BURTON DOES 3-D: Hollywood's TRANSITION TO CGI
© Continuum 2005
New York and London: Continuum 2005. 262pp, illus.
Hardcover: $85.00. Paperback: $21.95
It all started with Vincent.
Actually, Tim Burton's life-long love affair with stop-motion animation began when he was a child. His love of Godzilla movies led him to fantasize as a youngster that he would one day be the actor inside the Godzilla suit; by the time he was a teenager, he and a handful of friends were making Super 8 movies in Burbank. One of these was a stop-motion animation film, which Burton recalled quite clearly:
We made a wolfman movie, and a mad doctor movie, and a little stop-motion film using model cavemen. It was really bad and it shows you how little you know about animation at the beginning. These cavemen had removable legs - one was in the standing position, and the other was in a walking one - and we just changed the legs. It's the jerkiest animation you'll ever see. I used to love all those Ray Harryhausen movies - Jason and the Argonauts, The Seventh Voyage of Sinbad - they were incredible, I loved stop-motion animation as a kid. And as you get older, you realize that there's an artistry there too, and that's what you're responding to.
Burton began his commercial 3-D animation career with Vincent, a five-minute combination of stop-motion and 2-D animation that he made with Rich Heinrichs when both were still Disney employees. Burton had been working as a conceptual artist on The Black Cauldron (1985), but none of his ideas were used in the final film. However, his conceptual work got the attention of Disney producer Julie Hickson and Studio Vice-President Tom Wilhite gave him $60,000 to produce an animated film based on a children's story that Burton had already written.
The Disney Company was willing to fund such experiments because they were desperate for a new sense of direction. Walt Disney had died in 1967, making The Jungle Book his last film. The studio had been in the doldrums since then and was looking for a change. They were even considering making a children's movie using 3-D animation, but hesitated because of their unfamiliarity with the process. One of the ideas they were toying with was a shift from traditional cel animation, with which Disney had for so long been associated, to a completely different style. Burton and Heinrichs felt that '...we could convince the hierarchy that a feature-length, model-animated film with the Disney logo on it could be commercially viable, and Vincent was our way of showing them.' Vincent featured 3-D characters made from ball and socket models, and 2-D cel animation in a high-contrast black and white style. This new approach was partly inspired by Disney's own practice of making 3-D models for its 2-D animators to work from, as Rick Heinrichs described:
Disney likes to make sculptures of the characters for its animated movies so that animators can hold three-dimensional things in their hands. Tim and I learned that you can combine the graphic look that makes most of the design elements of a two-dimensional picture with something three dimensional. It finally came to fruition in Vincent....showing us we could have the best of both. We loved that expressionistic approach then, and Tim still strives for it in all his work.
Though the studio had provided the $60,000 budget, the project had 'off the lot' status - meaning they had to supply their own materials and find their own space. As a result they were able to make the film with little interference at the Dave Allen studio across town.
The film is a very creative combination of 2-D cel animation and 3-D stop-motion animation. It begins with a black screen with Tim Burton and Rich Heinrichs receiving equal credit for the film. Then the longest 2-D animation sequence begins: a black cat (vaguely reminiscent of the poster image for the film The Tomb of Ligea starring Vincent Price) emerges from behind a skeletal tree, hops onto the top of a brick wall, and the title appears on the wall in quasi-gothic lettering. The cat walks to the end of the wall and jumps into the window of a 3-D house - the first of many such seamless transitions from 2-D to 3-D. Inside the house, the cat, now a 3-D model, sidles up to Vincent, who is playing the tune on a recorder that we have heard since the opening credits. The music is a score by Tom Hilton, with melodic variations on "In an Egyptian Marketplace" and J.S. Bach's "Toccata and Fugue in D minor". When Vincent picks up the cat the voice-over narration, performed by Vincent Price, begins:
Vincent Malloy is seven years old
He's always polite and does what he's told
For a boy his age he's considerate and nice
But he wants to be just like Vincent Price
On this last line a stop-substitution, masked by a quick dissolve, transforms Vincent the boy into Vincent the boy-as-Vincent-Price, complete with long cigarette holder and evening gown vaguely reminiscent of Vincent Price's costume in Pit and the Pendulum (Roger Corman, 1961). He blows cigarette smoke in the air - a 2-D detail in an otherwise 3-D scene. He is still holding and petting the cat, who has been purring contentedly with its eyes closed, but when it smells the smoke it looks at Vincent and, startled by the transformation, screeches and leaps out of Vincent's arms.
Similar transformations occur on the beat of every second verse for the rest of the film. Vincent, still in his Vincent Price be-robed persona, walks into the next room, where we see his little sister and two dogs:
He doesn't mind living with his sister, dog and cats,
Though he'd rather share a home with spiders and bats
On "bats" Vincent pulls a light cord, which turns off the lights but leaves a pyramid-shaped lighted area that fills with fluttering bats. Again the transformation is smooth and seemingly instantaneous, all in 3-D, though the areas of light and shadow are indicated by 2-D drawing. The references, and the transformations that occur along with changes in lighting, continue. Vincent's aunt comes to visit and switches on the light; Vincent the boy allows himself to be patted on the head, but then he transforms into the Vincent-Price mad scientist persona, specifically inspired by the film The House of Wax (Andre de Toth, 1953) which was originally a 3-D film, meant to be watched with 3-D glasses. Like his hero, Vincent the boy lifts his aunt up on a chain and drops her into a vat of boiling wax. Still in his mad scientist guise he transforms his dog Abercrombie into a "horrible zombie" and the two skulk through the dark and misty London streets looking for victims.
Several critics have compared Vincent to the German Expressionist film The Cabinet of Dr. Caligari (Robert Wiene, 1920), as well as Nosferatu - A Symphony in Terror (F. W. Murnau, 1922) and M (Fritz Lang, 1931). Burton himself described it as "Dr. Caligari meets Ray Harryhausen," though the only obvious connection to Harryhausen is that the film is stop-motion. According to Burton, the primary source of inspiration was not Caligari, but the work of Dr. Seuss:
I certainly saw pictures of [Caligari], in any monster book there were pictures of it. But I didn't see it [the film] until fairly recently. I think it probably has more to do with being inspired by Dr. Seuss. It just happens to be shot in black and white, and there's a Vincent Price/ Gothic kind of thing that makes it feel that way. I grew up loving Dr. Seuss. The rhythm of his stuff spoke to me very clearly. Dr. Seuss's books were perfect: right number of words, the right rhythm, great subversive stories. He was incredible, he was the greatest, definitely. He probably saved a bunch of kids who nobody will ever know about.
The "London" set is the closest thing the film has to anything from Caligari. There are other details that could be seen as Gothic, but once we are alerted to the debt of inspiration owed to Dr. Seuss the homage to Seuss seems visible in many details: the areas of action illuminated with pyramid shaped spotlights that highlight the crooked checkerboard wallpaper; the sinuous wainscoting; the twisted skylights; the long silhouetted staircase that looks more like the teeth of a handsaw than steps, with a tiny, crooked door perched at the top; the feathered quill pen and the penultimate image of a door and floor that are curved out of shape in a hallucination that seems to be filmed through a fish-eye lens.
