After roughly three years of personally anticipating the release of Magic Leap’s debut product, The Magic Leap One is finally in my home. Going against the logic of many critics the device is actually not vaporware, but a tangible (albeit expensive) developer kit. Since its debut I’ve read a lot of mixed reactions; some of it anger, some of it confusion, some of it delight, and some of it simple excitement for a product that represents a vision and a tool necessary to make something very unique happen. As a longtime follower of Magic Leap, my own feelings over the last few years have spread across a wide spectrum as well, so I can understand how the device to newcomers must be conflicting. With this review I hope to strike a middle ground between the the criticism and the hype, the sweet spot if you will. So I hope this helps inform the curious where the Magic Leap One excels and where it actually falters. As much as I thought I knew about the device beforehand, even I was surprised!
THINK INSIDE THE BOX
Images appear inside the square within the lens.
Like most people, I had a custom fitting from a representative of the third party company Enjoy. While not entirely necessary, I’m glad Magic Leap is doing it this way. One of the things I’ve noticed in the three weeks of owning a ML1 is that too many people are wearing it incorrectly; something I’m sure that drives the human factors people at Magic Leap batty. The headset does need to be worn at a downward facing angle with the back of it perched almost on top of the crown of your head. Not doing so is not only uncomfortable, but also messes with the FOV (raising or lowering the position of the lens can shorten or lengthen the vertical view) and may even affect the eye tracking. After I had placed the headset on with the correct nose and forehead pieces, I booted up and saw what has been described as the “out of box experience”: an astronaut “leaping” between alien islands floating in the air. There were two things I immediately said to the Enjoy person. The first was my surprise at how light the lens tint was. She responded that everyone says that to her. From the outside the reflective lenses look impenetrable. You can’t see someone’s eyes behind them and so you expect not to be able to see that well out of them. However, from the other side things are remarkably bright and something I would compare to wearing a very lightly tinted pair of sunglasses. The tint is not something I regularly notice unless the room is for some reason not well lit. It’s also important to note that the Magic Leap One does not operate well in dimly lit rooms, contrary to early theories. This practically makes most concern about the dimness of the lenses moot because you’re never going to be able to use it in a room that doesn’t reach a certain level of brightness in order for the sensors to detect surfaces.
FIELD OF VIEW
Surprisingly, the 50 degrees FOV hardly ever bothers me.
The second thing I said was that the FOV was much bigger than I’d anticipated. This is a major sticking point for pretty much every AR/VR device. When the Rift and Vive launched there was a combination of disappointment and mudslinging over which headset had worse FOV, with the Rift on the losing end at 100 degrees versus the Vive’s 110. Standalone AR headsets like the Hololens, ODG range of glasses and Magic Leap have it even harder as the current technology limits this to sub-50 degree numbers. I owned a Hololens for several months and the FOV was my number one complaint. How I wished for just a few degrees more. The 35 (diagonal) degree FOV meant that I would have to keep perfectly still to look at anything virtual like a floating screen or menu. When it came to full sized humans avatars like in Fragments, they were always cut off unless I stood on the other side of my living room. Thus, I had extremely low expectations for Magic Leap One. But because of several factors (Like a UI smartly designed to accommodate the limitations, and a substantial increase over the Hololens) it is actually at the bottom of my complaints about the device. In fact, I had long thought that consumer AR would likely need to be somewhere around 90 degrees or more to be acceptable, but now I think even a marginal increase over 50 could be the trick. Even as is, I was able to replicate my 150 inch home projector. The FOV of the One truly impressed me, and while it is still limiting in certain use cases, I hardly ever think about it.
I’ll try not to retread topics that have already been exhausted. Yes, it’s lightweight, much more so than the Hololens which was a burden to wear for more than 45 minutes, making it a chore to watch a movie in. And the sound design is a stellar feature in its own right. “Ear candy” is a good way to describe it. The Lightpack is sturdy and heavier than I thought it would be. The cable sometimes needs to be moved out of the way, especially when sitting, but I have yet to snag it on anything. Overall, the ergonomics are a win. Even the claims of “tunnel vision” were overblown. The sides of the device hardly block your periphery as there is considerable gapping between the headset and the face (though this will vary depending on the person’s face). It had ben speculated that the One would be dangerous in certain work conditions if used for enterprise due to its lack of peripheral vision and unwieldy cable. Both of these are non-issues once you realize that the sides of the headset sit above your eye, not directly next to it, and that the cable can be covered with a lab coat or even just a jacket or hoodie, depending on the scenario. I’ve already spoken to some enterprise devs working with the Magic Leap One who are happy with the device.
Something that was definitely an oversight is the lack of a carrying case. The Hololens came with one, as do most portable VR headsets. Luckily, these are easy enough to come by for cheap. I picked one up off Amazon for twenty bucks. And despite rumors (again) that you would need an entire backpack-sized carrying case because of the multiple components, my Magic Leap One surprisingly fits into one as small as a lunchbox.
