Valve plans to manufacture and sell new SteamVR Tracking 2.0 base stations in 2018. The new base stations will bring a number of improvements over the current tracking beacons, including the ability to support more than two base stations for a huge tracking volume.

Update (1/9/18): SteamVR Tracking 2.0 has been shown off in a commercial product for the first time at CES 2018 this week. HTC’s new Vive Pro includes the 2.0 sensors, and the headset was shown being tracked with the new, rounded 2.0 base stations. 2.0 sensors are backwards compatible with 1.0 base stations, but 2.0 base stations are not backwards compatible with 1.0 sensors, which means they won’t be able to track the original Vive.

The Vive Pro appears to be on track to be the first product to launch with SteamVR 2.0 Tracking.

During CES this week, Valve said that SteamVR Tracking 2.0 allows the new base stations to be “smaller, more reliable, and offer improved performance,” including a 10 × 10 meter tracking volume. The company this week also affirmed the ability to connect up to four base stations together, though it’s unclear if the 10 × 10 volume requires four base stations or if it can be achieved with just two. We’ve reached out to the company for clarity. Valve has confirmed that four base stations are required for the 33 × 33 tracking volume.

Valve also says that now more than 1,000 licensees have signed up to use SteamVR Tracking technology in VR headsets and other products.

Original Article (October, 2017): Valve’s Joe Ludwig announced via the SteamVR Tracking HDK board today an update on the company’s progress in developing new SteamVR Tracking base stations which will be compatible with SteamVR Tracking headsets built with the TS4231 sensor. The update asks SteamVR Tracking licensees to submit their expected demand for 2.0 base stations month-by-month through 2018 so that Valve can gauge how to ramp up production.

At launch, the 2.0 base stations will only work with the current base station limit of two, but in early 2018 Valve says they’ll expand the system to work with up to four base stations which “should cover a single room play space of roughly 10 × 10 meters [33 × 33 feet].” Ludwig writes that the company is investigating functionality to support even more than four 2.0 base stations, but isn’t committing to a roadmap at this time.

SEE ALSO
SteamVR Tracking HDK Now Available for Anyone to Buy

The 2.0 base stations won’t be compatible with existing HTC Vives since they use a different method for sync timing, but newer headsets and tracked devices in the works will benefit from the 2.0 base stations which will be “smaller, quieter, lower power, more reliable, and less expensive than their 1.0 counterparts,” and offer better performance, according to Valve. An update from June talks more about details of forward/backward compatibility roadmap between base stations and sensors.

Image courtesy Valve

Valve expects 2.0 base stations to being shipping to licensees in early 2018, but notes that supply will likely be limited initially. OEMs will order the 2.0 base stations in bulk (at $60/unit + shipping) from Valve and repackage them as part of their products. HTC currently sells replacement 1.0 base stations for $135/unit + shipping, which could give us some indication of the cost savings from the new base stations.

Though the bulk units sold to OEMs won’t include any mounting solution, Ludwig writes that Valve is “creating a custom wall/ceiling mounting solution,” which they’ll talk more about in 2018. It isn’t clear if Valve plans to sell 2.0 base stations directly to consumers, though so far it looks like the company plans to work only with OEMs.

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.


  • GigaSora

    I do find myself slightly confused on Vive’s continuation of outside-in tracking research. I imagine they know their product will seem silly if someone else achieves good inside-out tracking. Maybe they’re hoping to pivot quickly when the technology is there? Not sure.

    • Joshua Corvinus

      Valve is doing the research, not Vive. Also, SteamVR tracking *is* inside-out, not outside-in. The Vive tracks itself and does not use external sensors. The base stations emit light, they are not sensors.

      • benz145

        Within the context of VR, “inside-out tracking” generally refers to tracking which is achieved without the need for elements that are external to the tracked object.

