Building the perfect beast
Movie special-effects maestro Phil Tippett has won billowing praise for the jaw-dropping digital transformations that turned models of alien bugs into the fearsome insect armies of Starship Troopers. But the 46-year-old founder of Tippett Studio in Berkeley, California, is his own most astonishing piece of transformation. If he's at the top of the bug-heap now, five years ago he was plummeting toward the bottom. No less a luminary than Steven Spielberg told him so.
Tippett is a big guy with a bald pate and a casual yet urgent manner; in conversation his mien runs from the boyish to the brooding as he careens from the grunts and growls of mythical brutes to the aesthetics of digital alchemy. At the crisis point of his career, he'd built a reputation as a design and animation paragon who put a distinct signature on medieval and futuristic monsters. His creations included the volcanic flying reptile of Dragonslayer and the elephantine Imperial Walkers in The Empire Strikes Back. These massive incubi had bone-rattling impact, menacing beauty, and the kind of horrific details that stick in a movie-lover's memory. Jim Morris, now the president of Lucas Digital, says, "When I started working at Industrial Light & Magic (ILM), Phil was doing the two-headed dragon, the Eborsisk, for Willow, and I asked Phil instantly, 'What's wrong with its mouths?' Phil said, 'It has diseased gums.'"
After leaving ILM's creature shop in 1983, Tippett funneled his own fascination with dinosaurs into an experimental short called Prehistoric Beast and an Emmy-winning CBS TV show, Dinosaur! When Spielberg set about launching Jurassic Park, he knew that Tippett was the man to make dinosaurs come alive. But Jurassic Park nearly became Tippett's Waterloo.
When I asked to see Tippett in action, his studio lent me a home video fit for a time capsule: a taped record of an early (October 1991) Jurassic Park work session at Spielberg's Amblin Entertainment headquarters in Los Angeles. It records Tippett and animatronics expert Stan Winston introducing the director to his main characters, the T. rex and the raptor, with detailed scale models and an array of drawings.
Tippett is the meeting's magnetizing force. While the others get caught up in camera angles and mechanics, Tippett keeps forging an imaginative connection to the dinosaurs as animals and characters. He punctuates his stream of perceptions with mime, whoops, and whinnies: His body language bespeaks an uncanny identification with the prehistoric beasts. With his arm sweeping up from his waist to his bobbing head, he talks about dinosaurs as "conscious beings who filter the world through their eyes to their brains" and are always "looking, looking." He could be describing himself. At one point, he reminds everyone that animal behavior includes stupidity: He imitates a behemoth slamming down on a Land Rover, then dumbly straightening up and twirling around because it doesn't know why it can't get in. "It's not necessarily dramatic," he says, "but it's real."
You can see Spielberg sparking to Tippett's gutsy, informed enthusiasm, his I'll-say-anything confidence. At that point, there was no reason for Tippett to doubt that he would be delivering dozens of dinosaur shots to Spielberg using "go motion," an up-to-date version of traditional stop-motion animation.
Go motion was state-of-the-art in the early '90s. But there was trouble on the horizon; you can sense it on the tape, when Spielberg and company ooh and aah over a rough computer animation of the T. rex circling the Land Rover. Computer animators at ILM, hired to embellish Tippett's effects, were instead conjuring ways for digital graphics to supplant them. Spielberg had scheduled the computer jocks to do only a couple of herd shots, but the results of their experiments knocked him out. He canceled the go motion. The way Spielberg has told the story (as quoted in Joseph McBride's biography), he and Tippett watched tests of computer-generated dinosaurs moving smoothly through bright sunlight. Then Tippett turned to him and said, "I'm extinct."
Cut to 1997--and Tippett lives. And thrives: Starship Troopers opened with a $22 million weekend gross. When I visited him in his lodge-like office at Tippett Studio in August, he took off his shoes, let down what's left of his hair, and described himself as being "physically debilitated" when Spielberg decided to work primarily with computers. "It was such a horrendous proposal," he said. "Basically, everything I'd done practically since I was able to walk was not to be used anymore." How Tippett got from there to here is the story both of one man's reinvention of himself, and of his fight to keep movie art in the computer age honest, messy, and true.
Tippett has been a lifelong devotee of stop motion as practiced by masters like Willis O'Brien in King Kong (1933) and Ray Harryhausen in The 7th Voyage of Sinbad (1958) and Jason and the Argonauts (1963)--an intricate, painstaking art in which animators pose and photograph miniature figures frame by frame. He wasn't alone. Just about every top animator or effects person today has favorite Harryhausen figurines, such as the part-rhino, part-centaur Cyclops, the undulating serpent woman, and the two-headed roc bird from Sinbad. In traditional stop motion (still practiced by Henry Selick in marvels like The Nightmare Before Christmas) the camera records a series of subtly different poses rather than actual shifting, so the resulting flow of images is inherently surreal--ultra-sharp and jerky. In the go motion Tippett helped develop at ILM starting with Dragonslayer in 1980, motorized and computer-governed rods produce believable blurred movement.
