per-second to 90 with the foveated rendering applied. This functionality will only be available with NVIDIA graphics cards at first, according to the company.

According to a spokesperson for Vive, the release of aGlass ties into the team’s stated goal for 2017 which is to “expand the ecosystem” for the headset by providing cutting edge peripherals like this, the TPCast wireless VR system and the Vive Tracker. To that end, Vive is officially referring to the aGlass system as an “upgrade kit” for the Vive.

7invensun is a member of the Vive X accelerator’s second class. This is Vive’s in-house startup incubator that previously gave rise to TPCast and other VR-specific startups.

The aGlass will only work with the Vive upon release. HTC emphasizes that they are not making that a requirement for 7invensun, which has full freedom to develop this hardware for other headsets in the future.

Update: after publishing, HTC confirmed the price and release schedule for the aGlass. That information has now been added to the story. 

What's your reaction?
  • Kimberly Martin

    I would like to see the whole video

    • According to HTC, the system will cost about $220 USD and will release first in China next month before rolling out towards Q3 in the west.

  • Herb Derb

    That’s all the vive needs more cabless and junk attached to it, and let’s face it no one is gonna develop strictly for an add on so i really don’t see the point

    • Lou Cipher

      lol. Nice one Herb. Cables and junk?? You sure you’re not confused a bit. Might be the Oculus you’re thinking about?

      • Me

        I think he has a valid point, the Vive is looking more and more like a gangsta car.

      • Bundy

        What are you talking about? The Vive has more cables and addon’s than the Rift. I do like all this tech pushing VR forward, but personally I’d wait until I could buy a kit with all of it built in.

        • Lou Cipher

          two Lighthouses that just need power vs 3 or 4 cameras that have to be connected by USB to your PC? yeah, the Vive is a real cable storm.

          • Bundy

            I was talking about the device itself. The Vive has that thick 3 cable setup and the headphone cable

          • Lou Cipher

            ah true. it is a bit awkward. there is a new integrated version but some people still prefer the old flat type.

          • Bundy

            It’s all good. In 2 or 3 years none of this will matter. The next gen setups will be much better

    • NooYawker

      Are you pitching for some unrelased hardware or something?

    • J.C.

      You clearly missed the part where it has foveated rendering that’s independent of the game you run. That’s a HUGE deal.

    • Me

      It’s something arcades wouldn’t care much about, but for end users willing to buy a product it might scare them away, not speaking of the added price and integration with all apps.

  • Herb Derb

    No confusion Lou my man, the Oculus is the one with all the Triple A games 🙂

    • NooYawker

      There are no AAA VR games out.

      • David Addis

        What about Elite Dangerous?

        • Adrian Meredith

          or dirt rally? or project cars?

      • Master E

        or Resident Evil?

      • Kev

        Not true. Batman, Dirt Rally, Trek Bridge Crew, ED and many others.

    • GOD (Verified ✔)

      Oculus is a decent alternative for poor people yes. Might as well save some extra cash and just do Gear VR.

      • Bundy

        Pretty weak trolling dude.

        • GOD (Verified ✔)

          But i’m so srs…..Since day 1 Oculus has been inferior and the pricing reflected that. It came with a stationary experience, bundled with motion sickness and eventually poor motion tracking. What you pay is what you get is usually the case in life.

          However at this point, I think Vive is overpriced for last year’s technology.

          • Bundy

            Room scale on the Rift works just fine. There are no tracking issues to speak of. Motion sickness issues are exactly the same on the Vive.

          • J.C.

            The Rift has a smaller maximum area for roomscale, but very few people have the space for that anyway. I use my garage for VR, which offers a large space, but concrete floor and questionable climate control. It also requires me to put the cars outside. It’s not convenient, and the general public craves convenience above all else.
            There’s nothing “poor man” about the Rift, and we really need to stop this whole hatred of “the other” headset thing. This isn’t a sport, we aren’t on teams. Oculus as a company will learn that eventually, and open their store up to all headsets. We, as the early adopters, need to keep our expectactions in check a bit, and stop being hateful shitbags. I know that might be hard, since internet anonymity lets you turn your other frustrations in life into cruel, and often false, comments about others.

            My excitement over VR has waned a bit, yeah. But the tech has a LONG way to go, and if it’s going to survive, it needs people to be excited about it. Don’t drag that down with Team mentality.

          • Bundy

            Well said.

            I have a Rift with about a 2 meter sq set up (3 sensor setup). I’d like bigger but it’s as big as my space allows in the room. I have tried a Vive and my only 2 takeaways were that I liked the longer cords, and the Vive accommodated my glasses better, but the Rift controllers and headstrap were nicer and more comfortable. Otherwise, it played and felt exactly the same.