Smith and Matthews have pointed out that the plot of Vincent refers to stories by Edgar Allen Poe (the story Vincent reads is conflation of 'The fall of the House of Usher' and 'The Premature Burial' as well as filmic adaptations of these stories directed by Roger Corman, many of which starred Vincent Price: The Fall of the House of Usher (Corman, 1960), "The Black Cat" which Corman made into a segment of his Tales of Terror with Vincent Price in 1962 and The Premature Burial (Corman, 1962) which did not star Price). Vincent's high collared coat in the mad-scientist scenes and the electrical equipment in the lab seem to be obvious homage to Frankenstein (James Whale 1931). But these stylistic references are fleeting, while the rhymed couplets and the way the alternations between light and dark, child-gothic-fantasy and parental insistence on daylight and the suburban status quo that happen on the poetic beat are much more reminiscent of Dr. Seuss. Vincent's obsession with Poe stories and Vincent Price movies is similar to the obsession of Dr. Seuss's characters in stories like Green Eggs and Ham (1960) and And To Think That I Saw It On Mulberry Street (1937). The plot of Vincent is vaguely similar to that of The Cat in the Hat (1957) and The Cat in the Hat Came Back (1958), both horror stories about two children locked in a house with a mad feline visitor who systematically wrecks all the household symbols of middle-class suburban life.
In Vincent, by contrast, it is his mother's order to go outside, "If you want to you can go outside and play, it's sunny outside and a beautiful day" is what drives Vincent to his final attack of gothic fantasy, "I am possessed by this house and I can never leave it again." Vincent's final visions are a conflation of his imaginings, from killer zombie dog, aunt-in-wax, and beautiful undead wife represented in a mix of 2-D and 3-D animation, until Vincent collapses on the floor, dead or fantasizing that he is dead, as the camera withdraws from him and allows him to be submerged in a pool of darkness as he quotes the final lines from Poe's 'The Raven' 'And my soul from out that shadow that lies floating on the floor, shall be lifted Nevermore.'Apparently Burton meant for him to be fantasizing his own death, as he has commented that by the end of the film, 'the kid basically has the impression of not living...even though it's in his own mind.'
Smith and Matthews have described this film as having Death as a theme but no actual killing; this seems disingenuous when we clearly see Vincent preparing to dip his aunt in wax, and though we don't see her actually sink into the wax we later see the effects; we also see Vincent going out with his dog in search of victims, Jack-the-Ripper style; and we can only assume that, like the character in the Edgar Allan Poe story, Vincent - as-mad-scientist - is responsible for burying his "wife" alive. Of course, all of these killings or intents to kill are fantasies, but they are clearly depicted, leaving little to the imagination.
It was probably these representations of killing, the fact that Vincent might be dead at the end, or the idea that a child could harbor so much rage, that led one of the Disney executives to ask for a "happy ending", which Burton successfully opposed:
"They wanted it to have more of an upbeat ending, but I never saw it as being down beat in any way. It's funny, I think it's more uplifting if things are left to your imagination. I always saw those tacked-on happy endings as psychotic in a way. They wanted me to have the light click on and have his dad come in and go, 'Let's go to a football game or a baseball game.' That was my first encounter with the happy ending syndrome."
Getting Vincent Price to narrate the story was a real coup, both emotionally and commercially, for Burton. This is how Price remembers it:
[Around the end of 1981] They asked me to come down and meet Tim and Rick Heinrichs...So I went down to Disney and they showed me the drawings and mock-ups they had planned for this short film, as a sort of tribute to me. Tim recited the poem for me and asked me to narrate it. I was really struck by his charm and enthusiasm, so I said, 'yes.' Tim is really in love with film and is a wonderful kind of mad fellow. I thought it was marvelous of Disney to give these two kids a chance to make the film. They were only about 20 when they did it...
Price's narration was recorded in one day in December 1981, before the animation had been developed, but his relationship with Burton did not end there: he acted again in Edward Scissorhands and Burton has produced a documentary homage to Price called Conversations with Vincent (working title) which has not been released because of conflicts over the rights.
"He was very supportive. I always had the feeling he understood exactly what the film was about, even more than I did; he understood that it wasn't just a simple homage, like 'Gee Mr. Price, I'm your biggest fan.' He understood the psychology of it, and that amazed me and made me feel very good, made me feel that someone saw me for what I was, and accepted me on that level...It's a scary proposition meeting somebody who helped you through childhood, who had that affect on you, especially when you're sending them something that's showing that impact in a kind of cheesy, children's book kind of way.
Burton asked Price to give voice to Santa Claus in The Nightmare Before Christmas, but the actor by then was in failing health and had to bow out. Although Vincent received the Critic's Prize at the Annecy Animation Festival in France, as well as favorable comments at festivals in LA, Chicago and Seattle, Disney only released it theatrically in one Los Angeles theatre for two weeks, as an opener for Tex starring Matt Dillon. It was then shelved and did not become available to the public until the release of the Nightmare Before Christmas Special Edition DVD. The concept of producing 3-D animation at Disney suffered a similar fate, and nothing along those lines would happen at the studio for another decade, when Burton returned with A Nightmare Before Christmas.
Disney financed two more short films of Burton's: Frankenweenie (discussed in Chapter Two) and Hansel and Gretel (1982). Hansel and Gretel was a 45-minute episode on 16mm film for the newly inaugurated Disney channel, starring an all-oriental, non-professional cast, with a man playing the witch and a script by Julie Hickson. Although the budget was low -- $116,000 - Burton took the opportunity to pay homage to the design style and color schemes of his beloved Godzilla movies. He loaded the program with special effects, including front projection, forced perspective and a little stop-motion by Heinrichs and Chiodo who had done the honors on Vincent. The episode was shown only once, after 10 p.m. on Halloween, and is not available to the public today.
Burton did not use any stop-motion in Frankenweenie, his last film for Disney as a contract employee, nor did he use any in the version of Aladdin and his Wonderful Lamp (1984) that he directed for Shelley Duvall's Fairie Tale Theatre, even though Heinrichs and Chiodo again provided the special effects, including a tunnel embedded with cartoon skulls that would be echoed later in Pee-Wee's Big Adventure.
PEE-WEE's BIG ADVENTURE (1985)
In Chapter 2 I analyzed how the narrative in Burton's first feature film, Pee-wee's Big Adventure (1985), was structured and it's relationship to cartoon narration. Peewee is in essence a road picture, a series of disparate episodes linked together by Paul Reuben's performance as a child-man who is searching desperately for his stolen bike. Each episode is shot in the style of a different kind of film genre, from sports movie to coming-of-age to slapstick comedy. There is a long, complex and very funny sequence at the end where Pee-Wee makes his way across various Warner Brother film stages where different kinds of movies are being filmed, from a Miracle on 34th St or Angel with Dirty Faces style drama to a beach-blanket bingo film to a Toho Godzilla set, complete with Japanese director. This set includes an actor in a Godzilla suit who ends up in a sleigh with a Santa Claus careening across the Warner Brother's lot, almost a flash-forward to the stop motion animated feature Tim Burton's Nightmare Before Christmas, a film Burton wouldn't make for several more years but which he had already designed and storyboarded. Another reference to Nightmare is the skeletal reindeer that delivers Peewee's toast, taken directly from the reindeer design for Nightmare.
