Over the weekend, Oculus founder Palmer Luckey “Threw Oculus Sensors in opposing corners of this room for the hell of it,” replicating a ‘room-scale’ VR setup like the HTC Vive. The result? “Works fine,” he says.

The notion of ‘room-scale’ VR—a large tracked space for virtual reality experiences—has been touted by Valve and HTC as a major advantage of the Vive headset. Oculus, on the other hand, has opted to focus the Rift headset and Touch controllers on seated or standing experiences with a smaller range of movement.

At E3 2015, Luckey told me that an emphasis on seated and standing experiences over room-scale was a matter of practicality, not performance limitations.

See Also: Oculus Demonstrates Their Own ‘Room-scale’ Tracking Capability at E3 2015

“The system may match [Lighthouse] in terms of capabilities, but we’re not trying to push that as something for developers to do. Most of the developers we’ve talked to don’t want to limit their audience beyond a subset of a subset of a subset of users,” he said. “You have people who have PCs that are powerful enough to run VR (or willing to buy one), then of that set, people who are interested in virtual reality. [Developers] don’t want to narrow it then down to people who want to clear out large spaces in their homes.”

SEE ALSO
The Future of Music Creation & DJ Performance with 'TheWave'
oculus-room-scale-setup-palmer-luckey
Oculus founder Palmer Luckey snapped this photo showing sensors in opposite corners of a ~15×11 room.

But the notion that ‘Oculus can’t do room-scale’ has persisted. In an apparent response to critics, Luckey this weekend moved the Oculus tracking sensors from their usual position (both facing the user from the front), to a Lighthouse-like setup with each sensor in the corner of a 15×11 room. Though not exactly a rigorous test, Luckey says that the setup “works fine,” and later responded that the ~15×11 setup “is limited by the room, not the sensors.”

Luckey told me much the same back at E3.

“[Not aiming for room-scale experiences is] not necessarily a technological limitation, I mean all these optical systems have similar limitations in terms of occlusion and in many cases range. It all comes down to what developers are going to make, and we’re not trying to push people to make large room-scale experiences.”

htc-vive-square
See Also: HTC Vive and SteamVR Hands-on – A Stage of Constant Presence

If Oculus’ sensors can track the Rift and Touch controllers as well in a room-scale configuration as Lighthouse, the only major downside to using Oculus’ sensors is that they need to be plugged into the host PC. Lighthouse trackers are ‘dumb’—they don’t send any data of their own and therefore only need to be plugged into a power outlet—instead the Vive headset itself picks up beams from the Lighthouse trackers and sends this data back to the host PC.

If the Rift and Touch do indeed work just as well for room-scale, it adds an additional layer to the interesting platform positioning developing between Oculus’ own platform and SteamVR which aims to support the HTC Vive and other headsets like the Oculus Rift.

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.


  • Carl Usick

    I guess it needed to be stated, even informally like this. I haven’t heard much about the “seated experience” in a while either. I’m sure many people will go to the effort to put together a room scale set up for themselves, but Oculus is probably making a good call to focus on the core use of the tech, which will mostly be in front of a desk, sitting or standing, at least until wireless is doable.

    • Lolo

      You’re right. Even if we have room 5×5 meters it will only works with games like job simulator, where you stand in one place, using an additional hw like a virtuix omni or make couple steps limited by wires. In my opinion most of players will use VR google just like you said – in front of a desk. I mostly play racing games and flight simulators, sometimes cs:go and others fps, so for me room scale is unnecessary. But definitly i will try it, out of curiosity.

    • Bryan Ischo

      As far as I can tell, the Vive approach is just as good if not better than the Oculus approach for every kind of VR, seated or room scale. First, Valve has specifically focused on ensuring that room scale VR works with regards to real life obstacle avoidance; Oculus has nothing for that that I am aware of. Second, the lighthouses are apparently much better at not losing tracking. Third, the light houses require much less CPU processing power to triangulate position than a camera with image processing as with Oculus.