            All this vitriol does nothing to serve the industry and only makes people look like ill informed idiots.

          • Lou Cipher

            i appreciate your candid thoughts on Oculus exclusives. cheers.

          • iorek byrnison

            oculus will not open their store up to other headsets

            because their store is, in their mind, the single biggest selling point about the oculus: you buy the oculus to access the store

            if any old headset can use the oculus store and oculus content with ease, the rift sees major sales damage. a lot of people have stated that store access is WHY they got a rift. those people would perhaps have gone vive with an open store.

            don’t get me wrong, i think oculus should open up the store – but oculus sees opening up the store as financial suicide, so they won’t be doing that.

    • Kev

      All of which you can also run on the Vive as well.

  • REP

    That looks good but too bad the Vive 1.0 has terrible resolution and FOV. To take advantage of this technology, HTC needs to come out with Vive 2.0 IMMEDIATELY with AT LEAST 4K resolution and more than 110 degree FOV.

    • Full_Name

      Not terrible, but yes, a bump in resolution would be welcome, and show off foveated rendering even more. That said, oversampling makes a lot of difference on the current screens, and with foveated rendering that can be done much more.

    • Me

      Yep, this tech doesn’t make much sense on the current gen

    • Robbie Zeigler

      Res scale 1.8 and up takes care of a lot of that.

  • Mike Moeller

    I have to agree…the Vive is nice, but we need more pixels.

    4k headsets please!

    • Bundy

      Agreed, except the GPU hardware was not up to the job. Hopefully with this kinda thing we can make that a reality.

      • polysix

        well.. DUH, that’s the whole point of stuff like in this article.

      • Superkev

        The GPU’s are up to the task. It’s more about the level of mesh detail, anti-aliasing, supersampling, shadow processing and more that you select. Yeah if you try to max all those things out then in effect you need a 1080Ti etc.. Most of the time you just mitigate some things and there you go. Foveated rendering could be an great help though in some cases. But yeah I agree with the OP – more pixels would be a massive step.

        • Basic

          You misunderstand… A 1080 Ti can push those graphics fine at the current resolution but we need at least 4K to get rid of the current pixel issues – which means pushing 4x more pixels per frame = You’d currently need 4x 1080s.

          • Kev

            Wow. Sorry have to totally disagree. 4X 1080’s is just a ludicrous response because there are numerous techniques to “get rid of pixel issues” as you put it. But the best way is to simply have more pixels in the headset itself to work with. In many ways the higher the pixel density the less anti-aliasing you have to do as it is inherently less visible. But as for the computing power required the PSVR is the prime example as it’s graphics computational power is about 25% less than the ancient AMD 7870 but with very good results. A 1080ti is around 810% faster than a 7870- easily enough to drive a much higher density HMD, heck you could do it with reasonable quality on a GTX980.

          • Basic

            Yes, but there’s a reason you can detect the difference in VR between the PS and HTC (I’ve not tried the Oculus since the DK2 so can’t comment).

            Sony do a number of tricks to try and improve performance (and full marks to them to get VR running on such low specs) but they’re all at the expense of quality and accuracy.

          • Kev

            Duh. A 1080ti is over 800% the speed and a 7700k is 500% the speed, the memory is like 300% the speed. Far less tricks required with the kind of performance PC’s can do – that is the beauty of it, once again your observation about 4x 1080 is just ludicrous.

          • Spuzzum

            “once again your observation about 4x 1080 is just ludicrous.” … lol.. considering Palmer Luckey himself, (Oculus founder and creator), said that in 10 years, 8K (4x 1080) won’t be so ridiculous. … “To get to the point where you can’t see pixels, I think some of the speculation is you need about 8K per eye in our current field of view [for the Rift],” he said. “And to get to the point where you couldn’t see any more improvements, you’d need several times that. It sounds ridiculous, but HDTVs have been out there for maybe a decade in the consumer space, and now we’re having phones and tablets that are past the resolution of those TVs. So if you go 10 years from now, 8K in a [head-mounted display] does not seem ridiculous at all.”

          • Kev

            There is a difference between what we would “like” to see and what is “required” for operation. The part I was debunking is that massive computing power that does not exist is “required” – it isn’t plain and simple. Would it be welcome – of course it is and of course it would. You threw out a ridiculous blanket statement it’s required and there’s no way around it which is totally false and laughable.

          • Spuzzum

            The original statement from Basic, was that although he 1080ti can handle the current resolution, we need at least 4K to get rid of the pixels that we’re still seeing. Your response was that 4K isn’t needed. I think you’re just a little confused. 4K is not what we’d ‘like’… it’s what’s actually being worked on as we speak. You’re the only thing that’s laughable here dude. Get over yourself.

