Pixar’s Scientific Method

13 July, 2015 / Articles

Pixar movies, those computer-generated wonders known for their visual and emotional sophistication, are also science movies. They’re created with complex engineering, and they’re populated by heroes—whether toys, humans, monsters, or feelings—who are inventors, Rube Goldbergian geniuses, or, as in “Inside Out,” diligent functions of the human brain. Pixar’s characters often want to increase productivity or solve problems; the films often feel like industrious American projects combining gumption and physics. Think of Buzz Lightyear’s attempt to prove he can fly, in “Toy Story,” or Carl Fredricksen’s balloon-powered mission to save his house and travel south in “Up,” or the journey that Joy, Sadness, and Bing Bong take through Abstract Thought in “Inside Out.”

Last week, the Boston Museum of Science, in an effort to make children aware of the ideas that power the fun factory, opened an exhibit called “The Science Behind Pixar,” which it created with the studio: interactive models, videos, workstations, and displays, themed around the movies’ scenes and characters. Dizzyingly intricate scientific processes are explained alongside, say, a model of the superhero-costume designer Edna Mode from “The Incredibles,” or of WALL-E next to a Rubik’s cube.

I visited the exhibit on the day of the opening-night party, whose expected guests included the museum’s president, Ioannis Miaoulis; Ed Catmull, the co-founder of Pixar and the president of its and Walt Disney’s animation studios; and the actor John Ratzenberger, a.k.a. Hamm the Piggy Bank (“Toy Story”), a school of moonfish (“Finding Nemo”), Fritz the Mind Worker (“Inside Out”), and Cliff Claven (“Cheers”).

That afternoon, Miaoulis told me that the Pixar exhibit is part of an effort to “introduce engineering into young kids’ lives.” He added, with a slight frown, that the current U.S. children’s science curriculum is “pretty much all science of the natural world.” Miaoulis, an engineer, wants to change this. The first engineering exhibit that he brought to the museum was “Star Wars: Where Science Meets Imagination,” co-created with Lucasfilm, in 2005. It was a “blockbuster,” he said, seen by three million people, and now part of science programs at schools near you. After “Star Wars,” Miaoulis decided to take on computational thinking, with Pixar. The museum spent several years developing the show, testing its effectiveness on some three thousand museum guests in the process. At a museum, he said, “You have a lot of guinea pigs.”

“Star Wars” could inform people about space travel fairly naturally alongside its costumes and Millennium Falcon cockpit, but “The Science Behind Pixar” has to introduce and explain more theoretical concepts. It opens with an introductory video that shows you the Pixar studio itself. The voice of Mr. Ray, the schoolteacher manta from “Finding Nemo,” says, “I want all optical orbits up front. And remember, we keep our supraesophageal ganglion to ourselves!” and then we see a bespectacled Pixar employee riding her scooter through the offices, encountering friendly Ph.D.s who explain the company’s moviemaking process. Story is first; then there’s modelling, rigging, surfaces, sets and cameras, animation, simulation, lighting, and rendering. (Each step in the engineering process has a corresponding display.) At the end of the video, Catmull appears onscreen. “Computers don’t make movies—people do!” he says. “It is art and it’s technology. It’s hard and messy, but it is fun.”

Catmull, a bearded seventy-year-old genius with the air of a math-curious Bob Balaban, helped develop many of Pixar’s pioneering computer-animation technologies. I was there to walk through the exhibit with him; he was seeing it for the first time that day. We began in front of a huge model of Buzz Lightyear, astronaut toy hero. “I founded the company and did a lot of the original technical work,” Catmull said. “Initially there was no field, so we were bringing people out of academia. A lot of the original art and math and science came from the people who had their Ph.D.s in this area and were working on a set of problems. When I graduated, my goal was to make the first computer-animated film. I figured it would take ten years, because I knew there was a whole bunch of work in front of us; it took twenty.” Catmull graduated in 1974; he and his peers worked for Lucasfilm, Steve Jobs, and then Disney, developing technologies like RenderMan, a 3-D-rendering program, along the way. “Toy Story,” the first fully computer-animated movie, came out in 1995. Behind Catmull, the giant Buzz Lightyear looked gleeful.