        • Joshua Corvinus

          That’s assisted vs un-assisted. Inside out vs outside in refers to where the ‘eyes’ of the system are (and which direction they’re pointing), and the frame of reference that results in. Here, have some proof: https://www.researchgate.net/figure/241686793_fig2_Fig-5-Inside-out-and-outside-in-tracking
          http://www.neogaf.com/forum/showthread.php?t=1004254 (this article is from before the misconception you’re referencing entered common discourse)

          • brandon9271

            I’ve had the argument before and it’s like talking to a brick wall. It’s just like the debate over the terms VR, AR, MR, etc. People seem to not care if a term is technically correct or not. It drives me bonkers. Hell, even Nintendo used the term “Sensor” bar for a device that wasn’t a sensor in any shape or form. I guess it’s for simplicity sake because they know 99% of consumers have absolutely no idea how any of this stuff works. If that’s the case they may as well call it magic and tell people it operates on fairy dust.

          • care package

            Well you can correct people all day long worrying about what is technically correct, or you can call things how people know/understand them. I know where you are coming from, but how inside-out vs outside-in is recognized at this point is pretty well established so the comments above trying to undue how it’s used are a waste of time.

          • Impfarts “impfarts”

            Lol. How pedantic.

          • benz145

            I understand what you’re saying, but that’s not how the terminology is used within the VR/AR industry.

            For the use-cases of VR/AR, whether the sensor (eye) is on the tracked object or on the external static reference frame doesn’t fundamentally change how the system can be used. You could for, for instance, replace the Oculus Sensors with a ring of IR LEDs and then put a camera on the headset to “see” them, and you’d end up with the same fundamental experience as an outside-in system, which is why the industry generally refers to all such systems (those which require purpose-built external components mounted to a static frame of reference) as outside-in tracking.

            What matters is whether or not you need those purpose-built external elements to achieve positional tracking, because if you don’t, that fundamentally changes how you can use the device.

          • Joshua Corvinus

            You say you understand what I’m saying but you clearly didn’t read what I said. You keep saying “the industry uses it like x” when I posted literal industry research that contradicts that claim.

            It’s ok to admit you were wrong instead of doubling down.

          • NooYawker

            If there are any external device used to assist tracking it’s outside in. Case closed.

          • Popin

            Incorrect

          • benz145

            I did read what you said. A research paper from 2010, two years before Oculus was founded, isn’t evidence for how the modern VR industry uses the terminology, it’s only evidence of how someone used the terminology at some point.As a VR journalist for six years now, I’m telling you how the industry uses the terminology and why they choose to use it that way.

            Industries frequently form their own jargon, often hijacking colloquial terminology. You can say it’s “wrong,” but that doesn’t change how the industry uses the terms today.

            Another example within the VR space is the term “Presence” which is not the same as the colloquial definition.

          • Popin

            How about the words from the lead engineer who developed Lighthouse? His word good enough for you?
            https://youtu.be/xrsUMEbLtOs?t=4m23s

          • Popin

            Also here is the Xinreality page regarding it
            https://xinreality.com/wiki/Inside-out_tracking

            “Inside-out tracking systems
            See also: Systems using markerless inside-out tracking
            Lighthouse – SteamVR”

            As @joshuacorvinus:disqus mentioned you are misconstruing assisted vs unassisted, also known as Marker vs Markerless

            https://xinreality.com/wiki/Markerless_inside-out_tracking

          • psuedonymous

            A common misconception. The directions photons are flying, where the ‘camera’ is, etc, are all distractions. The fundamental difference between Outside In and Inside Out is what the coordinate system is referenced to. For Outside In, the coordinates are fixed to the world. For Inside Out, the coordinates are fixed to the tracked object. Lighthouse is an Outside In system, because it locates objects relative to the basestions.

          • psuedonymous

            Many oversimplfy this because camera-based tracking is /almost/ the only optical tracking method in wide use. The underlying mathematics of having a fixed camera and mobile markers, and having a fixed Lighthouse basestation and mobile sensors, are identical.

          • Joshua Corvinus

            This is the best counter argument I’ve gotten so far.

        • kontis

          Inside-out and outside-in terms were used in the context of VR long before anyone showed a working markerless, self-contained optical tracking. Inside-out usually meant camera on the HMD + fiducial markers or LEDs in the room. I even remember Carmack using it that way in the DK1 era when he was suggesting solutions for the DK2.

      • GigaSora

        I dont really see the difference if it still needs external objects to track. Its outside in.

