Seeing the Oculus Rift at E3 feels like the end of a journey. It’s experienced a more dramatic trajectory than perhaps anything else at the show, from a simple prototype in 2012 to the flagship product of a company that Facebook paid $2 billion for. Oculus’ booth is lined with lavish prints of virtual reality games like Lucky’s Tale and EVE Valkyrie, and co-founder Palmer Luckey is chatting with journalists under the calming purple light. But really, it’s a beginning: our first look at one of the earliest attempts to make VR a real medium instead of a science fictional dream, complete with a totally new control system.
The finished Rift, which will see release in early 2016, is a surprisingly low-key device. It's big, but its size is tempered by gentle curves and a matte finish — plastic on the front, cloth almost everywhere else. It weighs about as much as Crescent Bay, the prototype it's based on, but notably less than either the second development kit or the mobile Gear VR (which makes sense, given that the Gear VR has to hold an entire phone instead of just a screen). Oculus has deemphasized specs ever since announcing Crescent Bay, but Oculus CEO Brendan Iribe confirms that it's got a wider field of view than the DK2, and its screen is high-resolution enough that the pixels aren't distracting.
That's not to say the screen is perfect — it's still probably the Rift's weakest feature. Oculus has overcome the "screen door effect" of the original headset, but when every flat screen around it is dense with pixels, the magnified Rift image looks noticeably less sharp. It was probably a little blurrier than usual in my case, because the demo unit had a fixed inter-pupillary distance — normally, there's a slider that lets you adjust the width of the lenses to fit your eyes, but it was sealed off for E3.
I got a choice of eight Rift game demos, of which I ended up checking out three: hack-n-slash RPG Chronos, a hockey game from a sports collection, and chilly survival game Edge of Nowhere. It's frankly amazing how well the developers have managed to overcome VR motion sickness. The slightly counterintuitive trick, as far as I understand, is to stop the camera from shifting in any way that might remind you of natural motion. In Chronos, for example, each room is shot from a fixed angle, like you're looking onto a stage — it's a strategy that has worked for older VR games like Herobound. The hockey game uses a first-person perspective, but you only play as the stationary goalie.
Is it still immersive if you're a disembodied camera?
Edge of Nowhere is particularly interesting. The game is exactly the kind of thing that could be nauseating in the Rift: an action game with lots of running and jumping (who knows, the Lovecraftian elements could end up being nauseating in their own right). But instead of the average over-the-shoulder perspective, its third-person camera is kept as far from the character as possible, and you don't get any say in how it moves. The effect is almost more like watching an interactive movie than playing a game. Occasionally, this makes it difficult to control — it's hard to time jumps if you're seeing them from an angle you're not used to, and it doesn't directly track the protagonist as much as slide around him. But it's mostly a problem of unfamiliarity — something that, if VR expands, could quickly fade away.
Since VR is often described as "putting you in the game," this might seem to defeat the purpose. What’s the point of playing something immersive if you’re immersing yourself as a disembodied camera? But besides getting what feels like the world’s biggest screen, seeing a third-person game in VR gives you a real sense of situational awareness. You know things your character doesn’t, and you can respond accordingly. It bolsters the argument that VR is genuinely a new medium, with conventions that are only just becoming clear.
On launch, it's not clear how many games there will actually be, and Iribe didn't clarify except to say that these might not all be out when the Rift is released. Eight games isn't enough to make a catalog, but we're seeing plenty of VR games — Adrift and Time Machine, to name two — that could help pad it out.
As weird as it sounds now, the consumer Rift isn't actually the most exciting part of Oculus' VR plans any more. That would go to Oculus Touch, the prototype motion controllers that are coming out in the first half of 2016 (the Rift comes out in the first quarter). Touch seems roughly in the state that Crescent Bay was before today, dotted with exposed sensors and apparently slightly fragile; it's even paired with a Crescent Bay headset. While most of the controls are solid, the face buttons are cringe-inducingly mushy. But it otherwise feels admirably non-experimental. Aside from the rings around the front — which lets the Rift's camera track them from multiple angles — the controllers are like nothing so much as Wii nunchucks, designed to fit the contours of your closed fist. There are two triggers, one under your forefinger and the other against the meat of your palm, and an analog stick for your thumb.
Touch feels less like a remote control than an extension of your hand
There’s also one intriguing, invisible feature: a pair of touch sensors on one trigger and the analog stick. They're not controls per se, but they let the Touch reflect your hand position — if you lift your thumb, for example, your avatar will raise theirs as well. It's so natural that it feels weird at first.
There are no Touch-enabled games at E3, but Oculus has developed a "toybox" full of moveable, breakable objects. I built block towers and managed a short, clumsy game of ping-pong. I played tetherball and, in a strangely meta moment, picked up a controller inside the game to drive a miniature tank. I fired a slingshot and managed to actually hit a few things. It feels roughly as accurate as, say, the Sixense Stem, but its design is far more intuitive and restrained, with a minimum of buttons. And while Touch doesn't give you as many options as Valve's trackpad-equipped controllers, it consequently feels less like a remote control and more like an extension of your hand.
Unfortunately, you’re still limited to a smaller space than Valve’s room-sized VR. Oculus added an extra tracking camera to the setup when we tested its motion controllers, so it’s possible you’ll end up needing more than just a standard Oculus Rift kit and the Touch controllers. A single camera theoretically wouldn’t be able to pick up, say, someone turning around and obscuring the controller — something that both Valve and Sixense have solutions for.
Motion wands are emerging as a standard of next-generation VR; Sony uses the Move, Valve has its controllers, and now Oculus has Touch. Iribe won't say that's the ultimate standard, but he comes close to dismissing one of the major competing trends: pure hand tracking like the Leap Motion or HoloLens' controls.
Oculus doesn't seem very enthusiastic about HoloLens-style hand tracking
When I ask whether Oculus is exploring motion tracking, he says you'll probably always want something physical to interact with — just waving your hand in the air won't give you the feedback you need. I'm inclined to agree, mostly because even the best motion tracking (and Leap Motion is quite good) involves a lot of guesswork. It's hard for our hands to make perfectly consistent gestures, and especially as Oculus branches into more demanding games, an input method that can fail isn't a valid input method at all. A Touch controller lost track of one of my hands briefly during the demo, but it was otherwise very consistent, especially for a prototype.
Even if Oculus hasn't given us a killer app, it's showing off a lot of very solid building blocks — though they're ones that won't all be in place when the Rift makes its debut. I've never been convinced that VR is The Future, singular, of gaming. But Oculus is making a good case for it as a future — something people might buy and enjoy as more than a curiosity.