      I do know that the DK2’s tracking is very easy to lose; turn around and you lose it very easily, cover up part of the head set with your hands (such as when cupping the sides to keep it in a comfortable position like alot of people do), move even a little bit too far to the left or the right … it’s not very good tracking.

      Maybe the production version improves on all of this to the point that the tracking is functionally just as good as the Vive’s. But it will still suffer from point one above (no dedicated support for real obstacle avoidance) and three (unnecessary CPU load).

      I think where it wins though is probably on price … which, I admit, is a big deal for many people. Not so much for me though … I’d gladly pay an extra $200 for much better tracking, or better frame rates as my CPU isn’t spending so much time processing camera images ….

      • whitedragon101@gmail.com

        First : Yep oculus haven’t shown this yet. Emphasis on the yet. But it is trivial to implement and not a technology as such. They could just come out with a software update and bang there it is. It could even be implemented by the game makers themselves.

        Second : This is purely controller design not sensors. The Vive Devkit controllers are large wands with large domes on the top. Its the huge raised dome that helps with occlusion not the tracking technology. Its just line of sight, a bigger thing is harder to occlude.

        Third : The difference in CPU power required between the two is negligible. This was even stated in the talk the other day by a Valve Vive developer in his talk about HDR in VR. https://www.youtube.com/watch?v=Y20twV0m__8&feature=youtu.be&t=846

        DK2 : Thats the old system (about 2 years old). There is DK2 then Crescent Bay then CV1 prototype then the tweaked CV1 they are showing now. CV1 has equal head tracking robustness to Vive.

        • Bryan Ischo

          Well, I hope you’re right. Having basically feature parity between the devices is going to allow game/experience developers to target a wider range of experiences more easily.

    • Just standing makes a big difference in open world type games. You don’t want to feel like you’re exploring skyrim in a wheel chair.

  • Foreign Devil

    I’m glad there will be at least the option to try room scale on the Rift. Ultimately room scale won’t really catch on until the high powered desktop systems can go wireless.

    • Goatmeal

      “until the high powered desktop systems can go wireless”
      That is going to take a very long time. I think the better avenue is looking toward more efficient room scale techniques.

      • Foreign Devil

        Well we do already have low latency streaming from Xbox to PC or PS4 to other screen devices. . I guess the challenge is to get the positional data back to the PC at high enough speed, not so much to broadcast the video feedback to the HMD.

      • No, it’s not going to take a long time. 1080p 60 Hz low latency streaming is already a thing, e.g. Valve’s steam link and others. 1200p 90 Hz is trivial if you’re willing to accept more compression artifacts to fit into the same bandwidth. If you can double the bandwidth it will be good enough for most things (waterfalls, flames and noise are notoriously hard to compress; fill the entire screen with that and it still might not be quite OK).

        • yag

          Problem is you really don’t want too much compression in stereoscopy… that would be a shame after all that work to get decent image quality in our headsets.

          • You’re not going to do wireless without any compression; that’s not even on the horizon.

            Nitero is now claiming they have the problem sussed out with 60 GHz wireless. We’ll see when they publically demo the technology.

            The problem will not get worse at higher resolutions (!!!).

            Foveated rendering. You only need to render at full resolution where the fovea is looking; eye tracking and foveated rendering have already been demonstrated. Eye tracking in general has been done a long time, but saccades are very fast and achieving the kind of speed and accuracy required for foveated rendering was only done recently (if the tracking is too slow, you see the cheating when your eyes move). Too late for gen 1, but every gen 2 VR headset will be using foveated rendering.

            8k by 8k per eye takes about as much rendering power and display bandwidth as 3k 16:9 without foveated rendering. There’s not much use going far above that resolution (human eye resolution is about 60 pixels per degree at the fovea; the moon is about half a degree wide from earth).

            Those kinds of resolutions do not seem difficult with OLED (sufficient DPI was demonstrated on lab scale 5 years ago). It is large panel sizes which are challenging. If VR takes off as expected it won’t have to ride the coat tails of phone manufacturers, who have no conceivable use for 16k by 8k screens. I don’t think it will be more than 5-10 years for those kinds of resolutions appear when VR is finally a commercial success (after 3-4 false starts in the last 20 years; who knows if this will be the time; it is never too late to snatch defeat from the jaws of victory)..