Although Burton was focusing on the live-action directing in Peewee he included two brief stop-motion sequences. A wonderful moment is a one-and a half second replacement animation by Stephen Chiodo, who transformed the ghost of Large Marge into a pop-eyed ghoul. Large Marge already had a hairdo and wide-eyed stare reminiscent of the Bride of Frankenstein in James Whale's 1935 film of the same name. The moment that Pee-Wee realizes she is a ghost, eternally doomed to drive her bus along the same strip of highway, is when her staring eyes expand into huge white eyeballs, her mouth becomes impossibly large and opens hugely wide revealing irregular, rotting teeth, and her long, stop-motion tongue uncurls towards Pee-Wee, like the tail of a noisemaker, followed by her protruding eyeballs that pop out on springs. There are similar moments in Beetlejuice (using rod puppetry manipulated by three puppeteers) and Sleepy Hollow.
There is also an homage to the stop-motion dinosaur movies that Burton loved, though only in the briefest of scenes: Pee-wee dreams that a red tyrannosaurus (a rubber doll with glowing white eyes) walks onto a miniature set and picks up his bicycle in its mouth, and begins to chomp down on it with its huge teeth. All of this is intercut with alternating shots of a stationary puppet of Pee-wee shot from behind and shots of the live Pee-wee from in front and above, in other words, from the point of view of the dinosaur. Although there was not much of a budget for this animation, the intercepting of the live Pee-wee with a model Pee-Wee is reminiscent of Ray Harryhausen's sequence in Clash of the Titans (1981) where he intercuts close shots of Calibos' expressive face with rear long shots of a puppet representing Calibos. Burton also has a brief reference to his own film, Vincent, in a moment where Pee-wee walks down an alley, preceded by his own huge shadow - reverse of the shadow of Abercrombie the zombie dog that follows Vincent through the London fog.
In his next film, Beetlejuice (1988), Burton had a bigger budget, though at $14 million, it was not huge, with only $1 million earmarked for special effects. This meant that the plethora of special effects required had to be done mostly on-set and in-camera. These effects included creature creation and makeup, opticals and miniatures, blue-screen, forced perspective, motion-control cinematography, and, not surprisingly, replacement and stop-motion animation. In most cases the stop-motion was tightly interwoven with live-action and puppetry.
The most important stop-motion is in the sand planet, the infernal environment populated with carnivorous sandworms that Adam and Barbara Maitland find surrounding their house after they have died. The sand planet was a mixture of three elements: a real set, puppetry, and stop-motion. Alan Munro, a former storyboard artists on his first outing as effects supervisor, realized that a twelve foot deep set with forced perspective (the distortion of the shape and scale of different set pieces and props which stand between the actor and the back of the set or background image, giving the impression of great depth on a shallow set) would be cheaper and easier than using blue-screen for the actors as had been originally planned. A six-foot-long rod puppet was built out of latex and foam to dive through the sand dunes, diving through channels in the sand that had been made for it to travel through. Forced perspective was used with the worm as well, with "sand flings" and clouds added optically later.
Another, larger puppet of the worm's head, with a mouth that could open to reveal another, inner worm that would threaten the Maitlands. Burton made the decision to augment the puppetry sequences with stop-motion, and hired Doug Beswick to build a three-foot version of the worm's head and upper body and the inner worm head. To make the worm move quickly Alan Munro skip-framed Beswick's animation. The worm originally had three stop-motion sequences, although one of these was dropped. The two remaining include the worm's first appearance, where it rears up out of the sand and reveals the inner worm, which tries to attack Adam Maitland. The worm's last appearance in the film also includes some stop-motion of the worm with Barbara Maitland on its back, breaking in through the ceiling, eating Beetlejuice, and disappearing through the floor.
The banister that turns into a snake was also originally planned as all-puppetry, but Burton was not happy with the results and insisted that the snake be re-designed as stop motion and hired Ted Rae to do the job. Rae chose to make a realistic-looking three-foot snake body with a cartoon-version of Keaton's face. A larger rod-servo version of the puppet was used for shots where the snake crawls between Delia Dietz's foot and under her dress, and for the dialogue shots, and the snake-like eyes were achieved with rotoscoped animation (where additional animated detail is hand-drawn over the already existing image).
The last bit of stop-motion animation in Beetlejuice takes place when Delia Dietz's sculptures come to life and trap the Dietz's so that they cannot stop Beetlejuice from marrying their daughter. The movement of the tree-like sculpture was modeled after the labored walking of a dinosaur, while the other sculpture inched along the ground like a worm. In total, the film has 300 effects shots, of which 15 are stop-motion.
Burton did not use stop-motion again for his next few films, Batman, Edward Scissorhands, and Batman Returns. After The Nightmare Before Christmas, discussed below, he produced Cabin Boy (1994) with Denise Di Novi, apparently intended as an "affectionate spoof of Harryhausen's Sinbad films" written and directed by Adam Resnick with stop-motion by Doug Beswick and Yancey Calzada, both veterans from Beetlejuice (the latter uncredited).
Cabin Boy is an interesting but ultimately a failed film, a fairy tale about an effeminate and arrogant aristocrat, played by Chris Elliot, who graduates from his posh and protected boarding school and sets out to get on a ship that will return him to his father and his pre-determined post helping run his father's business empire. Unlike the very macho and muscle-bound actors who played Sinbad in the Harryhausen films, who were fatherly toward their crewmen, seemingly without fear, and coldly gallant to women, Elliot's character is very much a boy, whiney, cruel, and inconsiderate, and an extremely arrogant abuser of his father's power and influence. His bad attitude gets him dumped out of his limousine before he arrives at his ship (modern limos and nineteenth century sailing ships, along with eighteenth century fashions are all combined). He ends up getting lost and boards a run-down fishing ship called The Filthy Whore by mistake, where he is forced to earn his passage by working as a cabin boy. This leads to a taming-of-the-shrew like conversion, and he is rewarded at the end by finding true love with a woman who is determined to swim around the globe and chooses to stay at sea with The Filthy Whore and her crew so that he can retain his new personality and be near his mermaid-like love.
The film has two stop-motion animation sequences, which stand out for their quality and because they include rare moments of warm feeling in the film. One is a brief animation of a "shark man", a character with a human upper body and a shark's lower body, as he slices through the water with his tail. A longer (18 shot) sequence of an "iceberg monster", reminiscent of the monster (a rod puppet) at the North Pole in the 1911 film by Georges Méliès called A la conquête du Pôle (The Conquest of the North Pole). In true Harryhausen style, the iceberg monster sequence creates a solid illusion that the stop-motion character interacts with the actors, poking one actor with its finger, attacked by the sailors with harpoon and oar, with Elliot's character stepping on its toe, a moment where it shows emotion convincingly, especially when its hand is melted away much to the monster's surprise. The final coup de grace is the moment when one of the sailors sprays the monster with coffee and the monster's beautiful wickedness begins to melt, ending with it shattering into icy pieces on the deck. Burton had another homage to Méliès in Nightmare in the form of the skeleton reindeer that pull Santa's sleigh, (which emerged simultaneously in Pee Wee's Big Adventure, as Burton was already designing Nightmare at the time, as the toy reindeer that drops Pee Wee's bread into the toaster).
