Magic Leap isn’t having a great time of things. The company’s actual tech demos have been panned, its hardware hasn’t reviewed particularly well, and now Palmer Luckey, the onetime founder of Oculus, has weighed in with his own thoughts on the platform, hardware, and capabilities. His conclusion is that the Magic Leap is more of a “Tragic Heap” to use his phrasing.
Luckey’s review covers several points I haven’t seen widely discussed in other ML coverage, including the shortcomings of the company’s decision to use magnetic tracking rather than line-of-sight or positional reporting. The controller doesn’t review well (it lacks a clickable trackpad and responds poorly to any kind of quick motion). The Lightpack computer, containing an integrated Nvidia Tegra X2 (Parker) SoC is praised for its overall performance — this seems to be the part of the ML that everyone likes the most, including the fact that the heat-generating and heaviest part of the system is sensibly clipped to one’s belt rather than hanging off your face.
Having only previously been familiar with Palmer via his public statements, I’ve got to say, the man has a talent for the not-so-subtly delivered insult. After noting that it’s a shame the ML battery isn’t replaceable, he breezily dismisses the point with “nobody is going to use their ML1 long enough for that to matter to anyone but collectors with an aim to preserve the history of AR and VR.” Granted, the Oculus Rift’s sales figures haven’t exactly set the world on fire — the PSVR is believed to command the vast majority of the VR market that requires a standalone computer or console — but compared to the Rift, the Leap seems to offer a weak value.
Luckey unloads on Magic Leap for its headset (Lightwear) design, noting the vast gap between the company’s rhetorical claims and the practical reality of its shipping product. While the Magic Leap does use a set of dual wave guides to give it two planes of focus rather than the single plane implemented by Oculus and HTC, its design falls vastly short of what the company claimed it would offer. Here’s Palmer:
The supposed “Photonic Lightfield Chips” are just waveguides paired with reflective sequential-color LCOS displays and LED illumination, the same technology everyone else has been using for years, including Microsoft in their last-gen HoloLens. The ML1 is a not a “lightfield projector” or display by any broadly accepted definition, and as a Bi-Focal Display, only solves vergence-accommodation conflict in contrived demos that put all UI and environmental elements at one of two focus planes. Mismatch occurs at all other depths. In much the same way, a broken clock displays the correct time twice a day.
He then goes into more detail on how stacking wave guides to create more focal planes is unlikely to scale well based on frame rate, image quality, and weight concerns. Headset tracking is also mediocre, the FOV is still quite limited, and the operating system is, to use his words, “Android with custom stuff on top,” rather than the unique, ML-specific OS that Magic Leap claims to have built.
All of these issues and flaws would be understandable in a first-generation product from a small Kickstarter team or similar research endeavor from a major company. What makes Magic Leap baffling is that this represents years of work from a dedicated team that received over $ 2.3B in investor funding. The only magic in Magic Leap seems to be how they pulled that off.