Framestore's Charles Howell on the Visual Effects in Gravity
Framestore is a London visual effects company that recently won an Oscar for its work on Gravity. Charles Howell, head of production, explains why moviemaking is shifting away from the auteur model.
A film’s look has traditionally been the result of two visionaries—the director and the cinematographer. Has the rise of visual effects changed that?
Moviemaking is getting more collaborative. Alfonso [Cuarón], the director of Gravity, always made a point of saying how involved Framestore and the visual effects were and what a huge collaboration the film was. Some directors don’t necessarily have the confidence or humility to acknowledge that. If it’s a really heavy visual effects movie, like, say, Lord of the Rings, you could have more than 900 visual effects artists.
Gravity’s director of photography, Emmanuel Lubezki, won the Oscar for cinematography, though much of the film was computer generated. How’d that make your team feel?
Gravity’s a unique case, because it’s a hybrid of live action and computer-generated images. And Lubezki was very heavily involved in our work. He came and lived with us for a bit, and lit the entire movie on a computer screen with us. So in that case, I think his win was justified.
Is the public aware of your company’s role?
I think more so recently, and that’s mainly due to a number of landmark projects that made people sit up and go, “Oh, OK, a bunch of guys did that on a computer.” Films like Forrest Gump, which did cool things, like putting Tom Hanks into footage from old newsreels. Before that was Jurassic Park, when everyone saw those dinosaurs and thought, “Wow, how’d they do that?” And Avatar was obviously a big visual effects movie, and Gravity, too. There’s still a way to go.
You’re a producer, not a visual effects artist. What are the big challenges?
I work with animators, technical directors, compositors, and paint artists, among others. They’re usually very creative people, and some are very technical, so communication with them can be difficult.
When will actors become obsolete?
Computer-generated humans are the last big challenge in visual effects. We’ve done everything else—water, fire, cloth, hair—all these natural phenomena. But there still hasn’t been a convincing computer-generated human. You can animate a human’s face, but there are so many controls on it, and the nuances are hard to capture. That along with the look of skin and the way it reflects light—there’s just so much detail. Movies have tried, like the film Beowulf. The characters sort of looked like real people, and we describe that as the uncanny valley—they kind of look good, but it’s not quite right.
How close are you to nailing it?
It’s not going to happen overnight. Maybe in 5 or 10 years you’ll see a performance by a computer-generated human that you won’t question.