It’s been more than a year since we first heard about Leap Motion’s new, Open Source augmented reality headset. The first time around, we were surprised: the headset featured dual 1600×1440 LCDs, 120 Hz refresh rate, 100 degree FOV, and the entire thing would cost under $100 (in volume), with everything, from firmware to mechanical design released under Open licenses. Needless to say, that’s easier said than done. Now it seems Leap Motion is releasing files for various components and a full-scale release might be coming sooner than we think.
Leap Motion first made a name for themselves with the Leap Motion sensor, a sort of mini-Kinect that only worked with hands and arms. Yes, we’re perfectly aware that sounds dumb, but the results were impressive: everything turned into a touchscreen display, you could draw with your fingers, and control robots with your hands. If you mount one of these sensors to your forehead, and reflect a few phone screens onto your retinas, you have the makings of a stereoscopic AR headset that tracks the movement of your hands. This is an over-simplified description, but conceptually, that’s what Project North Star is.
The files released now include STLs of parts that can be 3D printed on any filament printer, files for the electronics that drive the backlight and receive video from a laptop, and even software for doing actual Augmented Reality stuff in Unity. It’s not a complete project ready for prime time, but it’s a far cry from the simple spec sheet full of promises we saw in the middle of last year.
100 degree field of view just isn’t enough. What kills me is that they could make a feather-weight curved/flexible displays that could handle 180 degrees and you wouldn’t have to wear some giant contraption on your head.
It seems as though you don’t know about the state of the art of head-mounted displays, and the limitations that exist.
Both optics and resolution (and the ability to drive that resolution) are challenges that make anything over 100 degrees FOV a non-starter. There’s a reason why both Magic Leap and Hololens have very limited FOV’s, while costing thousands.
Oculus and many others are doing hard R&D to improve on both these aspects.
Gatekeeping acceptable VR/AR hardware specs is counter productive, got the software working and itterate the hardware over time. AR gear is stupid expensive at the moment, but the potential of the tech is amazeballs. I’ve had my castAR kickstarter kit a few years and despite it’s restrictions it still blows peoples minds when i show them. if leap motion can approximate a hololens/magic leap/meta type headset for a tenth of the cost, then I’m not going to sneer about the FoV, that’ll follow in it’s own time.
100 degree FoV is huge in the AR market right now. Curved displays would be more expensive and you would still need the contraption as it uses reflections for AR.
Ultimately if you want small we need light field displays and those are expensive and awhile from prime time with much smaller FoVs (Magic Leap is 55 degrees I believe). Just be happy someone is pushing the tech in the affordable range for once. Meta used the same technology and charged $2000 for it.
Well, if you’ve got a recipe, please let us know!
The project is open source. Please make a proposal.
You can make feather-weight curved/flexible displays that can handle 180 degrees, but you still are at a focal length which is unusable. You need optics to let your eyes focus at a distance, and still have the screen a few cm from your eyes.
Current AR headsets have a 40° FOV at most, so it’s not that bad. And most VR headsets are around 100° as well, the ones that go much higher (200°+) don’t get much traction and require a lot of GPU power.
A little experiment:
1. Get a piece of paper/cardboard/etc. with right angle corner.
2. (Gently!) place the corner against your skin just above your eye, so the two sides are at equal angles.
3. Close the other eye and notice how far the two edges of the paper angle out.
That is 90 degrees. 110 would be even better. Does that make sense?
This is an amazing project and putting on together right now will cost you about $350 if you already have a 3D printer. Parts are available from a prototyping shop, some are on preorder.
The issue I have is that there is no solution for 6dof tracking of your head. That’s completely left to the user to implement and not a small feat to accomplish! Even with Vive Lighthouses already in place you can’t simple get a sensor unit and everything works. This needs to get inside out tracking of some sort and that’s not a commodity, yet.
I was looking at taking the Lenovo Starwars Jedi Challenges AR apart and see how the tracking ticks. Seems simple enough and worked reasonably well.
” This needs to get inside out tracking of some sort and that’s not a commodity, yet.”
https://www.engadget.com/2019/01/23/intel-realsense-tracking-camera-t265/
” The lengthily-titled RealSense Tracking Camera T265 uses inside-out tracking (that is, it doesn’t need outside sensors) to help localize robots and other autonomous machines, particularly in situations where GPS is unreliable or non-existent. “
Interesting! This might love the issue.
I’m not sure but I believe the LeapMotion unit has a laser projector. This thing also looks like a laser projector. That will give interference.
https://newsroom.intel.com/wp-content/uploads/sites/11/2018/01/Intel-RealSense-415.jpg
Intel recently announced a tracking module dubbed RealSense T265.
It may provide the inside out tracking for the headset that you look for.
However, it will make the headset more expensive and probably bulkier.
A few LED’s, a webcam, and some software will give you 6DoF. I’ve built a couple setups myself.
Thats not inside-out tracking
I am really looking forward for the full scale release. The price is affordable, hardware and software are open source. Don’t have a 3d printer, need to order the parts somewhere…
Jerri Elsworth was looking into converting her CastAR headset into a projector on glass one. This due to the fact that there are two 720p or 1080p projectors over your head when wearing CastAR. I don’t know what happened with that idea or CastAR at all. Seems like something could get worked into solution with leap.