Magic Leap Joins The Khronos Group

[phoronix]

Phoronix: Magic Leap Joins The Khronos Group

The latest company joining The Khronos Group to promote cross-platform industry APIs is Magic Leap...

http://www.phoronix.com/scan.php?page=news_item&px=Magic-Leap-Joins-Khronos

[angrypie]

I wouldn't be so excited about a company that faked a demo video. And if they deliver anything, it will be a lot less exciting than what they've been promising all these years.

[starshipeleven]

I wouldn't be so excited about a company that faked a demo video.

*faked all demo videos they ever published

[speculatrix]

https://search.theregister.co.uk/?q=magicleap&page=1&source=osd

[coder]

*faked all demo videos they ever published

I'm pretty sure they're the ones who even put out a video of an AR fake moon landing, which particularly bugged me.

But, I will still buy, if they deliver the goods and get us the first wearable lightfield display.

[starshipeleven]

But, I will still buy, if they deliver the goods and get us the first wearable lightfield display.

I'll wait for them to go bankrupt and be bought out by some other company first. I don't trust them to make anything at this point.

[coder]

I'll wait for them to go bankrupt and be bought out by some other company first. I don't trust them to make anything at this point.

It was rumored that their lightfield display was the real deal (which is how they attracted so much funding, like $500 M from Google), but was too bulky and required a desktop for the rendering. It looks like they managed to squeeze enough horespower from a Nvidia "Tegra" TX2, which is also probably handling the entire SLAM workload.

[starshipeleven]

It was rumored that their lightfield display was the real deal (which is how they attracted so much funding, like $500 M from Google), but was too bulky and required a desktop for the rendering. It looks like they managed to squeeze enough horespower from a Nvidia "Tegra" TX2, which is also probably handling the entire SLAM workload.

"rumors", heh. If there is one thing they are good at is spreading misinformation.

The issue isn't just hardware but also software, even conceding that a Tegra Tx2 can do that, they still need a decent AI software to deal with converting sensor input into a virtual 3D representation of the room (so it can create "holograms" over stuff that don't clip through objects and such), and even MS's Hololens didn't really perform that well on that.

[coder]

"rumors", heh. If there is one thing they are good at is spreading misinformation.

These are rumors from investors that tried it, like Google and Baidu, who invested $500M and $800M, respectively. Anyway, the good news is that we don't need to rely on such rumors and hearsay for much longer. According to what they've said, it should be launching in the next month.

The issue isn't just hardware but also software, even conceding that a Tegra Tx2 can do that, they still need a decent AI software to deal with converting sensor input into a virtual 3D representation of the room (so it can create "holograms" over stuff that don't clip through objects and such), and even MS's Hololens didn't really perform that well on that.

They bought a company called Dacuda, which had a respected and relatively mature SLAM platform they were shopping around.

https://www.tomshardware.com/news/zeiss-dacuda-acquisition-tango-parity,33215.html

[DrYak]

Just to give the point of view of someone who is an expert in the field (and at the time of writing these articles wasn't a competitor. But in the past has done tons of HUD research, and might work on HUDs again in the future).

Basically, their claims seem unrealistic.
Beside paid actors, only few select people under very stringent NDAs were allowed to see it in very limited amount of time and under as close as perfect conditions.
Their design looks crappy (in terms of performance. Not about aesthetics). Based on actually fine analysis of the patents, not just on general guess work from the outside appearances.

According to what they've said, it should be launching in the next month.

Unless they have a hidden completely different technology, that they never demoed before, that they didn't even patent yet, boy we are in for a giant let down compared to their marketing hype train.

Don't get me wrong. I'm not saying that the device Magic Leap will be selling will be bad per se. (Common, you're talking to the guy that still proudly owns an eMagine Z800 3D Visor - which is now 10 year old tech by the current standard of Vives and Occuluses - and still makes good use of his Microvision ShowWX+ HDMI despite all the bashing about Laser-beam-scanning tech).

It's probably going to be a fun gadget at AR.

But given the ginormous level of hype, the absolutely outrageous amount of over promising, it's going to be a let down.
No you're not going to play doom live-scale with your co-worker in your brightly lit cubicle-maze office like the teasers animated by weta workshop.
You're going to watch bright shapes in your darklit basement (think whatching iPod ads art with the room dim-light set to the lowest dim level. There are a few pictures of - probably NDA-transgressing - footage floating around).

It's going to be fun for some games, but not the experience they've been (over-) selling.

(just like pico projector are fun for a small to medium video in a darkish room, but are completely hopeless for anything Circlet smartphone are trying to do)

(I'm basically putting this in the same bin as Elon Musk's hype. SpaceX is useful valuable tech. But no, you're not going to use giant space rockets as a form of public transportation within a few years. That's just stretching an idea).