Back in April Leap Motion first revealed North Star, a prototype AR headset that’s designed to replicate the features of a future high-end AR headset, as a platform for experimentation. Today the company revealed the first extended look of a demo shot through the headset, offering a glimpse of its capabilities.

Leap Motion is best known for its markerless hand-tracking technology, but in the last few quarters the company has been increasingly showing of its design chops in both software and now hardware. Project North Star, which was recently open-sourced, cares not about form-factor; the headset was built purely to push the limitations of the end AR experience, and thus to serve as an experimental platform for what might be achieved with features that will hopefully one day fit into a compact and affordable headset.

Image courtesy Leap Motion

The company claims North Star has “best-in-class field-of-view, refresh rate, and resolution,” and today revealed a new demo designed to show it off, including the company’s hand-tracking tech. Below you’ll see through-the-headset footage of Leap Motion’s table tennis demo on North Star, which has the player facing off against an AI opponent. In the video, the table and the player’s paddle are real, while the ball and the opponent’s paddle are virtual.

The demo appears to use a professional motion capture system for tracking the headset and paddle, while the user’s free hand is tracked with a Leap Motion sensor on the headset. The user uses a pinch gesture to spawn a ball for each volley.

Toward Truly Glasses-sized AR: First Look at DigiLens' AR HUD Reference Headset

Leap Motion has no plans to manufacture the North Star themselves, but believes the device could be produced at $100 per headset at scale, which could make for a excellent AR development kit.

Disclosure: Leap Motion’s Barrett Fox and Martin Schubert have recently published a series of guest articles on Road to VR highlighting their experiments in AR/VR interface design. The latest piece is here: Validating an Experimental Shortcut Interface with Flaming Arrows & Paper Planes

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."
  • JJ

    I can’t wait for this

  • MarquisDeSang

    What is the point? I prefer to play in full vr, not alone in my basement.

    • mellott124

      It’s a tech demo. It’s quite hard to do tracking that well.

      • JJ

        its practically impossible, the finger articulation is starting to get crazy when before it was pretty bad.

    • JJ

      yeah because other headsets are meant to be used outside in the public….???

      This ar tech from leap is massively beneficial to every aspect of vr, if you cant see it’s instant benefits in hand and finger tracking as well as wide FOV AR display then you are clearly retarded.

      f off troll

    • Facts

      I wouldn’t leave my basement with this ugly giant thing lol

      • dk

        it’s plugged in a pc …and it’s meant for developers

        • Ben

          The ugly/giant attributes are part of the lower cost, doubtful those attributes will change if a consumer version is made available.

    • crim3

      Research and development.

  • Daniel Lopes
  • gruguir

    “The demo appears to use a professional motion capture system for tracking the headset and paddle, while the user’s free hand is tracked with a Leap Motion sensor on the headset. The user uses a pinch gesture to spawn a ball for each volley.” Wait what ? :(

  • MosBen

    Leap Motion seems to operate like Valve, which has so much money that they can afford to just tinker forever until they get a product that they think is good. Unlike Valve, however, Leap Motion gets all of their money from investors, instead of a steady stream of reliable income from Steam. At some point Leap is going to have to release a finished product that people can buy, and that’s the point at which they’ll either become industry leaders or the whole house of cards will come tumbling down. I hope that it’s the former, and their tech certainly looks impressive in videos. That said, the longer we go without any real indication of final specs, prices, and release dates, the more I think that Leap is going to drop the ball, their investors will mutiny, and they’ll end up being bought up by one of the major players in order to have their tech incorporated into some other product.

    • JJ

      well said

    • dk

      at the moment they want to sell software to companies not hardware …but it could be interesting if that changes in the future

    • NooYawker

      People are confusing Leap motion with Magic Leap. Leap motion has released products that track hands and fingers to control your computer and VR motion controls. This AR device is new.
      Magic leap is the company that puts out cgi videos, hype, bullshit and has a lot of investors.

      • MosBen

        Huh. In my defense that seems super easy to confuse, but thanks for the correction.

        • NooYawker

          Pretty sure everyone has confused the two companies before. I have as well.

      • Brian Williams

        Whoa… you’re right. I thought this was referring to Magic Leap too. I was sitting here like: “That AR tech doesn’t seem *that* impressive???”

