Why body language holds the key to virtual reality
A few weeks after F8, I chatted with Eric Romo, the founder and CEO of AltspaceVR. I was in New York wearing an HTC Vive headset and holding a pair of motion controllers. Romo was in California doing the same. A pair of laser-emitting boxes affixed to the walls mapped the rooms we were in and tracked our motion, allowing Romo and I to walk, wave, and interact as if we were in the same space, and not 3,000 miles apart. If we got too close to one another in the virtual space, I felt legitimately uncomfortable. It was easier to avoid talking over one another, because hands gestures and head position provided conversational cues.
Compared with Facebook’s avatars, our body language was limited. We were like children’s action figures, given just enough mobility to spark the imagination. Our bodies had arms and hands, but they stayed locked at our sides. Our actual hand position was represented by a pair of controllers floating in front of our avatars. We had legs, but we couldn’t walk around. Instead we used a cursor to teleport from one position to another. Our heads moved, but our gaze did not. Our mouths didn’t attempt to track the sounds of our words.
As more accurate tracking technology becomes available, AltspaceVR plans to add it in. Right now, for example, anyone with a Leap Motion controller can have highly detailed hands, complete with flexible fingers, in place of two floating game controllers. When comedians perform for big crowds in AltspaceVR, they can wear a full body suit, allowing them to act out physical gags.
The first ground rule of avatar design for Oculus is simple. "Leave it to the human brain, don’t fake it."
Of course, with convincing body language now a matter of computer code, the person across from you may not be what they seem. Imagine you’re in a boring meeting with a lot of people: in real life it would be rude to walk out the door. In VR you can sneak away, while your avatar stays sitting, nodding intelligently, and paying attention. "That autopilot version of a person I’m sure everyone would love to have," says Sheikh.
After my visit to the Panoptic Studio, I wanted to try the cutting edge of social VR for myself. So a few months later I traveled to Facebook’s headquarters in Menlo Park, where I got a demo of its Toybox software. Inside a windowless room optimized for virtual reality, I donned the Oculus headset and an assistant placed the Touch controllers in my hand. Now I was across a table from my partner, a pair of floating blue hands and a head.
In Toybox, I had the visceral sensation of being with another human being, but the world had none of the constraints or consequences you find in your living room. We could smash a vase with our slingshot, then snap a finger and it would be back. At one point we lowered the effects of gravity, and sent our pile of blocks floating in every direction. They scattered to the four corners of our playspace, and a few minutes later, reappeared neatly where they had begun.
See the full story here: http://www.theverge.com/2016/12/7/13857144/social-vr-carnegie-mellon-panoptic-studio-facebook-oculus-toybox