          • Spuzzum

            “There is a difference between what we would “like” to see and what is “required” for operation.” That’s just it.. both manufacturers are working on 4K as we speak.. it’s not what we’d ‘like’.. it’s what we require, and what we’re doing about it. This doesn’t come down to ‘tricks’.

          • Kev

            Heh that’s funny you try to change your original statement. We all want higher resolution HMD’s with a wider FOV and more. This is literally your original assertion that I responded to- “GPU hardware was not up to the job”. Somehow you have morphed my response that the current GPU hardware could drive 4K HMD’s using countless techniques into I don’t want 4K. Of course I want 4K and I want a much wider FOV too – I merely think our GPU’s are not holding us back as you asserted. Your assertion in effect that brute force is required and the many techniques to reduce the load are insufficient is just plain silly.

          • Spuzzum

            lol… slow your roll dude, I was simply responding to your statement about what we’d like to see vs what’s actually required. You made it sound as if what we’d like is out of our reach when it comes to reality. I never said we needed better gpu’s or use ‘brute force’. Foveated rendering will take care of everything. You can take a 3MP image and reduce it to 1.7MP.. just by diverting your rendering resources to what the user’s actually focusing on. So using that technology.. we can have the 4K and even 8K.. without the need for multi-thousand dollar video cards. You’re the only one making assertions here dude. Get over yourself.

    • NooYawker

      I don’t think present gpu’s can pump out 4K at 90 fps per eye.

      • Your Mom

        Give it a year or so. See what happened only in the last year since Pascal got released.

        • Eduardo RFS

          well, probabily volta and vega dont have that performance jump

        • NooYawker

          I hope so. If VR sells more they’ll probably release an “optimized for VR” card.

      • Mike Moeller

        prob not. but considering tech like liquid vr, and an api like vulkan, i don’t see why its not possible?

      • Tommy

        We need more pixels the to eliminate the screendoor effect.
        You can still render the game at 1080p or 1440p if you can’t do better; he benefits in the screendoor area will still be there.

    • SpydreX

      Ummh sorry but 4k @60fps is unacceptable. Until someone comes out with a consumer screen that can push beyond 60fps at that resolution then we should either stick to 1080p or up it to 2k since we can at least get 90+fps on those screens.

    • Master E

      I think 4K headsets being a standard are a lot closer than one might think. Not sure I fully understand what Sony did with their sub pixel technology, but have been really impressed with several programs and the clarity on their HMD. I’m hoping since they have so much experience in the display market that they will be the ones who push things to the next level

      Regardless, when you have companies like Facebook and HTC pushing 100’s of millions of dollars into VR, the outlook is good for hardware getting better.

      On the software end foveate rendering will also create much less demand on the hardware. So hopefully software and both hardware will meet somewhere in the middle and create the experiences and clarity we all want.

      VR FTW!

  • Cl

    I hear this has 5ms latency. If you use it with tpcast will it be more?

  • Me

    Ouch the price… OK for a dev kit, not at all for a consumer product.

    This tech has a lot of potential, but I don’t think it’s a priority for this generation: resolution of the lenses won’t allow full use of the potential, and it’s slowly starting to pile up with all the accesories. The Vive is looking more and more like a Frankenstein invention.

    We need, in this order:
    – Quality content with a clear roadmap on that
    – Lower overall access price (by lowering the price itself and/or the specs needed)
    – Ease of use and ergonomy : we need something light, easy to put on and off, plus and play, and that allows you to keep up with your surroundings (social starts with the friends and family you already have around you, not the ones on the other side of the planet !)
    – Once these criteria are met we can speak about spec and features and start to improve.

    What I see here, wireless, eye tracking, business packages and their overall price strategy makes me think the Vive is going more and more in the direction of a service provided by arcades and such, not as a product for your home. This is in complete contradiction of what we expect in our occidental societies, but more in line with asian ones I guess.

    • Bundy

      2nd or 3rd gen of the Vive/Rift should have most of these add-on components built in. I don’t mind seeing this kinda thing so much because it gets the development out of the way. Then it’s just a matter of cramming it all in future headsets so that we don’t look like a mass of wires and parts.

      • J.C.

        Exactly. It may make the current Vive look weird, but it’s a way to get this tech into the current headset. If it works as intended, and well, I’m betting this company is pitching their design to be made part of the next Vive/Rift/anyothermanufacturer headsets. This thing is proof of concept, at a “see for yourself!” level.

        Anyone buying VR right now needs to understand that we’re in the flip-phones + texting era for the hardware. Remember texting on a phone using t9? It sucked! But texting is awesome, so hardware evolved for it. VR headsets will evolve, and the next ones may be so far ahead of the current ones that you’ll have a hard time selling your original Vive/Rift. If you aren’t comfortable with that, being an early adopter isn’t for you.