Shot through the lens
I’m not sure what to call the images one sees in a Magic Leap device, as apparently they don’t meet the criteria to be called holograms, so I’ll just refer to them as virtual images. There’s a few things to clear up on this topic. First, there is no set range for how transparent the images will be. Some people have said they look “ghostly”. And this has led to a lot of assumption that they’re always transparent. Even Magic Leap’s “Capture” algorithm approximates a constant transparency (something I have to imagine is them just trying to be a little too honest post-launch). But the truth is that the transparency is dependent on on the lighting of your room, the distance between you and the images, and the distance between the images and real world objects. Recently, Tested reviewed the One and stated that you will never see opaque images. So when I got mine I was surprised to see that they can in fact be, and often are opaque. Sometimes even to a fault. One of the first times I realized this was when trying the musical experience Tonandi: my hand disappeared when I reached into a virtual plant growing out of my floor. Because the system lacks hand occlusion this can actually be a bit disappointing. You don’t want to see your hand disappear behind an object it’s physically in front of. Of course most of the time, the opacity is more than welcome as you may see a turtle swimming across your living room and you don’t want to see your floors and walls through it.
Images can be opaque
Shot through the lens with an iPhone X.
I have yet to tell how effective the multiple focal planes are, which seems to be a consensus among owners. Most of the time when I notice it, it’s because something in the background looks low-res. Which isn’t great. It just makes me think the CPU is trying to keep a steady pace rather than make something look “out of focus”.
The bottom line, though, is that the images you see in the device are generally compelling not because of how real or fake they look, but because of what they’re doing. Tonandi is something I’ve now experienced half a dozen times. It’s wowed friends, and brought shivers down my spine. It honestly feels magical in ways that I’ve not felt in VR.
TRACKING AND MESHING
This is where my main gripes are. Though it’s not all negative, most of my wasted time troubleshooting with the device has been related to how it meshes my home. Basically, the device has to map your environment in order to know where the ‘grams, or virtual images, are in the physical world. And while I enjoy the mini game that Create has before each launch to map your space, it can get tedious. Especially just before writing this review I received several errors that required a reboot
Meshing is my biggest complaint with the device. The speed is great, and if you take your time, you can really get it to be finely detailed, which is important for environmental occlusion. But the inability to detect dark surfaces has crucial areas of my living room area unusable. I have a dining table that is an espresso black wood, and a console table that is a matte black metal. The TV is also a no-go. Those areas of my living and dining room are complete failures when it comes to meshing, which is very disappointing. It means I have to avoid them, which adds more conscious thought about my surroundings than should be required when engaging in the experiences.
The good thing is, these are issues that should be improved via software updates. The best aspect of the tracking system, in my opinion, is the hand tracking. This is something that really sets the device apart. It’s what makes Tonandi so special. I was initially bummed when I first discovered that Magic Leap One didn’t have the same fidelity of hand tracking as Leap Motion. It’s limited, but like the FOV, not so limited that you feel it’s a detriment. Hand tracking is actually a huge plus and maybe the biggest surprise to me. In Tonandi, you see glittering ornaments dangle from your ceiling like tinsel on a Christmas tree. Being able to touch this tinsel and brush it aside, causing an angelic sound to ring out, is more enjoyable than it should be. A friend who tried Tonandi remarked that he felt like a wizard as light beams poured from his fingers. And touching an undulating mass of digital water that reacts like a living organism to your fingers awakened all sorts of ideas about music composition in AR. It’s the feeling of transcending the digital world that makes this particular aspect such a compelling feature of the device.
Helios is Magic Leap’s web browser and is currently my least favorite aspect of the Magic Leap experience, so I’ve barely used it. I see the huge potential, especially once resolution is increased. But the execution here feels cluttered and slow. Using the trackpad to type URL’s or select recommendations from a list of websites requires multiple attempts. And I’m never quite sure what button will do what if I want to “go back” or pull up other options. It’s constantly a guessing game. This kind of stuff I’m confident will be refined over the next year. But the low resolution currently makes it difficult to read text on. Another gripe here is that I was unable to play anything from Netflix or Hulu. If I want to stream media the only major website that works is Youtube. My hope is that we hear of some partnership between Magic Leap and an Amazon, Hulu or Netflix this year. Just imagine the convenience of having your own movie screen projected on a blank wall while your stuck in a three hour layover at the airport. While this is already something VR headsets like the Oculus GO can similarly do, they completely restrict your view of the real world. Here, you could watch Game of Thrones on a private movie screen embedded in the airport terminal and not miss your flight— if they fix the above issues.
Overall, the device is a fantastic piece of kit, with some room for improvement. The ergonomics make it very comfortable to use for long periods, the tracking quite solid (with occasional hiccups), and the virtual images often appear very “solid”. Even the Control, which Palmer Luckey dissed for using electromagnetic sensing, has been extremely useful. Though that could be because there isn’t yet enough software to form a proper pool for testing. It could turn out to be too limited in certain applications that I just haven’t tried yet. But these are all building blocks of course, and Magic Leap One already has the front facing cameras to switch to optically tracking the controller if it wants in the consumer version. Personally, the reason this device excites me is that the company is thinking outside the box. By not delivering just a headset that could be tethered to a PC, and by announcing dozens of content partnerships right out the gate on top of a yet-to-be-realized distribution deal with AT&T, they are circumventing a lot of the pitfalls of other startups. I’ve said it before, but I think Magic Leap is gunning to be the Oculus of AR. Their first developer conference is in a matter of days. I’m excited to see what their roadmap shows, because even just the small handful of applications that delivered with the One are compelling enough. Having unique hardware such as this, coupled with strong partnerships, and a base of excited developers is a rich cocktail. I’ll be back after L.E.A.P.!