THE NIGHTMARE BEFORE CHRISTMAS
As we've seen, Burton had originally produced Vincent with the idea that it would persuade Disney to produce a longer 3-D animation film. Once done with Vincent he wrote a story-poem that followed the rhythms of The Night Before Christmas, prepared numerous drawings, and Rich Heinrichs produced a model of Jack Skellington, the lead character. Burton's original intention was to have Vincent Price narrating what he originally thought would be a short film or a TV Christmas special similar to Dr. Seuss's How The Grinch Stole Christmas (Chuck Jones,1966, narrated by Boris Karloff) or the "animagic" musical with Burl Ives Rudolph the Red-Nosed Reindeer (Rankin Bass Productions, 1964). Burton then took his materials and showed them to anyone at Disney who would look at them. Burton compared the process of shopping the project around to being in "that show The Prisoner; everybody's really nice, but you know you're never going to get out, it's not going to happen...as you proceed, it becomes less and less of a reality."
Once he was chosen to direct Pee-wee he had to set Nightmare aside, but apparently never stopped thinking about it. After his success with Pee-wee, Beetlejuice, and Batman, the time seemed ripe to try it again, and he had his agent, William Morris, "ask around" at Disney to see who owned the property. Apparently, Disney did, though when Burton developed it he wasn't even sure if he was a Disney employee or not. "There's this thing you sign when you work there, which states that any thoughts you have during your employment are owned by the thought police."
That Disney owned the property turned out to be a good thing. David Hoberman, then President of Disney's Touchstone Pictures, was overjoyed to have an opportunity to bring Burton back into the Disney fold, and agreed to produce not only Nightmare but Cabin Boy and Ed Wood - the latter a project that Columbia had put into turnaround. Hoberman even allowed Burton his wish to make Wood in black and white. GQ described Hoberman as "bent on making the flat-footed Touchstone into a haven for the hip and the bankably outré". And why not? Burton had yet to make a flop.
The first surprise with Nightmare came with Burton's decision not to direct it himself. "If I had [directed it], I'd be dead before I ever saw the final version... the reason I originally got out of animation is because I didn't have the patience for it." Instead, the task of directing went to Henry Selick. Selick chose to base Skellington productions in San Francisco, where he lived and where there is a large pool of stop-motion professionals. Burton traveled up a few times while making Batman Returns and Ed Wood, but basically entrusted Selick with realizing his vision. On the "making of" featurette on the DVD, Selick credits Burton with the design of the main characters, the tone, and the look. He specifies that the art direction was based on what the crew thought a Tim Burton film should look like - that is, a combination of "German Expressionism and Dr. Seuss". Selick describes himself as "from the same planet, if not the same neighborhood," as Burton. Of Selick, Burton says: "He's someone whose got a lot of talented passion, and he knows a lot of people who do this work and do it well."
Burton was also primarily responsible for the story, as he worked it out scene by scene with Danny Elfman, who would then compose a song for each major scene. Shooting on the songs began before Burton's story was adapted by Michael Powell, a horror-novel writer, and the screenplay was written by Caroline Thompson, who had worked with Burton on Edward Scissorhands.
Even so, a lot of the look developed under Selick's supervision while Burton was elsewhere directing live-action films. Rick Heinrichs showed the animation team films by Ladislas Starevitch. Starevitch was born in Poland but did most of his work in France in the early twentieth century. He made stop-motion animation films using the carcasses of real insects such as grasshoppers and beetles, as well as the corpses of birds and other animals. His 1911 film The Cameraman's Revenge consists of a love-triangle story involving insects which also commented on the voyeuristic aspects of cinema itself while making fun of the more melodramatic aspects of American cinema. He later began to work with puppets and continued to make animated films into the sound era. What stands out about Starevitch's animation is the convincing emotional performance of his insect carcasses and later of his puppets and models. The techniques of imbuing an otherwise lifeless figure with a lifelike performance is applied in simulated animation figures to this day. Selick also mentions the German animator Lotte Reiniger who worked with shadow puppets and Wilfred Jackson's 'Night on Bald Mountain" sequence from Disney's Fantasia (1940). They consciously imitated the delicate cross-hatching textures of the drawings of Ed Gorey and Charles Addams; some of the characters seem to owe a debt of inspiration to cartoonist Ronald Searle.
Selick supervised the production of the storyboards. Once they were complete, the storyboard for the entire film was filmed, creating a "story reel". Each storyboard shot was edited to run the same length as the final shot in the film and synced up to a temporary dialogue track. As each scene was created in stop-motion, the storyboard scene was removed and the stop-motion version was inserted, so that the film was essentially made twice. Burton approved of the "storyboard film" but that didn't stop him from shutting down the production for a week when he decided that Halloweentown needed to have a darker look.
Nightmare is sometimes credited with being the first feature-length stop-motion theatrical musical, but that credit goes to Rankin Bass's Mad Monster Party (1967), a film that inspired Burton when he did Nightmare.
Although he would have been willing to make Nightmare as a TV Christmas Special when he was first shopping the project around at Disney, by the time Nightmare finally went into production he felt the only way to do justice to the story was as a stop-motion theatrical. The fact that so few other feature length stop-motions had been made illustrated why Disney had initially hesitated to take up Burton's and Heinrich's suggestion that stop-motion would revitalize their brand. As Burton's co-producer, Denise Di Novi, pointed out: "Usually you do stop-motion for a one-minute commercial, but we needed to build a whole studio from scratch. And we had to comb the world for animators." The production required over a dozen animators and 100 other artists of different types, 40,000 square feet of space, 230 sets and 74 characters, each of which had several puppet versions of themselves and separate parts that constantly needed replacing. The replacement technique had been modeled on that of George Pal, whose "Puppetoons" of the thirties and forties were a model for Nightmare. The Puppettoons were made from wood and had dozens of replacements heads for different facial expressions and lip movements. Nightmare had puppets made of plastic, with replaceable heads for Jack and a replaceable face mask for Sally, along with a whole set of eyelashes, eyepieces, and other individual parts. In spite the of unusual, elaborate production, the film cost $18 million (much less than the equivalent cel-animation film) to make and eventually earned $70 million and inspiring more animated cuddly monsters such as those in Shrek and Monsters, Inc. It is still a cult favorite today. But did it bring about the renaissance of animation, with Disney leading the way, that Burton and Heinrichs had originally envisioned?
When asked this question, Henry Selick gave a positive response:
It really is happening. This is the golden age of animation - it wasn't 1939-41. There's far more production going on, there are more independent animators, though it's not easy for everyone. Disney wants to do everything in-house and gobble up the rest of the world. They want to own anyone who is doing great animation. I think they have a deal with [the British stop-motion animation company] Aardman Studio. I hope they leave them alone and let them do what they do so well, though I wouldn't trust them to do so. They are also doing a film with Pixar [Toy Story], which is a company that does computer films with John Lasseter.