      • Lucidfeuer

        That bullshit argument again? So I guess the recently released “Mixed-Reality” headset really are Mixed-Reality aren’t they?

        In everyday, casual or professional, home or conventions usage, there is no difference between the Oculus’s indirect outside-in tracking and Vive self outside-in tracking, but a HUGE and clear differentiation with untethered (as in external device/accessory independent) inside-out tracking with internal cameras/IR.

        This dichotomy might have been relevant when we were still figuring out which method would be the most efficient for the first headsets, but we’re far beyond the first headset with one only relevant (because convenient in size, installation, space, material, cost, limitation, functionality etc…) method which is inside-out tracking as everyone from Google to Asus, Samsung or Oculus is gearing towards.

        • Unimpressed

          Lol there is no difference? How about the absence of USB cables running all over the play area? Let’s be honest, as far as roomscale is concerned SteamVR base stations and method are vastly superior to any of the camera-based solutions. Now that the next-gen Lighthouse trackers will be a lot cheaper AND still offer better tracking, I don’t see why it wouldn’t be even more popular and relevant.

    • kontis

      Your confusion reminds me some analysts who were confused about Sony and Microsoft still pursuing traditional consoles in the post-iPhone world. Now PS4 is responsible for something like half of the Sony’s revenue.

      That analyst wasn’t stupid. Mobile gaming is far more popular and far more convenient. So why do consoles still make sense? Maybe there is another kind of value?

      Laser can achieve precision and reliability that computer vision won’t match for, I guess, 20 years and computational cost that will always be much lower.

      • GigaSora

        I like the answer, but I do question the speculation of 20 years. I guess time will tell. Thanks.

      • brandon9271

        Analyst are kind of stupid to not differential between the casual and the “hardcore.” It reminds me of all the focus on mobile VR like daydream and Gear VR. I consider that the casual VR and Rift and Vive are the hardcore. Both have their place. Personally I have zero interest in Gear VR. It’s the Wii of VR. Great if you want to show grandma your 360 vacation photos but if you want to play Doom, Skyrim or Fallout 4 VR you need the “real deal”

      • Lucidfeuer

        I don’t know what analyst you’re talking about, but it’s an established contextual market fact that consoles are on the way out as a device format. The question is when exactly but clearly no further than 4-7 years in the future.

        The same goes for computer vision reliability, although when we actually start using computer vision in the coming year for inside-out tracking I hardly see it not being accompanied by IR and light projectors, there’s no reason one will suddenly replace the other.

    • JDawg

      Inside-out refers to the “eyes” that does the tracking. Human head does “inside-out” tracking but loses tracking when the lights are turned out. Those “eyes” are on the Windows HMD and on the Vive HMD not on an external accessory. There are no sensors/eyes for tracking on the lighthouses. The lighthouses are the Vive’s “light bulb”. In fact I think the Windows HMDs can’t track in a super dark room.
      RoadToVR and others get this wrong over and over again.

    • Durante

      It’s higher-quality tracking, in a larger area, for all objects including controllers, at very low computational cost.

      Asking why they are improving it further is like asking why Nvidia and AMD continue to build GPUs when Intel has an iGPU in every consumer CPU.

      What we have seen so far in terms of computer-vision based tracking doesn’t even remotely match SteamVR tracking *1.0* performance, particularly for controllers.

      • GigaSora

        What we have now, and what we’ll have in the future are very different things. It doesnt matter how big the tracking area is if another headset has infinite area. It also seems that people are forgetting the most important win of inside out, giving the headset a sense of context with the outside world. Advances in AI will make that invaluable to games.

        • Cl

          You’re exactly right… In the future we will most likely have perfect inside out tracking, but right now outside in is a lot better. You kind of made the point yourself. You want them to not come out with anything until inside out is perfected? We would be waiting a while. Besides, I don’t see myself using a hmd other than connected to my computer in my house since standalone will never be as good. Even if they made it wireless I would need to be near my computer and it wouldn’t matter to me if I had a box on my wall or not.

          • care package

            Doesn’t matter if it’s ‘a lot better’ if inside-out will be sufficient.

          • Cl

            Lol that’s like saying why drive a Ferrari when a civic is sufficient or why use a chainsaw when a hand saw if sufficient. Why 4k when 1080p is sufficient. Gets the job done right?

