Excerpt from Page 52 of the print edition.
What many e-mailers and computer game devotees ultimately want, of course, are not smile symbols, but avatars, digital stand-ins, preferably with their own smiling faces. Mova, a California start-up, has developed technology to capture minutely detailed digital images of people's faces in motion. Instead of filming some 150 markers glued to the face, the conventional motion-capture method, Mova uses phosphorescent makeup to record at least 100,000 points. It can also manipulate the resulting image like a puppet, to express almost any emotion. Thought the company is just introducing the method to Hollywood, the goal is to make it cheap enough for home use. But the Clippy problem -- getting the facial expression to fit the context -- applies to avatars as well. Mova founder Steve Perlman says it's like getting speech recognition technology to distinguish between "how to recognize speech" and "how to wreck a nice beach," only harder, because emotional expressions are so nuanced.