Selick went on to mention MTV as a showcase for independent work, such as the Olive Jar Studio, Jon Lemmon Films, Sculptoons, colossal Pictures, Cosgrove Hall, and his own. In addition to MTV, Paul Reuben's Saturday morning TV show, Pee-Wee's Playhouse (1987-1990) was a showcase for stop-motion, from Aardman's (whose short Creature Comforts, won the Oscar for best Animated Short in 1990), to New York Broadcast Arts, to the work of independent animators. Stop-motion animation was used by companies such as ILM for certain special effects in Star Wars (1977), The Terminator (1984), and Robocop (1987). There was also Will Vinton's Claymation (stop-motion with clay figures), who made the California Raisins commercial as well as the feature-length claymation film, The Adventures of Mark Twain (1985). In addition to such theatrical and artistic shorts, there was The New Gumby, a ninety-nine episode TV series from 1988 produced by Art Clokey who had brought us the original Gumby and Pokey in the 1950s. According to Thompson, Nightmare benefited from the audience for stop-motion generated by The New Gumby as well as from the fact that several Nightmare animators learned their craft on the TV show.
Selick and Burton did not break new ground technically with Nightmare. As Selick described it: "We took an old technique and did the highest-quality stop-motion that has ever been done for that many minutes. I think we moved stop-motion to a high level of performance in timing, lighting and computer-aided camera moves. We made it a serious contender rather than things that look like toys on a table top with two glaring lights."
The computer aided moves that Selick refers to are really a wonder to behold. There were two ways to film the moving shots on Nightmare: by blocking out the scene for the puppets with tape, marking the different camera positions for the tape, and as each frame is exposed, move the camera forward one notch. This system was used mostly for simpler moves such as camera pans or tilts. When a more elaborate move was needed, such as a flying move, a motion control camera is used. This is a camera hooked up to a computer. The length, speed and direction of the shot are programmed into the computer, so that every time a frame of film is exposed, the camera moves forward a fraction. Nightmare is unusual for stop-motion is that almost every shot has a moving camera, and some of the moving shots run for as along as 540 frames. The crane shots, tracking shots, the long, elegant takes, are Selick's particular touch.
The difficulty, patience, and length of time, not to mention the sizeable number of specialized animators, made Nightmare a transcendent example of its genre, one unlikely to be improved upon, although both Selick (who would go on to make James and the Giant Peach and Monkeybone) and Burton would continue to try. Burton, for example, began working on another stop-motion feature, The Corpse Bride, in 1993, just after Nightmare was released. This film is finally in production, the result of a collaboration between Will Vinton Studios, Tim Burton Animation Co, and Warner Bros Feature Animation, and is slated for an October 2005 release. Both cast and crew feature a long list of Burton regulars: Caroline Thompson who wrote Scissorhands and Nightmare worked on the screenplay (along with Pamela Pettler who worked on the script for Charlie and the Chocolate Factory, though the first screenwriter to work on the project was Michael Cohn in the early 1990s), Danny Elfman is doing the score, and the cast consists of Helena Bonham Carter as the voice of The Corpse Bride, Johnny Depp as the voice of Victor, the man who abandons his real-life wife (Emily Watson) in favor of the Corpse Bride. Other characters are voiced by Albert Finney, Richard E. Grant, Christopher Lee, and Joanna Lumley. Henry Selick is currently directing Coraline, another stop-motion feature produced by the Will Vinton Studios, so Burton is co-directing with Mike Johnson, who worked as an animator on James and the Giant Peach. This movie is based upon a 19th century Russian folk tale which apparently has some basis in fact, as anti-Semitic gangs would often attack Jewish wedding parties, killing and burying the bride in her wedding gown.
Although star directors like Selick and Burton both have stop-motion features in production, it seemed that their original vision, that stop-motion animation would change the face of animation as we knew it, was not to be. This was partly because a new method of producing three-dimensional animation was about to take precedence.
At the same time that Burton and Selick were putting the finishing touches on Nightmare, Universal Studios released Jurassic Park, directed by Steven Spielberg. Ironically, Burton had been one of the four serious contenders (along with Joe Dante and Richard Donner) for the project when Michael Crichton's novel was put up for bid. However, Spielberg won Crichton's confidence and the right to develop the film.
One of the reasons for this was that the author of the novel, Michael Crichton, had wanted to "make a dinosaur story that really worked, that wasn't One Million Years B.C.." Spielberg and his creative team shared Crichton's concern for realistic dinosaurs.
"What we tried to do," said Rick Carter, Spielberg's production designer, "was find the animal in the dinosaur as opposed to the monster in the dinosaur. The idea was not to make them any less threatening, but rather to keep them from doing as much monster 'schtick.' For our human characters, we wanted their situation to be more like they were being stalked by an animal that is a carnivore, as opposed to something that is psychopathic and just out to get them. That's one of the reasons we wanted to have herds of dinosaurs, to show that dinosaurs were just like any other life-form and that they lived out their lives in a somewhat naturalistic manner."
As we shall see, this concern with a 'realistic' dinosaur performance is key.
Traditionally, realistic animals in a film of this type would be accomplished using stop-motion animation, but Spielberg wanted to go with full-scale animatronics so that interaction between actors and dinosaurs would be more realistic. Because Stan Winston had had such success with the monster in Alien, Spielberg gave the Stan Winston studio a $65 million contract to produce a variety of animatronics so that it would be easier for the human characters to have realistic interactions with them, as he had done in different ways with E.T. and Jaws. Winston and his team were also very concerned about making the dinosaurs come across as animals rather than monsters, and his artists based their drawings on the latest research that showed that dinosaurs were probably descended from birds and behaved in a birdlike manner.
Creating the T-rex was obviously the biggest challenge. Using traditional animatronic techniques, Winston's team built a 5,000 pound animatronic robot. Once built the problem was how to animate it so that it would move with convincing fluidity. Finally it was decided to divide the T-Rex into an upper half and a lower half and have different hydraulic platforms, designed by a company that built flight simulators, run the motion. But how was the motion to be programmed in? Stan Winston finally decided to use telemetry suits, a device which electronically links a human controller to a robotic figure, so that any movement the controller makes while wearing the suit is replicated, almost instantly, by his robot counterpart. "In the middle of the night I woke up with the idea of creating a small T-rex puppet that would link to our full-size one" said Winston, "so that every axis of movement would be covered - about forty of them in all. With four puppeteers we could move the small puppet around manually, and every movement would be translated to the hydraulics and duplicated by the big dinosaur." In other words, the puppeteers performance would be motion captured and replayed by the animatronic.
An animatronic version was also built of the velociraptor, but because of its smaller size there was a also a low-tech approach for the upper body shots: a suit that a human operator could wear, not unlike the way Godzilla had been actuated in countless Godzilla pictures. "It's always an advantage to have a people-powered character because of the range of movements the human operator can produce," Said Michael Lantieri, who had the privilege of wearing the suit.
Another human body suit was built for just the raptors' lower half. Because the raptor was supposed to be light on its feet, Lantieri had to be suspended from above so that there was just enough weight on his feet to allow him to feel where he was going.
Spielberg recognized that not everything he needed could be accomplished with animatronics, and so he hired Phil Tippet, one of the developers of go-motion, to do about 50 go-motion shots for Jurassic Park. Originally experimented with in The Empire Strikes Back and perfected by Tippet for Dragonslayer (Matthew Robbins, 1981), go-motion uses puppets with rods that extend from their extremities and are attached to computer-controlled stepper-motors. The result is a kind of motion-control for puppets - the choreography of movements can be stored in the computer and repeated with variations. Although traditional stop-motion animators did not like the high-tech variation, it did have the advantage of eliminating strobing, a jerky look caused by the lack of blur in traditional stop-motion. (Some stop-motion traditionalists claim that instead of proper blurs the "images merely lose overall clarity during broad movements" which they find more objectionable than strobing.).
