Doug James is making virtual reality more real.

Creating realistic sound in a virtual environment requires much more than simply playing a recording. Take a bell for instance. The sound generated when it's rung depends not only on how hard you pull the rope, but a host of other factors, including how the sound waves radiate from it, your position in relation to the bell, and even the shape of your head.

“If you ignore that kind of stuff, it just doesn’t sound real,” says James.

Even with the fastest computers, computing the physics involved takes a lot of time—and that’s just for sound. Creating a virtual world might also require calculating how to graphically depict the bell as it swings and the tug of the rope as you pull it. But unlike other engineering calculations that can be computed at leisure, these calculations are time critical, because even the most realistic bell tone won’t fool anybody if it comes too late.

“We know how to simulate most everyday physical systems, it’s just that it can be really slow,” says James. “Computers are only now approaching the speeds necessary for processing at interactive rates. Unfortunately, many phenomena are still many orders of magnitude too slow for interactive processing.”

So James looks for ways to make physics faster. “We’re focusing on how to make new physical algorithms that are better suited to time-critical and human-centered applications,” he says. “One long-term goal is providing interactive-physics applications with multi-sensory feedback, such as sound and haptic (touch) feedback in addition to traditional graphics.”

By devising algorithms that pre-process many of the necessary calculations and focus only on what actually changes, James can compute virtual physics several thousand times faster than traditional technology. “If you have very little time to solve a complex equation it pays to do things ahead of time,” he says. “We are figuring out how to do things in a completely different way that may seem crazy at first, but it can save us a lot of time in the future.”

This is good news not just for video games and computer animation, but also for applications like virtual surgery, prototyping new products, and assembly planning—James’ research supporters include Pixar and Boeing, as well as the NIH and NSF. “If we make models more realistic and/or real-time, we can make applications more compelling and more useful,” says James. “With better sensory feedback people can perform tasks more effectively in virtual environments.”

The most rewarding and exciting aspect of his work, says James, is coming up with ideas and applications that no one has ever thought of before. “The great thing about the stuff we’re doing now is that it’s a fantastic license for creativity,” says James. “You can take all this crazy math, physics, and algorithms and combine it creatively in new applications. It’s really a lot of fun.”