Virtual reality is broken. Oh, the feeling of VR can be amazing. But no normal human being can be expected to rearrange their furniture, and connect themselves to a surround-sound-system’s worth of wires, just for the privilege of painting in midair.
That's why Daydream VR, which Google announced at its annual developer conference yesterday, is so promising. By partnering with manufacturers to build Android phones to a common baseline that can support high fidelity VR, Google was able to design a minimal headset and controller that could serve as a standard to deliver quality VR—without all the cords or gaming PCs.
"Our goal is to make every Android phone VR capable," says Clay Bavor, VP of Google VR. But what might be more important than VR capability is VR usability. Google’s goal is to make the entire user experience of VR worthy of your attention and soft to the touch.
To be fair, all Android phones can already run VR through Google’s first VR project, Google Cardboard. These paper headsets, fitted with plastic lenses, fit on almost any Android phone to let you look around inside games and 360-degree videos. The problem is that—not to be rude—Cardboard just isn’t that good.
"We knew there was a ceiling on the level of immersion you’d be able to achieve with, let’s face it, some cardboard, and phones designed only to be phones," Bavor says, and he’s right. Cardboard is amusing for only a few minutes at a time, and anyone who’s tried a dedicated VR system like the Oculus Rift would see Cardboard as little more than a novelty.
What was so compelling about Cardboard was its price. While an Oculus Rift is an investment of $600 or more, Cardboard is sold for $10 or under by a variety of companies that print headsets from Google’s open-source spec. This low commitment level allowed organizations like The New York Times to mail Cardboard to its own subscribers; Google had shipped 5 million of the headsets as of earlier this year. "But in the end, we’re VR purists who want to build the best thing, and VR experiences that are not only mobile, approachable, and affordable, but super high-quality, and richly interactive," Bavor says.
Boosting the power and performance of Android’s VR was an engineering feat. Better chipsets. Code that lets developers milk the processors for as much power as they need. Layers of hardware and software innovations that let Daydream VR do something unbelievable: the system has just a 20ms delay between processor and pixel. By comparison, Cardboard’s delay was 80ms, enough to feel atrocious. Samsung’s Gear VR is around 20ms, and it’s enough to feel real.
But it’s not the performance that Bavor is so excited about. "Approachable" and "accessible" are words that he says again and again during our interview. When it came to the industrial design of the Daydream headset, Google wanted it to have a domestic friendliness—something akin to a piece of clothing or home appliance—to make it something that you’d actually be drawn to touch and wear, compared to the black plasticky headsets of today.
"We’ve looked at how headsets are built and, this is an obvious statement, but you wear them on your head. They need to be comfortable and light, materials need to feel good," Bavor says, being careful not to reveal exactly what the headset is made of just yet. "We've been exploring using a pretty different set of materials—there’s no cardboard in it, or hard-molded plastic. We wondered if things like fabric and softer materials can provide a lighter, more comfortable experience."
One small creature comfort Bavor alluded to was that the strap could be stored inside the headset, eliminating the pile of scuba goggles feeling of a VR headset left on your table or in a bag. It's a small concern, but just the sort of thing that makes the experience of living with technology actually livable.
The other key component to approachability is what Bavor calls "easy in and easy out of VR." To enter Daydream VR, you slide your phone into the headset, and close it with a single latch. Put on the headset, and you’re suddenly standing in a tranquil forest. Though Bavor didn’t use the term, architects call this effect "liminal space." The forest serves as a transitional area between the hectic real world and the relaxing virtual one.
This forest is Daydream Home: the hub of your VR apps and content. Exactly how the system is organized is still unclear—a traditional menu screen can be pulled up by command, but that may not be the core UI. The virtual environment seems destined to be populated with real objects that serve as keys to sleeper functions—a similar approach to the old desktop metaphor in PCs that organized files inside folders to make computers approachable to luddites. Rather than organizing VR by apps, Bavor suggests that Home can surface the freshest VR content from across your VR library—be that new DLC levels to a game, or new chapters in an episodic movie—in a single place. "People ask, ‘Where do I want to go tonight?’ not ‘What application should I load and start my journey there?’" Bavor says.
Does this mean Home will leverage its rich VR environment to stretch beyond conventional UI? Could a bird fly through the sky when Angry Birds offers more VR levels, or could you see a new character from a game walking through the trees? It seems possible, but Bavor doesn’t clarify.
Controlling the mobile VR experience was the other great challenge Google solved for Daydream. Right now, the industry has no standards in this regard. Cardboard has a single button that can’t do much, while HTC Vive has a pair of motion controllers that require you to install laser tracking boxes. Oculus is working on motion controllers of its own, but in the meantime, the company shipped its headsets with an Xbox 360 gamepad instead.
To build the controller itself, Google went through as many as 100 different prototypes and physical mockups before deciding on the small motion control remote that, as Bavor puts it, "a normal person could immediately get and be comfortable with." Bavor rubs it almost compulsively during our interview, even gesturing with it to cement his points. He’s clearly used the remote so much that it’s become intrinsic to his day-to-day life.