Tippet and his team spent four months creating two brief animatic sequences, (animatics are moving storyboards used as a guide for subsequent animation), almost fifty shots, using the go-motion technology. The two sequences were of the T-rex menacing the two jeeps that are stranded outside its electric fence and the two raptors hunting the children in the kitchen. These and other bits of footage were complied into a "Dinosaur Bible" and later used as the basis of pantomime by the on-set dinosaur operators. The animatics were produced by Stefan Dechant, using an Amiga personal computer and Video Toaster Effects to construct a 3-D representation of the T-rex which was then animated.
While all of this development was going on, Dennis Muren and his team over at ILM had used a combination of computer generated imagery and advanced morphing techniques, which, along with some puppet creations by Stan Winston's shop, had gone into creating the shape-shifting robot of Terminator 2.
Spielberg had asked Muren if he could create a stampede sequence for Jurassic Park using computer graphics, as it would be difficult to achieve using stop motion. Two of ILMs veteran animators, Mark Dippe and Steve Williams, thought they could do more than the stampede, and secretly built a T-rex skeleton in the computer, using scientific images sent to them from the dinosaur repositories in Calgary. Once they had a skeleton they animated it in a brief sequence. They showed this to Kathy Kennedy, Jerry Molen and Lata Ryan, and piqued everyone's interest, and were given the go-ahead to try something more ambitious.
For their second attempt they took a model of the gallimimus dinosaur designed by Stan Winston, and computer graphics artist Eric Armstrong fashioned a gallimimus skeleton in the computer and developed an animated running cycle for it. "After we built the skeleton," said Dennis Muren, "we animated about ten of them running along in a herd. For the background, we picked some photos out of a book on Africa and scanned them into the computer. ...we did two angles, ... one looking down over a prairie on these animals running along and the other was a view right down at ground level as they run past. It was the same animation in both cases, so we got two shots for the price of one." Although the animals were in skeletal form, everyone at Amblin was very impressed, and Steven Spielberg commissioned half a dozen shots: the stampede sequence he had asked for a originally, and a few grand vistas of dinosaurs dotting the countryside.
Still, the animators at ILM thought they could do more. Specifically, they wanted to take a shot at the T-Rex. For their next test they had to use film. They used a Cyberware scanner, which focuses a revolving laser beam on objects or persons and records the topographical data into the computer, on Stan Winston's fifth-scale prototype, and animated it against a still image of rolling hills. "The shot started out with the T-rex maybe a hundred feet away, about two-thirds of the size of the frame. Then it just walked toward camera, step by step, and we sort of titled up at the head as it passed by," said Muren.
This sequence would turn out to be the sequence that "changed the world", and just as it had been an animated dinosaur that had impressed audiences in Willis O'Brian's 1914 The Dinosaur and the Missing Link and especially his 1925 stop-motion version of The Lost World, so now it was another animated dinosaur that would change animation in Hollywood forever. As Spielberg described it:
"My intention had always been to use full-size dinosaurs as much as I could, but I knew that my long shots or wide-angle shots would need to be done with stop-motion or go-motion, just like Willis O'Brien and Ray Harryhausen had done. None of us expected that ILM would make the next quantum leap in computer graphics - at least not in time for this picture. We had seen the gallimimus tests... but they were just skeletons and they were on video. The T-rex was complete and on film and walking in daylight, making full contact with the ground. It was a living, breathing dinosaur, more real than anything Harryhausen or Phil Tippet had ever done in their careers. At the showing, Phil groaned and pretty much declared himself extinct."
It was clear that the go-motion from Phil Tippet would not be needed, and neither would any more animatronics have to be built by Stan Winston's shop. Instead, those contracts were re-assigned to ILM. However, Spielberg discovered that he needed Tippet's expertise at generating convincing animated animal performances to guide the computer animators. Tippet had to get some crash computer training, which alternated with him training the computer animators in pantomime to help them block out the shots using their own bodies. Furthermore, Tippet's animatics remained the definitive guide for the dinosaur movements.
Tippet had developed a system called Dinosaur Input Devices, or DID, (later was re-named Direct Input Devices), in which stop motion movements on a dinosaur armature were recorded by a computer using encoders. This information would then be used by an animator to generate the CGI footage. In other words, the CGI was motivated, when the DID system was used, by stop-motion.
In other words, the realistic dinosaur behavior that Steven Spielberg, was so concerned with was achieved by numerous processes that all accomplished one goal: the breaking down of a performance into various elements that can be then re-mixed and matched at the will of the films' animators, and ultimately, the director. As we have seen so far, telemetry, motion capture, and direct input devices were all used to capture performances that originated in another format, whether in a physical performance by an actor, a puppeteer, or a stop motion animation, and translate it to CGI. Various forms of scanning, from that of a still photograph of a valley setting or a dinosaur bone, to the all-around Cyberware scanner, feed images into the computer that can be manipulated by the computer graphics artists and then animated.
This process has become a concern, especially for actors, since the release of films like Star Wars Episode I: The Phantom Menace (George Lucas, 1999), Final Fantasy (2000), and Lord of the Rings II: The Two Towers (2002). Synthespians such as Gollum from Two Towers are a great concern, at least to actors, now, because they alter our notion of a performance and who can claim authorship for a performance - who can say "I created that character?" when we are talking about a character created out of digital imagery designed by a series of computer graphics artists, whose motions come from the motion capture of an actor's movements (in this case, voice and movements came from Andrew Cirkus), and whose voice might be dubbed in from yet another actor? This debate was highlighted in the 2002 Academy Race when New Line Productions tried to have Gollum nominated for a best supporting actor role.
However, in 1990-93, the years Spielberg was working on Jurassic Park, his principle stated concern was realism. As we have seen, from the author of the original novel, to the director, to the production designer, to the effects artists including Phil Tippet, "realistic" dinosaurs were a key concern. But what makes a dinosaur realistic? How do we know how dinosaurs really behaved? How do we know if they would be interested in eating people or not? We don't; and, probably, we never will. What Spielberg was really after was not realism, but credibility:
I never thought I wanted to do a dinosaur movie better than anyone else's, but I did want my dinosaur movie to be the most realistic of them all. I wanted the audience to say, 'I really believe this could happen today.' Close Encounters, in a way, was based on both scientific and popular belief that UFOs have existed, or at the very least, could exist. And there was credibility in that film that I drew upon in attacking Jurassic Park.
Spielberg accomplished his goal by changing genres: instead of making a monster film, or a disaster film, or an homage to either of those genres, he looked to scientific documentaries from television and the action-adventure genre he already knew so well:
What I was after was kind of like Nova meets Explorer, with a little bit of Raiders of the Lost Ark and Jaws mixed in. But if I had to aspire to a particular movie, it would be Hatari. To me, that was the high-water mark of man versus the natural in a feature film.
It was probably this change in generic emphasis more than anything else that made the dinosaurs seem so realistic, along with the incredible improvement in special effects.
