This week AR hype machine Magic Leap will finally – finally! – start selling its headset to the public but you'll need to go to one of three AT&T stores in Boston, Chicago or San Francisco, to buy them. And fork over more than $2,000 to get a pair.
Thanks to the company's effort to build buzz in the tech community, we finally got to try the glasses on without having to sign an NDA or go to its headquarters in Florida because it was giving demos at the Intel Capital conference in Phoenix on Monday night.
It's been eight years since the company first started promising to revolutionize pretty much everything. But it missed deadline after deadline, in large part because it can't make a critical component of its system work.
Well, we can exclusively reveal, that, yeah, it still hasn't fixed it. So if you want to fork over $2,295 for something that will, most likely, become obsolete almost as soon as you put it on, then go ahead. But don't say you haven't been warned.
Here's the truth: the Magic Leap headset is actually quite nice. It is not like the big, bulky VR headsets that we occasionally try on, it's pretty light. And it fits comfortably, although you have to wear it at a weird angle, so it feels a little like when your glasses droop off your nose.
Talking of glasses, it won't work with them. But Magic Leap has worked around that by producing a series of lens inserts in a range of strengths. Except the one that worked best for this reporter's eyesight didn't fit. Why? Because it was designed for the "1" headset and not the "2" headset.
What does that mean? We asked the guy doing the demos – Magic Leap actually has two sizes of its headset – to accommodate for different sized heads (different sized spaces between people's eyes). Someone had only shipped the size 1 lens additions and I needed a size 2 headset and lens. If you think that this is all a bit fussy, that's because it is.
But aside from the fact that, as a result, everything projected at more than three or four feet away was a bit blurry, there are other issues. The field of view of really quite narrow, so you frequently feel the need to move your head up and down – something that gets tiresome after a while.
Track and field
The biggest issue though is what we have long warned about with Magic Leap: tracking is off, and slow. Because the Magic Leap One is an augmented reality (AR) headset, you can see through the goggles to the real world and images are presented as holograms in front of you.
This is actually rather pleasant compared to the disorientation you experience in a closed-off VR headset, like the Oculus, but it is darker than we wanted. If feels a little like you are walking around in a room just after sunset; you persistently want to turn a light on.
Anyway, because you can see through to the real world, we were able to see the problem with tracking first hand with the little controller that goes with the system. What was cool, and in fact, rather lovely was that an exact image of the controller appeared in light through the headset and moves as you move it in your hand.
You can compare where the virtual controller is compared to where you are actually holding it. If you look at both, the tracking becomes immediately apparent: when you move the controller, the virtual image lags behind it. The faster and further you move it, the longer it takes to catch up.
AR upstart Magic Leap reveals majorly late tech specs' tech specsREAD MORE
This isn't really a problem in an experiential sense, it just makes you feel a little like you're in a laggy computer simulation. But it is a huge problem if Magic Leap wants to achieve its main goal – to play games over the headset.
There were three demos to play with. Notably, each one was on a different headset and you had to put on the right headset to view it, suggesting that there is some delay in going between applications. But more importantly, each demo was static.
We tried two: one lets you see a human skeleton and select and move parts of it around. It's easy to see why people get excited about Magic Leap and keep piling money into the company – the images are like Star Wars holograms. Solid, fun.
Another app was a rendition of a virtual building which, again, was quite engrossing because you were able to walk through it – architects would love it and the company rep said the system will work with most CAD files, so it is all too possible to imagine people testing out and walking around large-scale architectural designs, or even digitally rendered furniture.
But they were static – nothing moved in the virtual world unless you decided to move it. Which draws some pretty strong limitations around the headset.
We wanted to test another key constraint – that the further away an object is, the harder it is to be accurately rendered due to the way Magic Leap's system works. But it was hard to do so because, thanks to not having the right size lenses for the right size headset, it was already a little blurry.
In short, the experience was exactly what we expected. And in answer to each of our questions about the limitations, the answer was, of course, it will be fixed in the next version of the headset.
It is notable that even before the Magic Leap One was released in strictly limited quantities to specific people who, laughably, were obliged to place them in a safe each night, CEO Rony Abovitz was already fund-raising for the second version.
It may happen. Magic Leap may get enough financial backing to finally do what it hasn't been able to achieve for eight years and fix the tracking issue while expanding the field of view. Maybe. But we're not holding our breath.
Talking of AR headset and blurriness, there are another two companies at Intel Capital doing interesting things in this field.