Computers weren't content to stay in the wings. In 1985, director Barry Levinson (Diner, Rain Man) was pondering how to nail down a difficult effect for Young Sherlock Holmes. He had a stained-glass window with a knight on it: He wanted the knight to jump off the window, march down the aisle, and whack a priest with his sword. Levinson recalls poring over the storyboards with ILM visual effects supervisor Dennis Muren and thinking, "We should see the stained-glass man come toward us, and then [have the point of view of the camera] begin to counter-move so that we'd [end up] over his shoulder looking at the [priest]. We needed to get into the next generation of special effects to have those two things happen." To capture the action in a continuous move and countermove, Muren turned to Pixar, then George Lucas' computer-animation group; John Lasseter (who later directed Toy Story) was among those responsible for the effect. Levinson didn't dwell on it: "I got a rough look at ILM on a video screen when they were just doing early stages, and I thought, 'Wow, if we can get this to work, it will be a great little moment.'"
That stained-glass knight wasn't just a little moment: It earned pages of coverage in American Cinematographer. This knight was also a giant frog in the small pond of mid-'80s computer graphics. Digital visionaries and animators alike saw him as a clue to the next direction--one that could change the relationship between special effects and the rest of movies.
In the golden age of Hollywood, effects sequences were often the lonely high points of epics, spectacles, and fantasy or adventure films. They were isolated in their position in the movies and isolated in the way they were made. That began to change in the '50s, when puppet masters George Pal (Destination Moon, The Time Machine) and Harryhausen developed enough clout to seize control of entire productions. In the '60s and '70s, a series of collaborative leaps--made by Douglas Trumbull and Stanley Kubrick in 2001; by Trumbull and Spielberg in Close Encounters of the Third Kind; and by ILMers like Muren and Tippett and Lucas in the Star Wars trilogy and beyond--brought effects teams and directors close together. And after Young Sherlock Holmes, filmmakers began to realize that the computer enabled them to weave the most whimsical or dangerous effects even more intimately into the fabric of a movie.
That hasn't happened yet. In 1997, effects are largely still a carnival attraction. Levinson compares the digital boom to the emergence of color television: "When the sets for the TV shows all had to be very colorful, game-show sets had panels with nine different colors. Everything went haywire and became garish. Each new invention basically gets abused in some fashion until good sense takes over."
Levinson's one-time collaborators at Pixar were the first to find a way to integrate digital effects organically into a feature--by making an entire film in the computer. Their pop phenomenon Toy Story took place in a digitally stylized Everysuburb that the mass audience instantly read as "home." With Toy Story, Pixar's moonstruck wiseacres proved they could bring self-mocking fun to their most intricate bagatelles and whim-whams. A direct-to-video sequel, Toy Story 2, is scheduled for 1998, along with Pixar's next full-scale feature: A Bug's Life.
Steve Jobs bought Pixar from Lucas in 1986, but it always had a separate identity from the rest of the Lucas empire. Lucas hoped to use computer graphics as one tool in his live-action movie workshop. John Lasseter and company aimed to create shorts and then features solely from computer imagery.
ILM's digital effects team is best known for bringing photo-realism into fantasy, and that can become monotonous. The dinosaurs in The Lost World are imposing but not haunting; they lack the mysterious gravity of Harryhausen's handmade Beast From 20,000 Fathoms. Speaking with a gifted animation supervisor at ILM, I kept asking whether the push to have computer-generated imagery duplicate physical reality was self-defeating. What would be the point? After all, it's not the accuracy but the poetry you remember from Willis O'Brien's King Kong. He looked at me kindly and patiently, and replied that it would take years for a computer even to approximate the complexity required to render the tissue, bones, and muscles of, say, a wrist.
When I pose the same question to Lasseter in a phone interview, he gets it: "Reality is just a convenient measure of complexity, a standard you use to judge your tools and abilities. I like to take a full step back and create something that doesn't exist and looks real; and if you keep doing this, the world is wide open as far as what you can do...
"What I like to do is not blend computer stuff with live action, but to create another world and have the audience sit there and know the world doesn't exist, but still feel it's believable. The computer allows us to create the believable-unbelievable, or the unbelievably believable. When [celebrated children's book author-illustrator] Bill Joyce saw Toy Story, he said, 'It's not like something you made, it's like something you dreamt.'"