Burton was surely aware of the success of CGI in Jurassic Park, but he had a real loyalty to stop-motion; particularly, he seemed to like the way stop-motion captured an individual animator's energy:
"When [stop-motion] is done beautifully, you feel somebody's energy. It's something that computers will never be able to replace, because they're missing that one element. For as good as computers are and as incredible as it will get and is right now, it goes back to painters and their canvases. This project [Nightmare Before Christmas] and these characters and these visuals, the only way that it could have been done was with stop-motion. Therefore, it's very specialized. I remember getting shots and each time I would see a shot I would get this little rush of energy; it was so beautiful. It's like a drug. And I realized if you did it in live-action it wouldn't be as good; if you did it in drawing it wouldn't be as good. There is something about stop-motion that gives it an energy that you don't get in any other form."
Burton made the latter statement to Mark Salisbury before 1995 and he went on to plan his next film, Mars Attacks! accordingly. The initial idea came from The 1962 Mars Attacks! Topps Trading cards which were included in nickel-packs of bubble gum. Read and considered together, the 55 cards had a story that was inspired by 1950s B movies, pulp magazine fiction such as that of Amazing Stories, as well as the films of Fritz Lang and other movies such as War of the Worlds. Burton and his screenwriter, Jonathan Gems, were also inspired by another set of Topps cards produced in 1988 called Dinosaurs Attack! written by Woody Gelman and Len Brown, who had written the Mars Attacks! series. This series had been withdrawn from circulation in the sixties due to parental cries of outrage, but was re-issued in 1994. That was when Burton convinced Warner Brothers to buy the rights to both sets of cards.
Having just come off the production of Ed Wood, Burton and Gems had the same goal: to make a big budget B-movie, using the cards as a basis, the way they thought Ed Wood would have made it if he had had the resources.
In an interview with the London Sunday Times of February 9, 1997, Tim Burton said that he had originally planned to use a great deal of stop-motion animation for his sci-fi satire, Mars Attacks. According to the article, when Warner Brothers realized that his film would not play to the lucrative sci-fi audience, they forced him to switch from stop-motion to CGI - and shave twenty million from the budget. Burton fought the change, until Industrial Light and Magic (ILM) showed him what they had done on Jurassic Park and Jumanji. Burton was impressed with how good it looked - "better than stop motion." Burton's decision-making process basically replicated that of Steven Spielberg on Jurassic Park just six years before.
Burton had already hired Ian Mackinnon and Peter Saunders, owner-partners of a special effects company based in Manchester, England, and given them the job of making stop-motion Martians. Burton specified that he wanted the battle scenes between humans and Martians in Mars Attacks! to be reminiscent of skeleton fight in Jason and the Argonauts, in which Ray Harryhausen had brilliantly combined live-action and stop-motion. Burton had originally seen the film as a youngster at the Avalon Theatre on Catalina Island, a cinema that was decorated with an underwater motif, with art deco shells and so on. For Burton, the theatre, its location, and the movie, with the mythology it evoked, seemed to be as one, and it made a strong impression on him.
He also wanted the "hub-cap like flying saucers" to be produced using stop-motion, and the visual design of many of the saucer elements echoes Harryhausen's work in Earth vs. the Flying Saucers (Fred Sears, 1956). Harryhausen had given his saucers a revolving inner section, time consuming to animate, but the result is that the saucers seem electrically alive. Harryhausen's saucers also flew in formation, raise and lower themselves on a central stems, dip and swoop as they fly, and have an animated ray gun that emerges from beneath and turns to aim at its target. The screenplay also copied some elements of Earth vs. the Flying Saucers, such as the Martians attempt to send a satellite signal from outer space which the humans have trouble decoding, and the film used a lot of stock footage to save money, the look of which Burton copied for his film. Other images from the Harryhausen film that Burton paid homage to was the gathering of people at the end with the destroyed Capitol in the background, the composite image of the saucer in the lake with live action people in the foreground, the pace of the scenes of destruction in general, and the destruction of the Washington monument.
In order to facilitate the stop-motion work for Mars Attacks!, Mackinnon and Saunders moved to LA and began to arrange a collaboration with the Skellington group in San Francisco, the company run by Henry Selick that had produced Nightmare Before Christmas, but Skellington was too busy with James and the Giant Peach. Mackinnon and Saunders got to work on their own, setting up a facility in LA and producing various models of Martians for Burton's approval. After viewing their models, Burton made two key decisions: that all the Martians should look the same, and that they should never blink. According to Karen Jones, who wrote the book on the making of Mars Attacks!, Burton made these changes to make the Martians more frightening, but they are also the kind of decisions made by someone working in stop-motion who is trying to cut costs.
When Larry Franco, the producer for Mars Attacks, who was fresh off the set of Jumanji, saw the first stop-motion tests of the Martians, he was reminded of the early phases of the CGI animal animation for Jumanji. It was he who asked Burton to meet with the people at ILM. Burton agreed with hesitation: he didn't like the look of Toy Story, (designed by John Lasseter, who had been at Disney with Burton) and that's what he thought he was going to see. At ILM, Mark Miller, who had been the visual effects producer on Jumanji and would soon be playing the same role for Mars Attacks! and computer graphics supervisor Jim Mitchell spent a month preparing a screen test of digital Martians against a real background, with a flying saucer and sound track.
When Burton saw the ILM test, he was finally convinced:
"When I first saw it, I was amazed. Every type of animation has a different vibe, and it's not something that you can really analyze or verbalize. But there was SOMETHING about the computer medium that seemed to work with these characters, because they were all the same, because they had a certain quality in their movement. Also, because we needed so many of them, that would have been much more difficult with stop-motion. To animate ten of them in a room would have been a much more difficult task. ...At the root of it, animation is animation. Each form requires its own special set of circumstances and expertise.""
As a result, just as Mackinnon and Saunders were ready to go into full Martian-stop motion puppet production, in November of 1995, Burton pulled the plug. Karen Jones gives a more detailed description of the problem:
"Faced with the incredibly demanding production schedule, the marriage of live action and animation proved too difficult. Due to the extremely time-consuming nature of stop-motion, Burton would have had to film the live-action plate shots - the background shots into which the animation puppets would be composited digitally - months before the other scene elements even could have been conceptualized, particularly those starring the live actors who would be filmed playing opposite the Martians."
The only Martian played by a live actor was the Martian femme fatale, played by Lisa Marie, Burton's girlfriend (the two had been introduced by Jonathan Gems, the screenwriter for Mars Attacks!). Because the mayhem in the film required life-sized Martian bodies, Mackinnon and Saunders were asked to produce fifteen full-scale Martians for use as Martian corpses in the film, and their design work was passed on to James Hegedus at Industrial Light and Magic (ILM), who inherited the job of creating 3D Martians, but now in the computer. Hegedus had worked as visual art director for Joel Schumacher on the Tim-Burton-produced Batman Forever. The digital architecture for Mars Attacks kept some influences from its stop-motion incarnation: many of the sets were round, as round sets made it easier for stop-motion animators to reach in and make adjustments.
The switch to CGI had certain benefits: the realistic interaction between Martians and live actors that Burton wished for would be easier to achieve, and Burton could now film the rest of the movie in an anamorphic format, which would not have been possible if he were still working with stop-motion. Burton was able to return to some design elements, such as the teardrop shaped Martian helmets he had wanted but had to let go of for stop-motion. The ILM animators even offered to leave off the motion blur that is added towards the end of the computer animation process, to make the images look more like stop-motion, and keep Burton's original idea of homage to the work of Ray Harryhausen. But Burton refused: if they had the ability to make the Martians look real, then that is what they should do.
