HTC today announced the launch of a new Full Face Tracker accessory for Vive XR Elite. The device attaches to headset to capture eye and mouth movements which developers can use for a variety of purposes.

HTC says the Face-tracking add-on for Vive XR Elite is available staring today, priced at $200. The device attaches magnetically to the XR Elite and connects via USB-C. It adds two cameras for eye-tracking and one camera for mouth-tracking.

Image courtesy HTC

According to HTC, eye-tracking is captured at 120Hz while mouth-tracking is captured at 60Hz. The mouth-tracking camera is thoughtfully placed on a hinged flap to be less intrusive while transporting the headset.

Image courtesy HTC

Data from the Vive Full Face Tracker can be used to understand where the user is looking and how their mouth is moving which allows supported applications to animate avatar faces more realistically, like Meta has shown with face-tracking on Quest Pro.

This is a benefit for social VR applications, but face-tracking can bring a wide range of other improvements to the VR experience, like foveated rendering, analytics, and better understanding of user intent.

In fact, HTC says the Full Face Tracker also adds automatic IPD adjustment to Vive XR Elite. This detects the distance between the user’s eyes and changes the headset’s physical IPD slider accordingly. This means anyone who puts on the headset gets a correct IPD setting automatically.

SEE ALSO
Zuckerberg: Quest 3 Beats Vision Pro in 'vast majority' of Cases in Mixed Reality

Face-tracking must be implemented into VR applications on a per-app basis, but HTC says the Full Face Tracker is designed with OpenXR in mind. Thus it could potentially work with applications that have previously implemented OpenXR-based face-tracking.

“We’ve built this using our Vive Wave platform, with support for OpenXR too, so developing for it is really easy. Because of this, anything built for the Vive Focus 3’s eye and face tracking will work with the new Vive Full Face Tracker,” says Shen Ye, Global Head of Product at HTC Vive.

Image courtesy HTC

Indeed, HTC previously released similar eye and mouth-tracking accessories for its Vive Focus 3 headset, as well as a mouth-tracker for PC VR headsets.

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.


Ben Lang
Ben is the world's most senior professional analyst solely dedicated to the XR industry, having founded Road to VR in 2011—a year before the Oculus Kickstarter sparked a resurgence that led to the modern XR landscape. He has authored more than 3,000 articles chronicling the evolution of the XR industry over more than a decade. With that unique perspective, Ben has been consistently recognized as one of the most influential voices in XR, giving keynotes and joining panel and podcast discussions at key industry events. He is a self-described "journalist and analyst, not evangelist."
  • Yeshaya

    I really wish this was feasible to add on to the Quest 3. I’d pay $200 extra for eye tracking

    • Stealth Ico

      seriously, meta needs to do a quest 3 pro at some point and just add back the face and eye tracking

      it is unfortunate they didn’t think ahead and add some expansion ports or anything, all we have is the usb-c port on the side

      • Cl

        And oled

      • Blaexe

        Even worse battery life and more weight in the front? I’m sure people would love it /s

        This Quest 3 Eye or whatever would be such a hard sell that I don’t see a reason to do it.

        Let’s say it costs 700 Dollars – only some enthusiasts would switch to that and it would be extremely confusing for the general consumer to explain the differences and what the impact would be.

        A Pro model would need to have bigger differentiators, what is probably exactly what they’re planning to do with the next Quest Pro.

        Quest 3 will be fine until the Quest 4, which will hopefully include eye tracking at an affordable price.

        • ViRGiN

          imo eye tracking in quest 4 depends primarirly on developers adoption of it’s tech. those who are interested, have a really decent devkit in form of quest pro. that’s how i always viewed that headset anyway. so far, there has been basically zero developer interest, just like valve knuckles ‘finger tracking’ is still not adopted on pcvr after _years_ of providing devkits prior to consumer release of index.

          • Blaexe

            Why would there developer interest if there’s no user base, not even an expected user base in the near future?

            Of course devs will wait for that. You could have said the same about Mixed Reality on Quest 2. There was not much interest either aside from some tinkering. Now this has changed because of Quest 3.

            Same will be true for eye tracking. Hardware (announcements) first, software after.

        • Stealth Ico

          foveated rendering can offset some battery life loss, and it would fill the niche for a good eye tracked headset with pancake lenses of which currently none exists besides the quest pro

          also 30g more weight is not that much

          • ViRGiN

            how do you want to offset battery life with foveated rendering, when eye tracking itself is a big power drain lol

          • Stealth Ico

            it can help in local rendering, for remote rendering though I understand what you mean

          • Blaexe

            Foveated Rendering doesn’t offset battery life loss because you would invest the additional performance in higher resolution or refresh rate.

            It’s a niche, exactly as you say. Meta is not interested to fill a niche. That’s why they dropped the Quest Pro so quickly.

          • Christian Schildwaechter

            Pretty sure that ETFR is used mostly to improve battery life on AVP, where a lot of the displayed image will not be some fully rendered 3D VR content, but pass-through with 2D overlay at rather high resolution. Limiting the processing power needed for pass-through alone might be worth the eye tracking, as we’ve seen on Quest Pro that enabling pass-through tanks battery life.

            So far ETFR has mostly been used to increase resolution or graphics, but up to now we’ve pretty much only seen HMDs for VR content, not some that run with very high resolution pass-through as default. Using foveated rendering to reduce power consumption is absolutely an option, just not one that has been prioritized in past implementations.

