Both Oculus and Valve have gone on the record to say that they’d be opening up their respective tracking systems for third-parties to make use of, but after a year, neither company is ready to talk specifics.

So called ‘6DOF’ (degrees of freedom) tracking is critical to virtual reality. VR systems need to know where your head is and precisely how it’s moving through space in order to render a virtual world which moves around you as you would expect to see in real life. 6DOF tracking is also important for adding motion input to allow users to effortlessly interact with the virtual world.

oculu rift teardown (1)
See Also: Oculus Rift Teardown Offers Glimpse of Unique Lenses, LEDs, and More

Between the top desktop platforms, two leading systems have emerged. Oculus’ ‘Constellation’ tracking uses an array of IR-LEDs tracked by an external camera while Valve/HTC’s ‘Lighthouse’ system uses an array of photodiodes to track lasers emitted from two base stations.

But extending those tracking systems beyond the head and hands has a wide range of uses; tracked third-party peripherals could open up a world of new opportunities for VR interactivity. One major use-case is simply mirroring the virtual item—be it a bat, golf club, sword, etc.—to the real object that the player is holding. This enhances immersion because not only is the object shaped and held just as it would be in real life, but the user benefits from all the expected forces like weight, leverage, and momentum from the object’s mass.

striker-vr-arena-infinity-prototype

StrikerVR, for instance, is creating a VR gun peripheral which includes a powerful force feedback module so that it doesn’t just feel like the player is holding a gun, but it feels like the gun is actually firing when the trigger is pulled.

SEE ALSO
HTC is Giving Devs a Big Revenue Share Boost on Its VR Platform

tactical-haptics-htc-vive

Third-party controllers with features that go beyond the first-party offering would also be viable with access to an established tracking system. Tactical Haptics, for instance, is creating a VR controller with ‘Reactive Grip’, a unique haptic solution which can create feedback not possible with rumble alone.

Fove4

Not to mention other VR headsets; FOVE, a Japanese firm making a headset with integrated eye tracking, had to renege on a Kickstarter promise to use Lighthouse tracking, instead opting to build their own tracking system for the time being.

But companies like these—which would prefer to tie their devices into the existing tracking systems rather than create a redundant tracking system of their own—have been stymied by a lack of communication by Oculus and Valve regarding their plans for third-party access of Constellation and Lighthouse.

Both companies have gone on record to say that they plan to open up their tracking systems, but have been extremely tight-lipped about timelines or specifics. I checked in with both companies recently and was told that there were ‘no details to share at this time’.

Continue Reading on Page 2

1
2
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 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."
  • Sam Illingworth

    I think the foot tracking thing can’t come soon enough. It will almost give us full body tracking, letting us play games where we have to avoid incoming objects by jumping and ducking and weaving, etc.

    I wonder which system will b easier to extend. On the one hand Lighthouse involves building advanced sensors into your equipment and doing the processing there in the equipment as well, whereas Constellation just means sticking some little lights on the device, but then you need to update the user’s Constellation system itself to understand how to track the new item, which could also add to the work the computer had to do.

    • Graham J ⭐️

      I’ve been imagining leg bands too. With that and the controllers I think IK could fill in the rest well enough. The thing is you don’t actually have legs in most VR apps today but in the future when seeing others’ avatars is more common it will become more important.

    • BlackMage

      Unfortunately Constellation isn’t as simple as sticking some lights on. There’s an article somewhere explaining how it all works but the devices need to chatter with the camera to pulse the lights at a specific frequency so that each device being tracked by Constellation can be uniquely identified. I think both require the same effort from a 3rd party perspective: buy the things to stick to your device from Oculus/Valve, wire them together so that they can send/receive data with the PC, plug it into the API and you get positional data.

      From the Oculus/Valve perspective lighthouse is overall easier as there isn’t really an increase of computational complexity per additional tracked device. There is no theoretical ceiling of simultaneous devices as each one only cares where it is in space and reports that data. Whereas each additional tracked device for Constellation becomes computationally harder and harder for one or two camera to track on their own. They’ve already had to add an additional camera just to handle 3 devices reliably after all. Constellation also requires gyros to fallback on for rotational data whereas lighthouse is currently using no fallback mechanism at all.

      • psuedonymous

        “From the Oculus/Valve perspective lighthouse is overall easier as there
        isn’t really an increase of computational complexity per additional
        tracked device. There is no theoretical ceiling of simultaneous devices
        as each one only cares where it is in space and reports that data.
        Whereas each additional tracked device for Constellation becomes
        computationally harder and harder for one or two camera to track on
        their own.”

        This is incorrect. The major difference between Lighthouse and Constellation is how the array of marker coordinates is populated. The model-fit process and sensor fusion (the actually computationally complex parts) are identical, and both will scale up in required CPU power with more tracked objects. Constellation will scale in computational power required with number of cameras, but the CPU power required to to blob tracking on a camera is utterly minuscule.

