We tested the Magic Leap for months: This is what we think

We tested the Magic Leap for months: This is what we think

Is the Magic Leap ready for industry?

While this may seem like a trick question to some, it was front of mind for the RE’FLEKT team during the unboxing of our very own Magic Leap. We’ve read the reviews and seen the tests but how would it really match up to the current seasoned veteran in the Mixed Reality market - the Microsoft's HoloLens?

After months of testing and application with real use cases, here are our first thoughts.

 magic_leap_box

Diffraction gratings, external cables, and tracking

One of the most notable differences between the Magic Leap and the HoloLens is the diffraction waveguide - or for the layman, the layers of glass that block out natural light. Since both devices are constructed differently, there are slight differences in the way light is perceived/not perceived. In comparison, the HoloLens display produces clearer images as there is less "stray light" from the excess light in the system (the light that is "discarded" by the diffraction grating).

However, the brightness also plays a role here. Karl Guttag states that the HoloLens blocks 60% of incoming light while the Magic Leap blocks a whopping 85% (so pretty much all the real-world light).

While this is fantastic for allowing the holograms to really pop out and reduce other optical problems on the Magic Leap, it is, unfortunately, let down with more pronounced ‘lens blurs’. This is a little ironic as the darkened waveguides (which are supposed to control the light and make everything darker) end up creating pockets for light at awkward angles to get captured in and create lens blurs. This is especially noticeable when a light source is either directly in front or above the device.

What does this mean for industrial usage? Well, for areas with stable light conditions such as medical environments or operating theatres, where light is always shone directly onto the subject, this should not be an issue. However, for environments with changing light conditions could experience low visibility as light conditions change - such as oil rigs or factories where external lights are ever present in the background.

The bottom line here: the Magic Leap clearly has higher Field of View (FoV) but lower image quality than the HoloLens. Also, the light blocking is quite intense in the Magic Leap - making it hard to recommend for physical tasks using a lot of interaction with the environment. On the other hand, we see clear benefits in a more VR-like entertainment scenario.

More Power but More Cables

Another major difference between the two devices is the Magic Leaps Lightpack - a tethered puck shaped pack that attaches to the hip. Removing most of the processing power from the headset not only allows for more but dramatically reduces the headset weight by 300 g (compared to the HoloLens) leading to a much more comfortable experience and increasing the amount of time the device can be worn. The big downside is the external cable that could pose as a potential snag hazard in industrial environments (not so much of an issue for entertainment use cases). What’s more, the cable is built into the device, so if a snag was to occur it would not simply detach but pull the entire device and wearer.

REFLEKT Dev team

All this extra computing power has a second downside. Due to the battery size and all the chipsets being located into its puck design, the Magic Leap requires fan cooling which may pose as a potential problem for industrial environments requiring a sterile environment as this must be maintained - compared to the passive cooling of the HoloLens.

Object Tracking

Despite boasting a design that encourages more GPU performance on the Magic Leap, our initial test still found the HoloLens to outperform with object tracking.

Overall, the Magic Leap digital content appeared to be a little jittery when seen up close or while moving around objects. That said, this is still early stages for the Magic Leap and we expect the tracking to be improved over time.

While not ideal, our experience with real-world use cases in industrial environments we have noticed that most maintenance procedures do not require too much movement around the object being serviced. For example, repair procedures could require you to get pretty close to an object or get inside a car. For this, you need large tracking FoV and robust tracking in general. Some of these cases are still challenging for the HoloLens, but would be even more of a challenge on the Magic Leap.

Display technology

The display is what everyone expected Magic Leap to improve, but it feels as if they only built a slightly more ‘up-to-date’ version of the HoloLens waveguide display with some extra layers. This was expected since the HoloLens is already a couple of years old, but the release of the second HoloLens (code named Sydney) should close or surpass this gap in 2019.

Virtual images are clear but lack the ‘Magic’ expected. We can’t wait for its light field technology to be implemented so we can experience the ‘real’ Magic Leap.

Ecosystem

The upper hand in ecosystems goes straight to the HoloLens. As a service company, Microsoft is able to throw all of their expertise behind it, from development aspects to the add-on features. It is clear that Magic Leap really had to reinvent the wheel and start from scratch for all applications, software and everything else around it (including the device itself). They did a great job, but if we compare to Microsoft, who already have a massive ecosystem of applications, software, and hardware, it was more a case of UI, software and service integration (and headset design) rather than starting from zero.

This is most apparent with the Magic Leap”s UI concepts - a wild mix of gestures and touch input. We have no doubt it will be improved over time.

Hololens (1)Image source: Microsoft

Overall

While the Magic Leap has a stronger consumer focus than that of the HoloLens, especially for the entertainment industry, there are a number of challenges the Magic Leap will have to overcome to truly be industry ready. While the overall package is really not that bad, expectation management from huge pre-release campaigns really led to considerable disappointment upon release. However, the elephant in the room is the missing light field technology. We can't help but theorize that shareholder pressure led to the release of a HoloLens like device to simply satisfy the market while development of the ‘real’ Magic Leap continued. If this is the case we are truly excited to see if the Magic Leap will be fast enough to keep up with Microsoft’s strong lead in the market.

Want to know more? Check out this great review from the Verges, Adi Robertson, which also gives a very accurate view on what we also experienced.