        If Foveated rendering can actually give users a large jump in performance, and can be reliably done at driver level (no developer input needed), it’ll allow for higher resolution headsets that run on the same hardware. Higher resolution screens can allow for lower Super sampling, lowering hardware requirements, or at least not raising them much.

        I expect that we will see all of this tech as a standard within a year or two, and we’ll wonder why we were ever impressed with the original headsets. We won’t have realistic haptic response for a LONG time, but I have no doubt the visual side of the tech will be spectacularly better.

        • Bundy

          That’s the big advantage here, higher resolution headsets. 4K headsets were always great on paper but if everyone needed to run dual 1080 Ti’s to get the job done, it was DOA. But with this it’s a possibility. And also headsets with wider screens for a better POV.

          But just wait, the “my peripherial vision can see EVERYTHING, let us see it all in focus!” trolls haven’t emerged yet. And they will, without a doubt they will.

          • J.C.

            Oh, I’ve got no doubt there will be people complaining about it. At least there will likely be an option for turning it off, so they can see that no, they are not some sort of evolved creature that has different eyesight from the rest of the population.

            Thanks for being one of those people with the understanding that 4K isn’t gonna work at 90fps without INSANE hardware running it. Most of the people on here want 4K, and they want it for half the price of the current headsets (or less!), and they want it all to run on their 3 year old laptop’s video card. Not gonna happen.

            Adding in wireless and/or foveated rendering + eye tracking is going to kill any possible price reduction for new headsets. I fully expect the next gen of headsets to be the same price or maybe even a bit more expensive than the current ones. These things are still the Ferraris of display tech. Everyone thinks they’re cool, but few people allocate the resources to get one.

          • Bundy

            I suspect the resolution increase/eye tracking will the first components to get added into the next gen kits. They’ll provide the most noticable improvements. But, that also further delays a wireless solution as bandwidth also jumps up.

            I’m not completely sold on wireless at this point. The latency, especially with higher res could be a big issue and the battery life I’ve seen so far kinda stinks. I wouldn’t mind staying on a cord (as long as it was sufficient length, looking at you Oculus). AR seems better suited to being completely wireless to me.

          • silvaring

            Do you even know why the Rift and Vive are so expensive? OLED. Samsung have been like the only supplier of these high res Oleds (aside from Tianma who only just started making them for the OSVR HDK2). If LG start mass producing the same panels or if Japan Display reveal a low latency LCD that can compete then it could bring down prices considerably.

          • NooYawker

            Can dual 1080 Ti’s even pump out 90 fps?

        • Diego Rens

          This comment is perfect. Thanks for bringing some sense

  • yexi

    Good article, but I can’t find any specification on the frequency of the sensor ? I Only found ‘High speed’.

    In addition, I know that foveal work great, but “performance jump from 45 frames-per-second to 90” seem nothing with reprojection technologie. It can be an unstable 80fps which becoming an unstable 90fps.

  • zflorence1

    So foveated rendering is going to end up being an external PC computation versus developer designed? I suppose leaving it to the HMD companies will save developers the hassle so it makes sense. If this is the case then it shouldn’t even effect game design work flow, correct? I have to say between this eye tracking add on, wireless add on, and vive tracker, Oculus is going to have some catching up to do with CV2 as far as feature set goes. Hopefully they will integrate eye tracking so it is a step further to become a standard so that developers can take a leap in graphical fidelity.

  • YCAU

    What is a USB chord? Serial over broad spectrum audio?

  • Ted Joseph

    I would take a 180 deg FOV before I take the current 110 with eye tracking. I CANT WAIT until the “blinders” are off.

  • Lavender

    I’m very concerned with the rush towards eye tracking. I dearly hope they are not basing their technology and research into developing something purely for those with 20/20 vision with no problems. A product like this long term becoming integrated into the Vive could leave those with an amblyopia or other vision problems at a severe disadvantage if the technology doesn’t understand what that means and how to correct it for the user.

    • David Addis

      Is there any evidence for VR making Amblyopia more likely to develop? As far as I’ve seen, VR and stereoscopic equipment generally comes with an age warning, so I would hope infants and toddlers wouldn’t be using it anyway.

      On the other hand, separating the images arriving at each eye gives us an interesting opportunity to treat Amblyopia.

      • iorek byrnison

        what he’s saying isn’t that it’ll cause ambylopia

        what he’s saying is that for people who already have ambylopia, an eye tracking system that doesn’t properly accomodate them could render VR unusable, and exclude them

        • David Addis

          You’re quite right. I’m not sure how I misunderstood that!

          On a side note, though, VR headsets could potentially treat (and with good enough eye tracking, diagnose?) Amblyopia, which is pretty exciting.