        As for implementation complexity: Lighthouse is pretty complex. Every device needs a precision timer and all devices need to be genlocked, every sensor needs an analog amplification frontend, and every sensor needs a DAC. Constellation also requires genlocking, but the timer granularity is a LOT looser (frame capture start, could even be implemented as a dumb radio pulse), and all the onboard logic required is to step through a set sequence once per sync pulse. You could literally use a shift register to implement it!

        • BlackMage

          Meant software complexity, all the complicated bits of the sensors and onboard parts both companies have already figured out and will probably just start selling them out like that little chip in the article.

          Adding additional devices that an API can track seems pretty much plug and play with lighthouse, the device communicates enough data for its position by itself and the data isn’t polluted by other devices. Valve don’t really have to do much to support 3rd parties from a software perspective in this case. Give an API that reports the positional data in the same coordinates as every other device in the room.

          Oculus need to keep their camera’s recognition software up to snuff to support every third party device, and I don’t think it’ll fly if you suddenly need three cameras to track one or two more objects. Any change in the way the camera works to improve support for their own devices suddenly need to be validated by any third parties.

    • Well it won’t help with kicking or doing your kung-fu moves, it is worth mentioning that jumping, ducking, crawling, and other motions can all be faked with inverse kinetics. After all, you’re body is a mechanism that needs to support itself with gravity. If you know where the head is, you can approximate figure out where the rest of the body needs to be to support it. Add in the location of the hands and, unless you take off and drop your gear, a body position isn’t too hard to figure out.

      • Sam Illingworth

        Yeah, but you can pull your legs up when you jump, it won’t know how well you’ve done that without tracking them.

      • bennymann

        Inverse kinetics ?

  • Graham J ⭐️

    I don’t see why Lighthouse can’t be used already since how it works is understood, hardware is relatively simple to create (IR photodiodes), the processing necessary to calculate position fairly is straightforward, and no interaction with HTC hardware is necessary.

    Rift tracking will be more difficult because Oculus will need to provide access to the software driving the cameras and you’ll probably need a device registration process so it can tell the devices apart. And they’ll likely lock down that whole process (like the app store and their signing system) making it harder for indies to get into.

    • crim3

      Lighthouse peripherals must communicate their position and orientation to the headset via bluetooth.
      Edit: Now, that I think, a custom peripheral could talk directly to the computer, so you are right.

  • NeoTokyo_N0r1

    Great article, nicely written. I hope it applies some pressure on them to announce something sooner : ) I am quite sure that valve is prepping the lighthouse devkit. It’s just that we want it in our hands like yesterday already! besides staying true to their vision of making it widely available. “like usb” I believe was the phrase..

  • veritas

    HTC CEO Cher Wang recently at the HTC Ecosystem Summit in Beijing sort of implied that they will ship out at least 1 million units of Vive (2016?) and come out with “Killer Product”. This killer product could easily be Lighthouse Tracking Module or Collar – I would be my guess.

    • I would bet on positional, maybe even hand, tracking for smartphone VR. Well mobile VR lacks the vast system resources of a $1500 gaming machine, VR through the GearVR is *PRETTY* good. It’s just the lack of positional tracking that tarnishes it. If their device wasn’t married to a particular cellphone brand, like their own, they could even take Galaxy phones away from GearVR. Bet that would tick Samsung off!

      • JoeD

        why would HTC care about positional or hand tracking for phones? They are focused on the high-end VR system.

        • CoD511

          The only reason we have displays available for use in VR is due to phones. Samsung has a 4K one already demonstrated at 5.5″ and if that’s not high end… (yes, it’s more pixels than the HTC Vive; 5 million for Vive and Samsung 4K goes to 8 million pixels.)

  • Brett Wagner

    Chet Faliszek at Valve just tweeted this the other day:

    https://twitter.com/chetfaliszek/status/726526658925064192

    So it’s in process, meanwhile mums the word from Palmer & Co.

  • Smokey_the_Bear

    I’m looking forward to VR accessories, I’ll take anything that makes it even more immersive! :)

  • RavnosCC

    I would love to wrap a bunch of lighthouse sensors around my office chair, so that way I can take a rest every now and again, and stay immersed. Or alternatively use it as a cockpit seat in flight sims/driving games, all w/o having to leave VR.

  • James Friedman

    I just don’t understand why Oculus hasn’t released the touch controllers yet? Are they broken? Are they trying to make them even better? It’s just very fishy to me seeing as there are plenty of games that would support it.

  • Theo M

    To be honest, it would be much better for Oculus to release an API because it would be much easier for developers to use. for example, to make a peripheral for the lighthouse system, a developer would have to put a lot of research into figuring out how to use the sensors, when with the constellation system developers could just strap IR LEDs to something and do the rest in code. also, this would be a much cheaper way to have room scale VR because instead of paying $100 more for a Vive you could just get a rift and use an IR wristband or something.