            The headset itself will have to do composition (can’t reasonably do 16k by 8k even over several cables).

            Phones are 10-15 years behind; so extrapolating out 10-15 years a phone will have the performance of a 980 ti today (and desktops will be much better still). More rendering power will really be required for resolution then, but it will always be welcome (e.g. real time global illumination, real time raytracing; it is always possible to squander infinite amounts of power).

          • yag

            Oh you’re right, I totally forgot that foveated rendering is now and suddenly much closer than expected (I was very surprised because other sources were saying that a +1000hz eye tracker was needed to do FR, do you have some info on that ? Btw I like to think that Oculus or Valve have currently an even better eye tracker in their lab.)
            Now indeed we won’t need to worry about compression artifacts at all (except for a few % of the image) Low latency transmission will be a piece of cake at any res. and freq.

            Foveated rendering is quite the panacea of VR…

            About the performance boost with FR, I like the idea that the expensive high-end GPUs we are buying now for gen 1 will still do great for gen 2 and maybe higher (not to mention all the VR optimizations to come, we have so much margin with 3D engines only).
            But the gain won’t go very high with actual rasterized engines, the real boost in image quality and performance will come with (you mentioned it) path-tracing engines (see Brigade engine) which are perfectly suited for FR.
            A new name to come : foveated rendering + path tracing = “foveated tracing” ?
            It’s a bit mind-blowing to think that in only a few years we may have really photorealistic worlds in VR (and the fun part is games will look much better in VR than on a flat screen – except for the apparent resolution).

            Now as you said the more complicated and expensive part will always be the displays. We are already lucky to have VR dedicated displays for the first gen, when there is no market yet (and nothing sure). But we won’t go very far in this area without selling millions headsets, as you said.

        • Robert Jenkins

          i have a steam link…meh..its good but the latency is noticeable. low latency doesn’t mean no latency and man…its noticeable. i wouldn’t advise using it for vr. I mean even if u keep a good frame rate the delay makes the game really choppy sometimes. even when i had both hooked to my router. Its good but it aint that good.

      • Ned Hoon

        I dont actually think a wireless option will take that long.You will probably have a peripheral booster you hook into your GPU rather than rely on the actual PC to do it.

    • Liam Armstrong

      I think a springy cable suspended from the ceiling would do the job well

  • Paul Jensen

    They aren’t pushing for room scale because they are Facebook. The will be running things like Farmville. For him to say room scale won’t take off is cazy.

  • CazCore

    yah, i wish people would stop talking about room-scale VR for the home. NEVER GONNA HAPPEN, cuz there will never be a critical mass for it.

    it will be great for arcades/them-parks/location-based entertainment, but its ALWAYS gonna be a tiny percentage of people that will dedicate the space for it in their home. even when everybody has HMDs.

    and from the design-side, it limits what kind of games you can do.

    • CHEASE

      There doesn’t need to be a critical mass for it because it can be implemented in any game basically just as a controller. With VR games having the ability to track arm movements and implement them into the game, tracking whole body movements is not going to be difficult to implement as an option or even post-hoc by dedicated mapping software – IMO.

      • CazCore

        true, but that has nothing to do with my concern

    • Liam Armstrong

      https://www.youtube.com/watch?v=O9U8-4AT7fs
      I’m fairly sure I could do something like this in my relatively small bedroom. Not to mention room-scale can be insanely fun, as demonstrated by this video.

      • CazCore

        yeah you probably could. but then its not room scale. its just stand-in-one-spot which ALL the headsets can do. i’m not sure if you’re INTENTionally supporting my statement or not. :)

  • lagann

    I tried “room scale” experiences with the gear vr in my house and yah….not gonna happen.

    You will need to have a big empty room in the house and fill the walls with foam padding for the eventual hit to the wall you are going to have.