Back in 1992, Phil Tippett saw his dreams crumbling on the computer screen. But in the early days of Jurassic Park, there was a terrible crash at the intersection of art and technology. It soon became clear that computer animators weren't immediately qualified to visualize mammoth reptiles dynamically and persuasively. As Spielberg's "dinosaur supervisor" (as his credit on the film read), Tippett schooled a corps of ILM and Tippett Studio animators in animal motion and behavior, encouraging them to prepare to "play" dinosaurs as actors would, with everything from mime and dance classes to field trips to animal sanctuaries and museums.
The ILMers, says Tippett, had to key into the manifold bizarreness of real-world movement--"a twitch [for example] a dinosaur might make before it started to turn. Only then could they begin to understand the kind of reflexes and action they needed to emulate."
In collaboration with ILM, Tippett's close associate Craig Hayes developed the direct input device (or DID, also known as the dinosaur input device or the digital input device). The DID, which Hayes had been thinking about for years, is basically a skeletal stop-motion puppet rigged with electronic sensors at each joint. The sensors record information on a controller box that translates it for software and use in a computer. From Tippett Studio's perspective, the DID allowed stop-motion artists to keep a tactile connection to their work and animate computer-graphic characters without learning a whole new technology. At the ILM end, it enabled computer animators not yet at full dino speed to study data that signaled the weirdness and anomalies of animal movement.
For Tippett, "it was extremely painful, the entire process of coming to grips with the computer." He still insists, "The computer doesn't like to do anything that's really good," and regards the video display terminal as a "one-eyed monster with a keyboard. We're people who live in a multitude of environments; to just sit in an efficient work station is pretty criminal. And it's a false economy." Tippett holds no rancor toward computer animators. But generally, he says, "I prefer people who have some experience working in the real world. They have more of an overall idea of things. If you live exclusively in a virtual world, there's a litany of details that you don't think are important, but they are."
Tippett wants to stay in a warts-and-all mode of filmmaking. ILM has by and large jettisoned the DID, but Tippett Studio has continued to refine it. Hayes designed one that could be worn as a body suit for the "animatics" (animated story boards) on Dragonheart. For Starship Troopers he came up with another that could move (in Hayes' words) "like a doll, in real time" (rather than frame by frame), an enormous help for conjuring crowded action scenes of bugs "jumping over walls, snapping, slamming into people."
Starship Troopers is one of the most effects-intensive projects in movie history, containing more than 500 effects shots. Tippett pulled off 250 of them. (The Lost World, by comparison, had about 200.) He did more than rise to the occasion: He made the occasion with his most tumultuous and imaginative menagerie yet. The soldier bugs have heads like staple removers and limbs like corner pieces in a razor-sharp erector set; when they cover the horizon like the massed tribesmen in Zulu, they erupt with a primal xenophobic aggression far more persuasive than the humans' teen spirit. At the high end of the insect spectrum are obscenely endomorphic brain bugs that are like abstract cartoons of effete, sedentary, polymorphous-perverse intellectuals. It's a bestial rogues' gallery worthy of the medieval craftsmen who took their cues from Revelation.
It's arguably the biggest foray into total filmmaking for any maverick creature-builder since Harryhausen. The project has led Tippett to treat the computer with grudging respect: "It enables us to blend every kind of special-effects technique with animation, so that you're not doing your shots in a vacuum, but you're involved with the plan and construction of the movie, with the director and producer, the editors, all the way through to the sound people."
Judged strictly by the calendar, Starship Troopers has been a four-and-a-half-year project for Tippett. But it's rooted in his long professional history with producer Jon Davison, starting with the jovial 1978 Jaws rip-off Piranha, and with director Paul Verhoeven, which began with the 1987 smash RoboCop (on which Davison was executive producer). As Verhoeven explained to me over the phone from Los Angeles, his impetus for Starship Troopers came from an earlier project he'd hatched with Tippett: "It was set 60 million years ago, when dinosaurs were the rulers of the world. It was going to be about a little doglike animal who befriends them before a meteor hits and destroys them all...There were battles between good and evil, but no human beings. It would have cost anywhere from $30 to $45 million, and Disney thought that was too high."
When Starship Troopers came along, Verhoeven saw it as a chance to do a similar Darwinian spectacle with Tippett, whom he regards as "a genius in his field. I had it in my contract that I would do the movie only if Phil Tippett were doing the bugs...Even before the filming, I felt that Phil was acting as a co-director.''
Craig Hayes was also in from the start. Tippett notes, "Craig designs the things to look good; I have to make sure they work." And Hayes agrees: "I try to create something visually strong. Phil's input comes with the movement, how a bug would reach in and grab someone and cut him in half." The two of them consulted with Verhoeven, Davison, and screenwriter Ed Neumeier. Hayes says they agreed on a central concept: "This was to be not a fantasy film but a war film. The bugs would almost be mundane--instead of having five thousand legs floating in some other dimension that would take an audience the whole movie to comprehend, we wanted to keep them grounded in reality, so people could register what they were and get on with the rest of the movie."