            Also why is “a lot better” in quotes?
            Because its subjective? So is “sufficient”. Except it really is a lot better at tracking and inside out may not be sufficient for many people.

          • care package

            Horrible analogy there, but more than likely it won’t end up being “a lot better”. It’s in quotes because I quoted who I was replying to duuuhhh. Ferraris are a luxury car. People buy civics because that’s what they can afford, yet a Rift costs the same as any new bundled windows VR HMD. The extra cost for a 4k TV is negligible compared to a 1080p, if they are even available anymore. Inside out tracking IS the 4k TV nimrod. It’s the ‘evolution’ of VR, in other words GEN 2.

          • Cl

            The analogies was about being sufficient. That was your argument. Nothing to do with price. Civic is sufficient because it gets you where you need to go and so does a Ferrari.

            You could say 4k is the inside out and my point is still valid. Until 4k tvs became more mainstream they kept improving 1080p panels just like they are improving outside in tracking. When outside in becomes unecessary they will keep improving it. By that I mean inside out is on par with outside in.

            Maybe we will always have both since each have their benefits.

          • care package

            Here is the problem with your car analogy. Anyone would rather have a Ferrari instead of the Civic. The Ferrari is too far out of 97% of the consumers price range. Anyone would rather have a 4k TV instead of a 1080p TV, yet they barely differ in price. Anyone who can afford a MS HMD bundle (inside out tracking) can afford a Rift (outside in tracking). If one is so better than the other, then why the same price? Because it’s a trade off, unlike a 4k TV or a Ferrari. The trade off is (Rift) better tracking but more stationary and more installation. MS HMD less FOV tracking but plug into any capable PC and less installation. If inside out serves to be ‘sufficient’, the average consumer is going to pop for the simpler solution period. I myself prefer the high end experience. Oculus’ Santa Cruz has a camera at each corner of the HMD significantly improving tracking range, so ya, it will get better.

          • Cl

            You know what I mean though jeeze. Why do you need a perfect analogy? My whole point is that just because you think something is sufficient, it doesn’t mean they should stop improving other options. Just being unreasonable. Also that sure IN THE FUTURE inside out will probably be the only option, but right now ouside in is better. I already said that.

      • Impfarts “impfarts”

        It’s a dead end. Consumers will choose microsoft’s all-in-one solution over ugly base stations and cameras in their room.

        • NooYawker

          That depends how accurate MS’s all in one solution is. I’m not counting any tech out just yet. It’s too early in the game.

        • care package

          I would agree. Even if inside out tracking isn’t as good, all it needs to be is good enough. Just like electricity, the average consumer likes the path of least resistance.

        • Unimpressed

          No it’s not. Enthusiast market will always go with the better solution. Lighthouse base stations are demonstrably superior to camera-based tracking (both outside-in and inside-out versions). They provide more precise tracking AND don’t require USB cables running all over the play area.

          Let the average Joe buy MS headsets, high-end enthusiasts will go with the best SteamVR solution (whatever that may be for Gen2, may not necessarily be Vive).

          • care package

            High end enthusiasts go with SteamVR? lol. You’d have to be an idiot to by a Vive over a Rift at this point.

    • Impfarts “impfarts”

      I think so. Every developer I’ve talked to says inside-out tracking is the future. No question. Too convenient to go back external cameras/sensors. There are limitations with tracking outside the fov but I think that will be gone by gen 2.

      • RFC_VR

        wireless standalone headset with inside-out tracking…imagine wandering through your house with the headset updating the tracked universe as you traverse. huge potential for multi-room gaming / experiences.

    • Warren Lester

      Steam VR isn’t outside in tracking. The lighthouses are just beacons, the sensing is done in the headset…. looking from the inside to outside.

      • Andrew Jakobs

        Yeah, but it’s still considered outside in, as it still relies on an external device for tracking.

        • Warren Lester

          We should agree to disagree I think and leave it at that:)

          • Andrew Jakobs

            there is no agree to disagree, it just as it is, it’s just fact. Lighthouse is seen as outside in tracking.

