Magic Leap reveals more details about its technology | Hacker News Reader

Magic Leap reveals more details about its technology

engadget.com

41 pointsjamsc25 comments

I'm still not convinced. AR through visual SLAM is not extremely difficult. All the demos here are indoor and from the occlusion handling for the first scene (0:13) you can see a slight shadow on the right of the table leg so I assume it's using some kind of structured light to sense the depth (versus multi camera/stereo systems). Outdoor tracking with this kind of approach is very difficult because sunlight basically destroys all the information.

The difficult part is where the system portrays this information to the user, which they have not shown yet. Their patents indicate some kind of projection directly onto the eye but there are obvious problems with that, such as when you move your eyeball. Also, running visual slam on a mobile device at real time with a respectable framerate is another challenge.

For the display they use a vibrating optic fibre. At least according to their patents. The fibre spins vibrates in a spiral pattern and uses a modulated RGB light to build up an image (like a CRT).

The latency on this demo looks pretty bad though.

This patent? https://www.google.com/patents/US8950867

Having a vibrating optic fibre in front of your eyes does not sound like 'something that you won't be shy to wear in public and that it will maintain "normal relationships with people."'

"The OS has to control atoms, photons". Uh, every OS does this already. Unless you mean directly in which case that sounds like some pretty amazing software+hardware if you're instructing atoms and photons directly in the OS.

All these folks have to do is find someone trusted to go get a demo and put everything to rest. A simple tweet from a well-known, non-celeb would do it: "Tried it. Looks like demo."

Benedict Evans from A16Z did pretty much exactly that today;

https://twitter.com/BenedictEvans/status/656678054064865280

He's not completely impartial (A16Z invested in Magic Leap) but Evans seems to be pretty quick to call out underperforming companies even if they're an A16Z investment.

Not sure that is a ringing endorsement.

At first I thought it was the other leap (https://www.leapmotion.com/), it looks like the two can be a perfect combination.

I did as well until I read this comment, thank you! The names are similar, but these are vastly different products.

"We're gearing up to ship millions of these things," Abovitz said. "We're not announcing when we're shipping. But we're not far."

So this is a total SWAG, but I'm thinking that translates to "At least a year, maybe two"

Looks incredibly cool. The technical challenges must be amazing! Can't wait to see one in real life.

Can't wait to see one, if it lives up to the hype.

Up till now, they have only revealed a tiny new part of their technology while emphasizing all that they have revealed before. To me, it seems like they are waiting until they need new funding to talk in public, and then only revealing tiny details so they don't commit themselves to certain features.

What I find most interesting is that even the darker areas of the 3d model are opaque.

  > Abovitz describes the Magic Leap tech as unlike anything you can get today.
  > With virtual reality, you're basically strapping two screens to your head.
  > Magic Leap is something else entirely.

I find this intriguing. Oculus and Hololens are a bit of a nonsense to me as I can't imagine a scenario in which I'd want to strap tech to my face in that way. Even if it was shrunk to Google Glass size, I think the general reaction to that shows that it might struggle to gain traction.

Can anyone think of a different form factor that might offer an alternative?

The patented technology that Magic Leap is using is optical, it projects the image directly onto your eye.

So it will basically just be as small as an optic fiber directed at your eye.

Source: https://www.youtube.com/watch?v=1UF9fJtZHAY

The US patent that use that technology shows that Magic Leap inverted the "camera" to do the opposite.

Source: https://www.google.com/patents/US20140003762

My guess is that they will use it in both ways, as a camera to see the world through "your" eyes and the projector to place the objects correctly in your field of view.

The demo video shown here focuses the "eyes/camera" on the objects and they respond correctly by going out of focus - this means it has to know how the eyes behave.

How do you think the system deals with your eyes moving around? So if you move your head left and swivel your eyes right and focus in close... I don't think I'm understanding how all that works together? They are detecting all movements - head, eyes, lens focus?

Eye tracking have been around for quite a while and it seems that even normal cameras can track the eyes and eye focus (Tobii being world leader).

If they use the technology mentioned here, it would be very high resolution (4K mentioned in the demo video) at a short distance (mm's) at full focus.

But - just as you say, no one yet knows how Magic Leap reads the environment to do proper mapping and focusing on an object in real life (rather than on a screen like some VR headset do) should be very unpredictable.

Hi, do you think your theory could explain this video? http://www.magicleap.com/video/magic-leap-video-3-hd.mp4

Depends on use case - I view Oculus and Hololens as useful in more specialized situations (gaming, etc). Sounds like the folks at Magic Leap are thinking this is more of a mass market device ala Google Glass. In which case, I think the form factor is going to be critical. To get people to put something weird on their face/head, you'll need to provide an amazing, compelling experience. If they were able to make it into something that was less obvious, then people would probably be more tolerant of less functionality.

Still, with the caveat that I'm a total sucker for this stuff, if only from a technical or "gee whiz" angle, I think once you get something that functions properly, hardware and processes can improve pretty rapidly to address ergonomics.

I know the mobile tech analogy is overdone but when you compare an early laptop to the laptops, tablets, and smartphones of today you see a steady increase in capability in increasingly smaller and lighter form factors.

I understand that something you wear is even more sensitive to this than something you carry and there are issues of physics that aren't as relevant when you're still just looking at a flat illuminated display. Still, the power you could pack into something the size of a deck of cards or a thumb drive 10-15 years ago was peanuts compared to what you can buy for $50 or $500 today.

Sure, you could design and build a computer with the power of a Nexus 5 back in 1995 or 2000. It just had to be in a big box on your desk and cost a grand or two instead of $400. Even these early Hololens or Magic Leap or Oculus demos make me smile because those goggles and things strapped to your face could very well become a lot more powerful and slimmed down in 10-15 years.

Honestly I think power and heat will remain the biggest issues with this stuff. Even if you managed to build a device into a form factor that looked like a pair of Ray Bans, where would the battery fit and how hot would it get (especially as they attempt to make the devices more self-contained and less like standard peripherals for more powerful computing devices)?

I really like magic leap but I just don't understand how they make these demos... They keep claiming it isn't manipulated but there's a change of depth of view/it's inside/etc. I hope whatever it is it just comes out soon. I'm tired of hoping it's real.

If the projection is into the eyes, how are these demos being recorded?

Or are these demos just being 'rendered' ?

If you can project it into an eye, why wouldn't you be able to project it into a camera?

I'd imagine it'd be the same reasons as to why a TV or projected screen looks great in real life but terrible when it's through a camera.

So....it's basically a technology that duplicates the retinal projection in Star Trek TNG's "The Game"?

Interesting if it works that well. I could see people using this far more than the others coming out.

I have a feeling that they want to create their own "OS" just to avoid the growing field of patents surround AR and VR right now.

Reply on news.ycombinator.com