  • Rick

    Looks promising

  • oompah

    at least it looks not so bad

  • Trekkie

    I find it odd.. Magic Leap gets a bad rap for seemingly over ambitious demos that don’t reflect the functioning of the real device and is compared to Therapist. Leap Motion releases a demo which is clearly something the $100 device will never be able to play, unless augmented by external tracking and processing thus pushing costs to over $1000 and no one complains. Leap Motion is half-baked and flaky, is yet to release a mobile solution and the existing PC unit adds one more cable to the already bothersome headset cable. Something tells me it’s all about fooling potential users with one thing – money. Just tell them you will get the moon for $100 and they will come to you in droves. ML made the mistake if saying they will be priced like a premium laptop and because of the “sour grapes” phenomenon most people will poke fun at it. I’ve nothing to do with ML but as a developer I compare specs before comparing prices so that I know I am not comparing apples and oranges.

    • Johnatan Blogins

      The only thing in common between Leapmotion and magic leap, is the word “leap”…

    • Ben

      I get a very different impression from Magic Leap’s demos, and what Keiichi Matsuda has been showing on twitter from Leap Motion and the North Star prototype…actually, the impression I’ve been left with is almost exactly the opposite of yours.

      • Trekkie

        The key word is “impression” :) The “impressionable” ones are very good at falling for these marketing tricks. Not sure you are aware of this: Matsuda is renowned for the great short film “hyper-reality” showing a fully augmented world. I enjoyed it but that stuff is SciFi and 20+ years away. He was then hired by LM where he is busy continuing showing what CAN be done – as opposed to what is available now. The LM hand tracking device (I own a few) is cool but a million miles from commercial production and deploying. As is Northstar which currently lack SLAM, spatial mapping, AI capable compute etc… I see the ML device having all of that right now and is generations ahead. Basically you need to dig a bit deeper to see my point.

        • Ben

          Keiichi has always been upfront about what he is working on for Leap Motion, there is no illusion. What he is working on is experimental UX and he states that directly.

          Magic Leap on the other hand captions their demos with the text: “Shot directly through Magic Leap technology, without the use of special effects or compositing.” You call it “ambitious.” What need of ambition is there if you believe that ML has “all of that right now”?

          That’s the difference here, Leap Motion has always presented it’s demos as research and experiments, and even made them available to tool around with and see the rough edges. They have basically let EVERYONE into the lab to see the evolution of their technology.

          Magic Leap on the other hand tells us they have something tangible for years. They only let a handful of people try it out, and far fewer are allowed to talk about it. We are presumably months away from an actual product from them and know little more than what it will look like aesthetically, and that the FoV is a little larger than HoloLens.

          I would be happy to dig deeper if there was actually some depth there to dig for, but guess what, you can simply brush the dust off the top of 3 year old news about Magic Leap and pretty much be up to speed.

          • Trekkie

            Making a self contained device which can do SLAM for full positional tracking, Spatial mapping with proper occlusion, hand tracking / gestures and a powerful GPU for graphics/AI is not a joke. Its much harder than cobbling together bits and pieces of tech and showing the semblance of AR. Take a look at their UE4 SDK for detailes – have you ?, Now I dont work for ML but I am a developer/programmer who cannot be fooled and I know where we are in the hardware curve. All I want is people to stop throwing half baked stuff and confuse potential users.

          • Ben

            We need good AR, hope you are right.

          • Trekkie

            True !

    • NooYawker

      Magic leap gets flak for its bullshit marketing and hype. Having actors tell people they’re going to change the world is just annoying crap.
      Leap motion at least shows they have an actual working device even tho it’s in its early stages.

      • Trekkie

        FYI certain members of my team have access to a device and it works. A project is underway. I am working with the SDK – I have a fairly good understanding of what it can do at the C++ level. ML is a million miles ahead of anyone in achieving the dream of true are AR like the world in Keichi’s awesome film. The tech is being closely guarded as it’s WIP and so only a few of my team have access and it’s locked up. However I agree with u, we are frustrated by the time they are taking to get a device out to even guys like me. But as a programmer I can tell you it’s a hard problem to solve and when companies claim it can be done for $100 with 3D printed parts I think it is they who are misleading, even though they may sound warm and fuzzy to you.

  • Johnatan Blogins

    It was my impression from a Twitter reply I got, that the paddle tracking is achieved with leap sensor as well, I’m with you that position is done with some other dedicated solution:

  • ProfessorK

    So basically they are cheating the ball/paddle sync with an external motion detector.

    How then is this handtracking? The hand is attached to the paddle