          • Warren Lester

            I know, it’s a common misconception that I find myself trying to correct frequently. The difference, as we know, is where the sensors are. If they’re on the object that is being tracked, it is classed as inside-out. But even inside out tracking requires “devices”: lighting, beacons, fiducials or even just natural features that it can sense. Without anything that can be identified in the environment, it is impossible for anything to determine where it is.

            There’s some more information here: https://www.theverge.com/2017/1/12/14223416/vr-headset-inside-out-tracking-intel-qualcomm-microsoft-ces-2017

    • Caven

      Until that someone else also develops reliable controllers with inside-out tracking, the Lighthouse system still has significant advantages over Microsoft’s implementation.

    • Andrew Jakobs

      You still need controllers which don’t have inside-out tracking anytime soon. And only tracking within range of the front of the headset just sucks.

  • Duane Aakre

    I think the interesting question is what new headsets is Valve expecting in 2018 that will be compatible with these new base stations? It sounds like they are expecting them from more than one vendor.

    Hopefully, in combination with their new lenses we will be getting the second generation Vive (or equivalent) in the first half of next year. It seems like if something is coming in that timeframe, they would want to announce it soon, to mute interest in the Microsoft variants, which are barely generation 1.5 (based on the resolution of the displays).

    • Muzufuzo

      do you think there will be Vive 2 next year?

      • Shawn MacDonell

        Doubtful, HTC’s efforts are likely towards the Daydream standalone HMD at this time; the SteamVR HMDs are likely come from LG and maybe other manufacturers who want to ship out SteamVR 2.0-powered hardware next year. HTC has said they don’t plan to ship out a new headset until there’s meaningful jumps from the 1st gen. Vive; SteamVR 2.0 alone doesn’t make the cut.

        HTC is likely waiting on more financially sound (as in not as expensive as they are currently) eye-tracking methods and improved optics/display technology before committing to a next-generation SteamVR-powered Vive product. With more SteamVR HMDs hitting market over next year, HTC will be pressured within the SteamVR ecosystem to commit to lower launch prices as the Vive originally launched at most likely.

    • Fredrik Sjöborg

      We know pimax is planned release for this year.

    • MosBen

      I think that Holidays 2018 makes sense for a new Vive/Rift. Both parties have been very cagey about what kind of lifespans these devices will have between generations, but the consensus seems to be that we’re looking at something more than annual cell phone revisions but less than gaming console generations, which to me sounds like 2-3 years. With wireless accessories already maturing, computer hardware necessary to run VR becoming cheaper, and higher resolutions screens becoming available, I would think that late 2018 would be a good time for second generation HMDs to come out.

      • GreasyMullet

        HTC has gone on record to say no new VIVE in 2017. If they wait for 2018, smart money is it being close to March’s NVIDIA announcement of Volta consumer cards.

        • RFC_VR

          should have news end of Q1, 2018 – to market Q3?

        • chuan_l

          There will be a flood of headsets —
          In 2018 based on the Qualcomm VR 835 platform. It came out too late this year for the mobile refresh cycle , but ” Oculus Go ” and the HTC portable will both use it.

          The 10 Nm performance & power consumption envelope is a perfect fit for mobile use. It will be a little bit of a step back going 3DOF again but ” Santa Cruz ” addresses that.

      • Impfarts “impfarts”

        Oh, yeah. No way they wait until 2019 for a new headset. Too many competitors coming to the market.

    • A Hyena

      Unless they’ve given up on it, LG was making their own thing they showed off earlier this year using Valves stuff. WEouldn’t be surprised if they take a lil longer to adapt their stuff to the new 2.0 tech, maybe even the knuckles too if Valve has em done by then o/

    • NooYawker

      I think there will be products offering both. Low cost ones using the original stations and higher end ones with new ones. Buyer beware.

    • Andrew Jakobs

      the pimax 8K, and the LG headset at least.

  • K E

    60 bucks each?? So, two base stations and two controllers will be at least $200, and then a headset on top of that? Seems that Valve has given up on growing the VR market. Strange, as a software company, isn’t that what they should be interested in.

    • Jean-Sebastien Perron

      Only Oculus, PSVR and MS will survive VR.