Tippett was on-set for the filming of the battle scenes, where the actors had to imagine their enemies. He tried to embody the insects to help Verhoeven pitch the tone for the actors and--most important--signal where they should look during one-on-one or one-on-swarm combat with arachnids. He rigged two poles with a length of rope, like a surveyor's tool, to show the actors both the length and position of the insects. And, Verhoeven says, "Phil always took the front pole, stretching the pole forward and impressing the actors with the threat. He went into 'attack' mode."
The attack mode is a good one for an effects specialist to be in these days, when frustration and anxiety temper the exhilaration of being in demand. The typical request for effects shots per picture has skyrocketed from the dozens to the hundreds. The chaos has sent the small but burgeoning industry into a tizzy. Special-effects houses can go under from not winning big contracts or underbidding for them, by failing to meet deadlines, or by producing less-than-brilliant results. The cost of software, computers, and animators has shot up quickly and fatally. For its investment to pay off, a company must generate a stream of assignments. But sometimes it's hard to snag new business when you're swamped by a single huge project--like Titanic. ILM's chief competitor in the live-action field, L.A.-based Digital Domain, got the contract for Titanic, but industry talk has it that the company underestimated the project and lost money. Digital Domain laid off 31 workers when the company couldn't segue into another jumbo production.
Jon Davison admits to feeling worried for Tippett, who expanded his West Berkeley base and built his staff to 100 while laboring on Starship Troopers. "One of the things that happens is, you end up acquiring more people, more machines, more software, more space--more overhead. And as the overhead expands, so does the risk that you won't be able to support it." Davison takes the collapse of the 90-strong Boss Studios in Los Angeles last August as a cautionary tale. The studio's founder, Richard Edlund, achieved industry-wide fame at ILM for his effects work on the Star Wars trilogy; later, Boss won its own acclaim for films like Ghostbusters and Die Hard. But, Davison says, "the profit margin in the effects business is very small, and because every time out of the box you're trying to create something new and different, it's a risky financial venture. You can get a director who changes his mind and sinks you; software problems can sink you."
Right now, Tippett's working on effects shots for Virus, with Jamie Lee Curtis and Alec Baldwin, and Disney's big-screen My Favorite Martian, starring Christopher Lloyd. His next all-out project--tentatively called Expedition--could be an artistic breakthrough for the same computerized image-making that almost put him out of business. As far as the plot goes, he and producer Davison will say little more than it's "about an expedition to another world." If Tippett gets his way, everything in it will be computer-generated, including its human characters. Of course, Toy Story was fully computer-generated, too. But that was a family cartoon, and its humans were deliberately peripheral and more two-dimensional than its toys. Tippett insists that he can bring off Expedition as an adult feature whose computer-generated protagonists are as complex as humanity. If he does, he will open uncharted realms of expression for high-tech artisans.
Although he won't divulge any details, Tippett has a fix on his goals. He refuses to make his animated figures merely mimic humans. His background is in fine art, not computers; he thinks that any attempt "to get a photographically representational human being that looks and acts like a real live person" would end up "boring" or "grotesque." He wants his computer-graphics dramatis personae to be stylized and of a piece, and to contain potent emotion. Tippett feels that special-effects films risk crippling live actors, who can't interact with their environment when they're emoting in front of a blue screen. But a computer actor in a computer set, in Tippett's view, "doesn't miss anything; he's just reacting to a different kind of vision."
The attempt is full of paradoxes, but paradox is Tippett's brain's default mode. Despite a project that sums up the state of intelligence (artificial and otherwise) in the world of movie arts and crafts, he contends that he's "no visionary." While saying he's "technologically indifferent," Tippett is driving his studio to a serrated cutting edge. He still adores Harryhausen's home-grown surrealism, and believes that it's crucial for filmmakers to connect with what thrilled them as children. But, with computers, he pushes his own monsters and robots past the jitteriness and funkiness that are part of the old artifacts' charm.
Tippett abhors any novelty, digital or otherwise, for its own sake. When he's viscerally engaged, he feels he taps into his (and perhaps the collective) unconscious--images and feelings that exist in millennial man's psyche but could also be seen at the dawn of art: "You want something possibly magical or deeply mythological to emerge, and the skills and level of imagination you're using go back to the Lascaux Caves.
Get the This Week's Top Stories Newsletter
Every week we collect the latest news, music and arts stories — along with film and food reviews and the best things to do this week — so that you’ll never miss Observer's biggest stories.