          • Blaexe

            We’re talking about the implementation inside Quest 3 specifically – I don’t see a realistic way that it would be used to improve battery life. After all it uses up battery life on the Quest Pro aswell.

            Also we don’t know specifics about AVP yet. Is that some guesswork or do you have a source?

          • Christian Schildwaechter

            Technically the Quest 3 would need about 20 IR LEDs and two IR cameras for less than USD 10 to implement basic eye tracking, which would add maybe 10g close to the face. Eye tracking through pupil tracking just to select objects is computationally cheap, doing motion estimation fast enough to avoid artifacts in ETFR is what’s expensive. We don’t know how much better motion estimation would have worked on XR2 Gen 2, or how much more powerful e.g. its Hexagon DSP is, so there is no way to tell if it would improve battery life.

            I assume that Meta sees Quest 3 as a “Quest 2 Pro” as in “PS4 Pro”, a mostly graphical update running the same apps. Which would explain the seriously underclocked CPU (27%-57% lower clock for the same core type than SG8 Gen 2, so there is headroom for eye tracking). And eye tracking makes little sense when only implemented in a higher end enthusiast device that’s expected to sell significantly less than the much cheaper base model, as developers won’t/can’t use it in games for interactions, and Meta can’t rely on it for the UI. So I’d guess that not including it was mostly a strategic decision, even though a Quest 3 like HMD could benefit at least in some ways.

          • Blaexe

            “Technically” is not worth much though. So far every eye tracking implementation has been pretty expensive – not arguing about the reasons – and also it did lower the battery life. I’m not an expert to say otherwise and at least I assume you aren’t either.

            If they could have implemented it in Quest 3 at the same cost and without drawbacks they absolutely would have. 100%.

            It’s also exactly what Bosworth said. Too expensive and require processing power.

          • Stealth Ico

            meta dropped the ball on the quest pro hard, should have never released as it did, they only have themselves to blame

            missing depth sensor, shit color passthrough, using the weaker XR2 chip

            should have just waited to make a Quest 3 Pro instead

          • XRC

            Dynamic foveated rendering using eye tracking requires more power budget.

            Getting around 4 hours with eye tracking switched off in my Pimax Crystal and about 3 hours with it switched on.

          • Stealth Ico

            it can help in local rendering as in playing purely in standalone, for remote rendering via wireless vr I understand though

        • Yeshaya

          I think it would do extremely well in the wake of the AVP. People see Apple commercials about navigating their headset just using eyes and gestures, but 3.5k is a lot. If Meta comes in and says you could 50% as well but for 20% the price, I think a lot of people would go for it.

          • Blaexe

            First – you don’t get the advantage of eye tracked navigation until you try it. Second, people would be completely let down because the UX in Quest – eye tracking or not – is substantially worse than AVP. There is no eye tracked navigation to begin with and it would be naive to expect Meta to reach a similar level anytime soon.

            They CAN already say “you can do 50% for a fraction of the price”. Quest 3.

          • Christian Schildwaechter

            … and it would be naive to expect Meta to reach a similar level anytime soon.

            Which is probably very true, but I still can’t fathom how that’s even possible. Someone copied the AVP eye tracking plus hand gestures UI for Quest Pro, and even for Quest 2 with just head tracking and controller buttons within days after the AVP presentation. And Meta changed … nothing.

          • Blaexe

            It’s not about just implementing the basics though. AVPs hand tracking gestures are seemingly very accurate and work in a very large range – that’s not true for either the Quest 3 or Quest Pro.

            If you can’t trigger the gesture reliably then it would only lead to frustration on the user side.

            Having your hands just hanging on your sides and clicking is very different from having to put them in front of your face. Only in the first case it’ll feel “like magic”.

          • Christian Schildwaechter

            The AVP UI clone hacked in by some Unity developers within days worked better even with just head tracking and controller button presses on the Quest 2 than Meta’s laser pointer UI, so neither eye nor hand tracking are valid reasons for the Quest UI being so much worse.

      • Yeshaya

        I agree, but then I used my Index for years and used the Frunk exactly once, so maybe Meta was scarred away from the “just add the port and figure out what to do with it later” approach. I like in general how the Quests start off barebones and people can spend money on add odds like better straps, grips, speakers, etc. Would’ve been really cool to take that to the next level with eye tracking too but I guess we’ll need to wait on that

    • Zack71

      Me too.
      Quest 3+ eye tracking based foveated rendering = Quest 4

  • ViRGiN

    if they want to move units, they need to collab with VR Chat and sell it under it’s brand. Literally noone else needs that. Well maybe ex-furry reffugees from NeosVR – Resonite.

    • Stealth Ico

      your comment is an unpopular take for some reason but this is literally the only market that would shell out the money for this thing lmao

      • ViRGiN

        Yeah, I got downvoted 3 times already by people I have never seen here before xD
        They literally must be vrchat moderators. Literally brand new accounts with 0 activity.

  • knuckles625

    Under 99% of circumstances, I couldn’t care less what a VR headset looks like, however, this is my 1%.

    The industrial design legit makes you looks like squidward or a proboscis monkey.

    • Arno van Wingerde

      I was thinking of a fly, particularly twith the mouth piece…

  • Lucidfeuer

    Just when I thought they could possibly do worse…

  • Scientism

    invokes Groucho Marx/Squidward glasses

  • Xron

    Dang, with this you will look like Cell from Dbz!

  • A few weeks ago on Reddit already there was the news this would have been launched. Btw the PCVR one has been deprecated