    I can see the niche working as a paid experience some business has set up for clients….but for home use its going to be one of those things where you are excited to try out and once you try it you forget it ever existed.

    Oculus has the right focus here. Most VR experiences will be seated. I’m glad they are focusing on nailing the seated experience first.

    • quintushalls

      Gear VR doesn’t use tracking, as far as I know. It’s for a mobile experience.

      You don’t need foam walls. You just need to be between both sensors. That’s for both devices.

      The huge differences between the devices is that VIve camera. Vive will let you see the “4th wall” and the Rift won’t. You’re blind to that dog entering the space and attacking you, while you could see it green-ish with the Vive.

      • Nil_Einne

        I think you missed the point about foam walls, lagann said:
        “You will need to have a big empty room in the house and fill the walls
        with foam padding for the eventual hit to the wall you are going to
        have.”

        Being between both sensors isn’t going to help you when you hit the wall because you’re too immersed in the game. The front camera on the Vive would at least allow you see the wall if you look at it, but it may not help if youre quite immersed into a particularly active game.

        • noname

          The Vive literally pops a grid in your face when you get close to a wall. walking into things wearing the Vive is almost impossible if you have it set up correctly.

    • igotnthnfraname

      I just tried the Vive yesterday, users of the Vive will have no problem with room scale at least in reference to running into walls. The chaparone system works great, I was trying to move around in the space as much as I could and I had no problems discerning where the boundaries are. Users will not have to pad the walls, it very clearly lets you know when you are getting close. IMO I at tired of traditional video games, to the point of about giving them up, but when we can include movement now we are talking fitness as much as video games, I can say I will spend an hour monday, wednesday and friday on the treadmill and tuesday thursday and saturday in vr. We are talking real bodily reflex and hand eye coordination. we are talking increased heart rate in a shooter or adventure game. we are talking simulated physical interaction in puzzle games. We are talking real wold senses stimulated in simulated fictional spaces. To me this means new experiences to really enrich peoples lives and providing real valuable experiences that could not be obtained otherwise. That is where my money will go, for me that is the future of vr and the true paradigm that is this newly introduced technology.

  • Rob Xsiq

    I like the idea, but really outside of maybe a bit of a step in a direction, I am not seeing it. I think my in world experience should be AR more than VR, Hololens gen 2 will be great for room experiences.
    But I would like a arms length experience. standing up and reaching, bending, leaning over and even taking a step left/right/front/back sounds good.
    Either way, all exciting technology.

  • Nathan Daughdrill

    Room VR will be a thing when the headsets are wireless and I can go to a warehouse and the tracking is like 5,000 feet.

    • Ned Hoon

      I would agree wireless is required.I think the tether is a potential safety hazard for room-scale gaming and if you trip without being able to associate the fall with reality someone could potentially be seriously injured.

    • noname

      I’ve seen some videos of setups like that (warehouse set up to mimic the digital space). They are, in my opinion, the future of VR arcades

  • Paul Donovan

    For people who flipped out over OR’s de-emphasis of room-scale VR, please consider this:
    Have you created a room-scale setup for repeated use? How many people do you think have a room of the right size, that they can keep empty or quickly convert to “empty” all the time? How long would you maintain the setup until some kid, pet, or spouse knocked things over or moved things in or around or complained that your pet vr project is taking up space for hat racks and armoires? I’m not sure if you’re reading the problem correctly. Motorcycle racing is common, but it is not *mainstream*, and these companies are really truly trying for mainstream adoption (just like every other broad spectrum consumer software and hardware manufacturer). He’s not saying that there isn’t an audience, he’s saying that it’s a very limited audience that is going to stay limited as long as people don’t reinvent their living spaces en masse to support VR (in addition to spending thousands on hardware)… The only way this gets mainstream cheap is if it has a mainstream audience, and vice versa.

    Also, remember: If you’re reading every article like this, you’re probably not the mainstream consumer that these companies hope to attract – you’re the VR consumer who’s already in the bag with a fistful of hundreds (like me, starting about a decade ago).