Catmull grew up entranced by Disney movies like “Pinocchio” and “Peter Pan.” “There’s an interesting thing that almost nobody realizes,” he said. “When Walt Disney invented animation, it was based upon the brand-new technology of the day, which was filmmaking. They were coming up with new ways of using color and sound using the xerography process. The Xerox had just been invented. So he had a special room built by them where the room was the xerography machine, and there was a person inside of it with conveyor belts going in and out. Today people don’t think of that as technology anymore. They just think of the art. But part of the energy came from the fact that he was adopting all the new stuff that was coming. And when he died, it turned into a pure art form and it went downhill.” When Roy Disney, Jr., took the reins in the late eighties, Catmull said, “the one thing he insisted on was to bring technology in for reinvigoration.” Enter Pixar, and “Toy Story.” “So then we brought in 3-D animation, and of course it took off from there.”

At the Modeling station—character design, involving sketches and clay models—Catmull said, “One of the things I came up with in college was a way of working with non-four-sided things.” He held up his hands as an example—not four-sided. “I converted the four-sided stuff into basically what amounts to high-school geometry; once I had it in geometric terms I could basically make up rules for combining them. See, now I had made a new kind of surface.” At the Rigging station, Sully, the giant blue fuzzball from “Monsters, Inc.,” was lumbering along on a video screen, with yellow bone-like structures illuminated inside him, like an X-ray. “See those yellow things in there?” Catmull said. “That’s called rigging. In these characters there are literally thousands of things to control.” Bones, joints, muscles. “If you think about the face, there are muscles that get pulled around the corners of the mouth and eyes and so forth. There are so many that it’s almost not possible to control them all.” Rigging makes some of these things automatic—programming relationships between simulated muscles to make characters’ movements more logical and efficient. When you walk, he told me, “when your heel hits down there’s a shock wave that goes up your body. No people ever notice the shock wave. It turns out if you don’t have the shock wave, your brain says something is wrong—it doesn’t have mass.”

Sets & Cameras featured a model of a tree from “A Bug’s Life.” “There are new kinds of mathematical surfaces that are used, in making these objects,” he said. “But there are techniques to understand how light interacts with them. With humans, it turns out that light goes into skin and bounces out, so it’s like subsurface scattering.” When you go outside, he said, sunlight shines through you. “If you’re outdoors, the light goes through your nose. Some illumination happens inside the nose. You never think about that. Except that if you don’t have it, and you look at it, your brain says something is wrong. You look like you’re made out of some opaque material, like clay.”

At the Lighting Design station, Catmull paused in front of a model of Carl Fredricksen’s living room, from “Up.” “It turns out you are affected by the colors and so forth,” Catmull said. The artists create a script that plans out the colors and moods for each film. He moved levers on the control panel to adjust the lamps in Fredricksen’s living room. “And then we can move the sun around,” he said, pushing a button. Sunlight and shadow drifted across the room, as if sliding from afternoon to dusk. For a moment, the model room had the feeling of a childhood afternoon at a grandparent’s house, light moving across the floor, the day moving through its phases. It felt like real life.

I suggested that much of what the exhibit might help improve, even among the science-challenged, is observational skills. Catmull agreed. “There’s one thing I don’t think people really appreciate, which is how closely aligned artists and technical people really are,” he said. “Culturally, they bond because they’re really based upon observation. I think there should be more art in schools. When they cut funding, they cut from art programs, shop programs. But they’re really about seeing and developing observational skills, which is useful in science. A lot of the people here grew up on art and math, and this is a place to mix them together.” Observing science, I might add, is helpful to artists.

Catmull continued to the Animation area, which featured Edna Mode, posing extravagantly, and other characters from “The Incredibles.” “Animators have well-trained eyes,” Catmull said. He mentioned another animation insight: “If you turn your head without moving your eyes first, it looks like you’re dead.” He was like a magician revealing sleight-of-hand secrets—all of the tricks made instinctive sense, because they were things you’d seen but not considered—except that he kept describing how to avoid making things look dead.

The Simulation station had a screen of Merida and Queen Elinor, from “Brave,” on horseback, beneath a title that said “Hair & mane simulation ON.” The hair bounced and fluttered in the simulated breeze, and the manes flopped realistically. “The hair was a big friggin’ deal,” Catmull said.

Rendering, the final stage and the final station, featured “Inside Out,” and showed, on a series of screens in the round, how all phases of the process—including story—come together. The rendering process, demonstrated with a video of Joy at her neural workstation, began with a grid-based image that evoked “Tron,” and filled it in, pixel by pixel, glorious orbs and colors appearing more and more fully, with a legend above it updating us on how long the process would actually take. (“Actual render time: 33 hours.”) Catmull explained that the reason that “Inside Out” ’s emotion characters are brightly colored and oddly shaped, with force fields of little dots coming off them, was to make them less humanoid.

“You don’t want it to feel like it’s a person inside your head,” he said. Having a computer in your head, though, didn’t seem to be a problem.

SOURCE