      • cirby

        I own a Rift and a Vive, and have a WMR set on order. You’re hallucinating if you think the Steam version’s going away…

      • NooYawker

        HTC just got a sugar daddy like Oculus so they’ll be around too. If HTC or Oculus relied solely on profits from selling hardware they’d both be gone.
        Steam VR will be the center of all VR. Most companies will be building around Valve tech and Steam VR. Even MS will be using Steam VR.

    • benz145

      $60 each in bulk, HTC is currently charging $135 each. Not clear how much margin they are making, but unless they are doing nearly 2x, it seems like the 2.0 base stations could be quite a bit cheaper.

    • brandon9271

      That isn’t any more money that buying Oculus Touch for $99 plus additional cameras at $59 each and Lighthouse 1.0 is superior as it is. 2.0 will be even better. I don’t think many folks will mind spending the money for an upgraded product.

      • Andrew Jakobs

        except these new trackers don’t work with your HTC Vive.

        • brandon9271

          They definitely don’t work with MY HTC Vive because I don’t own one ;)

  • mellott124

    Very cool. Can’t wait to extend my tracking range. Not terribly excited about upgrading HMDs already though. If they’re 4k then it may be worth it.

  • Sean Lumly

    I think this implies valve knows that a mass-market, wireless headset is on the horizon…

    • Impfarts “impfarts”

      Yeah. They must have hired a really good fortune-teller and saw mass-market on the horizon. Go, Valve!

    • RFC_VR

      Gabe and his VR team talked of this on youtube many months back. the tech is available, it just need packaging together in a coherent way (and the market needs to sell through 1st gen whilst establishing a more mature application ecosystem). Developments in display panels built specifically for VR application (rather than re-calibrated smartphone display) will see big increases in resolution, reduction SDE

      • Sean Lumly

        Indeed.. The pieces are there, and I would bet that Valve would leave this market alone IF there were options that were in line with Valve’s vision, but this is not the case. Most HMDs are poorly differentiated from one another, and many upcoming devices are lacking basic features (eg. tracking on mobile units). Due to this, I think Valve feels compelled to [once again] lead the market by setting the bar to hit.

  • Vak

    Do you think this tracking 2.0 can track more than one headset on the same area without any synch problem ? Like an lan local party on a warehouse with everybody in VR.

    • benz145

      AFAIK, the current base stations support any number of objects so long as occlusion doesn’t become an issue. I believe the 2.0 base stations should be able to do the same.

    • Caven

      Version 1.0 tracking already does this. My Vive shares the same space as another Vive, and one pair of Lighthouses works fine with both headsets. Absolutely no configuration is needed to support multiple headsets in the same space. I rely on the Lighthouses that came with the other Vive, while my Lighthouses remain in the box.

      • gothicvillas

        oh, you run 2 x Vives on the same PC? What GPU you have there? I would love to attach 2nd headset if possible.

        • Caven

          There are two separate PCs in the room–each running a Vive. I have no idea if it’s possible to use two Vives on one PC–except maybe by having one connect to a virtual machine. Even if that worked, I wouldn’t expect very good performance out of a setup like that.

  • Lucidfeuer

    Plot twist: Oculus announces multiplayer tracking with several base stations…which would have been the main interesting with this tracking space extension.

  • Binary Banana

    Still less than Microsoft Mixed Reality devices! :D Unlimited, that’s the solution!

  • Andrew Jakobs

    With this news (of not being backward compatible) I won’t be buying myself a vive any time soon. Don’t know if they’ll release an updated version of the vive which will work with these new trackers..

  • Peter Hansen

    Hm, 10×10 is pretty great, but not actually what we were hoping for, right?

    I personally work in a lab with a rectangular floor space of 8 x16 m. I was seriously hoping for a way of concatenating two square spaces of 5×5 m each to be able to seamlessly cover the central area.

    • Cl

      Should still work with 4. 1 in each opposing corner 10m apart and 2 in the middle opposite of eachother facing the ones in the corners.

  • HybridEnergy

    Nice , more reason to kick the wife out for extra space.

  • 87%ValdeDetta

    HTC Vive should buy Pimax! And make to different HMD, one Highend and one for Midend.