The final design is no more intimidating than an Apple TV remote. It has a thumb trackpad and just two buttons. "We looked at in contrast to a gamepad with 15 buttons," Bavor says, "Compared to when when you’re in VR someone says push R2 and triangle, it’s super simple."
Inside Google’s base VR apps, it operates more like a laser pointer than a TV remote. "It kind of works the way we work. Little kids, one of the first things they learn is to point to things," Bavor says. "So it works like that, but it’s very flexible in what it can turn into." Aiming is a natural extension of the laser pointer metaphor: while in Street View, for example, you can point your way down the road, teleporting your way through cities. But there’s a whole lot more the controller can do, too.
Google already has a developer demo that features 16 unique functions that are programmed into the controller—they’re basically mini games, showing possibilities like simulated fishing and wood carving (Google will share the code behind these interactions for developers to use in their own apps).
Yet these represent a small piece of Google’s larger experimentation around how people will navigate experiences in VR. In fact, four designer/developers inside the Google VR team have devoted the last year prototyping a few VR experiences each week, figuring out what works, what doesn’t, and the best practices for the future.
"They’ve explored everything from what it’s like to assemble a puzzle in VR, to VR whiteboards, to architecting a house and stepping into it to dance and hang out," Bavor says. This four-person lab actually built a presentation simulator that allowed a speaker to practice a talk in front of an audience that could be scaled up or down in size, and whose demeanor could range from impressed or bored. Then they put the experience inside a digital rendition of the Shoreline Amphitheater so the Google VR team could practice their own I/O keynote in front of a test audience.
But truth be told, Google’s most important decision in building the Daydream controller goes beyond any of these possible use cases. It’s more important for the standard it represents. Google has mandated that Daydream headsets must be sold with the Daydream controller. The two devices cannot be unbundled.
That means Daydream developers will have the impetus to design controls to the same core spec—and given Google's reach, that standard could snowball, an app at a time, until any VR developer on any platform would want to include support for a Daydream-style remote.
Even if Google can standardize the headset and controller, it's discovered through testing that many experiences inside VR shouldn’t be presented the same way to all people. Think about it: In real life, some of us love roller coasters, others hate them. Some of us sit in the front row of the movie theater, others sit in the back.
"We’ve been doing a lot of testing," says Kurt Wilms, product manager on YouTube VR. "Obviously, people prefer different things."
In turn, Daydream VR will be careful to label and classify the intensity of content inside Google Play. So when you decide whether or not to download a new VR app, you'll be able to see some of the app's motion metadata—essentially categorizing how motion sick they could make you.
Meanwhile, inside YouTube VR, you’ll be able to very literally customize your theater experience for yourself, setting the screen to appear as if you’re 30 rows away, or as if you’re right up against it and need to pan your head to see the whole frame. It even sounds as if users will be able to decide whether they’re sitting in a simulated environment, like an old-timey movie theater, or floating in black ether with nothing but the movie.
On this topic, Wilms and Bavor seem to have entirely different preferences. Bavor prefers the virtual theater. Wilms scoffs a bit at the notion of fake seating, and leans toward the purer video display. The Android VR team’s own diverse tastes are proof of the need to offer users choice.
All of this makes Daydream seem like the perfect couch-level VR experience. There’s no strapping into some big machine. Instead, you can put on the headset, lean back, and watch cute kitten videos on a 300-foot screen. Daydream is to Oculus what tablets are to PCs.
However, the Google VR team has one groundbreaking technology that could let Daydream compete with the bigger VR immersion rigs, like the HTC Vive—and that's Project Tango.
Project Tango is a technology out of Google’s Advanced Technology and Projects (ATAP) group that debuted at I/O two years ago. Using a smartphone equipped with a depth-sensing camera, Tango allows you to walk around a house and turn all of its walls, hallways, and even staircases into a 3-D map. Having seen the technology work in real time, I can say it's every bit as impressive as it sounds.
It’s easy to imagine using a Tango-enabled Daydream headset that eliminates the necessity of installing the laser tracking boxes required by the HTC Vive's room-sized VR experiences. Instead, Daydream would always see and understand the spatial constraints of your environment, so it could shape the virtual experience to fit inside your home. Imagine sneaking your way through a virtual hedge maze that, thanks to a phone that can see that couch and ottoman in your living room, knows when to make you turn left so that you don't trip over your furniture.
When I ask Bavor whether Tango might come to Daydream, he nods to the inevitability, mentioning that the Tango team has been part of Google VR for a year now. "There’s a reason for that," he says. "For Tango, the technology is there, but there are many other things that need to come together with positional tracking. Once you can walk around, how do you keep people comfortable and safe? There’s a bunch of other things we’re thinking about."
In other words, just as Google redesigned VR to work on your phone, it’ll need to redesign it again—to make it work for the entire world around you.
All Images: via Google