Like any animation format, CGI has a long lead-time, and with a December 1996 release date, time was tight. A team of 60 at ILM was responsible for around 200 shots, mostly involving the Martians in such a way that they would match with the live-action footage and interact with the live actors. Warner Brother's own digital company, Warner Digital Studios, was responsible for the remaining 130 effects shots, including the Martian robot, the flying saucers and the scenes of exterior destruction. An in-house model shop built exact replicas of such monuments as the Eiffel Tower, England's Big Ben, and the Taj Mahal to be exploded on film. None of these model shots ever stood alone; all required computer graphic additions, like flying saucers and death rays. The visual effects division at Warner Digital was run by Michael Fink who had previously won an Academy Award nomination for Best Visual Effects for the penguins and bats he had produced for Batman Returns. Fink commented on Burton's photoreal, but completely stylized look for the animation:
"Unlike other films, where the effects you create are entirely photorealistic and completely modern, Mars Attacks! has a very different kind of production design. What we tried to do was re-create the feel of the fifties science fiction invasion from Mars kind of movies, but make it contemporary and modern, and completely photorealistic. It's a very fine line to walk."
The appearance of the Martian robot, like the style for most of the film, was based on the Topps bubblegum trading cards banned in the 1960s for their "violent subversive images" of Martian firing ray guns at semi-naked blondes. The Martian robot is a closely modeled on Topp's Trading Card (#32, "Robot Terror), as well as related denizens of '50s Sci Fi movies. For its movement Warner Digital Studios took cues from other Tim Burton characters, such as Edward Scissorhands and Jack Skellington. The robot, which Fink describes as "a two legged army tank," was transported directly from the page to the screen. "We had the reference from the trading card #32(Robot Terror) and we also had a reference from Wynn Thomas, who had an illustrator draw a proposal for a robot. Based on these, we actually created a robot in our computer. Quite often, we'll actually sculpt a creature in three dimensions and then digitize it, but in this case, we started from scratch on the computer."
By the time they had built and designed a robot model, the production company was already in Burns, Kansas, to shoot the robot sequence. There they filmed all the background footage or plates. Through computer-generated prestidigitation, these would eventually include the robot running down the road chasing Richie in his pickup truck, demolishing the trailer park, getting caught in the power lines, and finally, crashing to the ground.
"We photographed all those plate shots and, as carefully as possible, measure the road, the camera position, and any other things we thought we would need to know in order to recreate that world in the computer. We re-createdŃwith computer graphics - the power lines that the robot gets tangled in and the structures that tear down or that bend as he falls, so we had to decide where those things went," Fink explained.
"We took all the data and the background plates and brought them into our computers to re-create the environment. Texture-map artists were working on developing the proper textures and contours that were painted on the surface of the robot."
The animators themselves accommodated details like shadows cast by the robot and the clouds of dust that erupt as the machine runs down the road: "The robot weighs about one hundred tons, so when its foot hits, it MUST shake the ground. We jiggle the camera a little bit and added the dust it would raise and the dents it would make in the concrete," Fink adds.
The 30-second sequence took three months of effort. Warner Digital also created many of the effects featured in the desert landing sequence, in which the Martians first descend upon Earth. Warner Digital designed and animated the spaceship from which the Martian Ambassador emerges. But the ambassador was animated by ILM. Real helicopters were used in the scene, but the helicopters that were destroyed by the Martian spaceships death rays were animated.
In addition to Card #32, (Robot Terror), other cards were specific sources of inspiration: Card #22 (Burning Cattle) for the opening scene' Card #2 (Martians Approaching) for the appearance of the Martians, Card #19 (Burning Flesh) is the model for General Casey being disintegrated from the stomach outwards by a Martian ray gun, Card #21 (Prize Captive), is the source for Natalie's abduction, Card #5 (Washington in Flames) shows the Martian attach on the Capitol building, Card #13 (Watching from Mars) shows Martians gleefully watching the destruction of the Earth from their ship (Burton also showed the Martian watching Godzilla destroy Tokyo on their spacecraft TV for entertainment), Card #4 (Saucers Blast Our Jets) is the model for a brief scene in which a jet gets blasted out of the sky, Card #50 (Smashing the Enemy) shows human soldiers breaking open the skulls of Martians (who have red blood in the cards, and not green as it is in the movie), Card #40 (High voltage Electrocution) was the inspiration for the destruction of the Martian Robot, although in the cards it was a giant insect that was destroyed, Card #36 (Destroying a Dog) inspired the destruction of the Presidential pet in the film, Card #11 (Destroy the City) shows Martians running amok, Card #41 (Horror in Paris) depicts the destruction of the Eiffel Tower, except there is an enormous caterpillar involved, and Card #24 (The Shrinking Ray) is the inspiration for General Decker's shrinking end.
From the Dinosaur Attacks! card series, Card #43 (Business lunch), Card #45 (Anchorman's Peril) and #30 (A Kid Strikes Back) were also influential.
HOLLYWOOD'S TRANSITION TO CGI
These two high-profile cases --Jurassic Park and Mars Attacks! -- are indicative of a change that has taken place industry-wide, where much work that would have once been done with stop-motion and animatronics is now done with computer-generated graphics. Some critics, such as Mark Langer in his article "The End of Animation History", have pointed out that both practitioners and scholars need to come up with a new definition of what animation is, a definition that isn't based on calling animation "not live action cinema" but puts animation and live-action into a new relation to each other. Langer goes as far as to say "... a hybridization of animation and live-action ...this collapse of the boundary between animation and live-action ... can no longer be viewed as an aberration, but as a major trend of contemporary cinema."
We can put the relationship of animation and live-action cinema today into perspective by looking back at the relationship between the two at the very beginnings of cinema's history. Traditionally in cinema studies we have seen animation as a sub-set of live-action cinema. But if we compare the relationship between the two at the beginning of cinema's history we will see the aberration is not that the boundary between animation and live action cinema is collapsing now, but that the two were ever seen as separate to begin with. Langer mourns the loss of cinema's indexicality (the link between an image and what it refers to, a link live-action cinema is assumed to have naturally) and connects it to an overall cultural fear that we can no longer distinguish between simulation and reality. But cinema has always been about simulation; even live-action cinema is staged, arranged, carefully lit, filtered, and manipulated in countless other ways. Furthermore, live-action cinema gives the viewer a sense of photorealism. Until recently, we have associated photorealism with realism, but our thinking of it that way does not make it so. In fact, I would go as far as to argue that live-action cinema and animation were never really distinct mediums, and that live action cinema should be seen as a sub-set of animation.
I believe that we have misunderstood the primary drive behind changes in cinema production and exhibition. The primary drive is not a drive towards increased realism, based on audience demand, but a drive to mechanization. In my book on the first woman filmmaker, Alice Guy Blaché, Lost Cinematic Visionary, I argued that what processes such as the impulse toward color and synchronized sound in the cinema, which have usually been interpreted as responses to audience demand for increased realism, were really the result of an industrial drive to mechanization - to put it simplistically, to need of the film manufacturers to standardize production and exhibition in order to more reliably define markets.
The drive to mechanization in live-action cinema made itself felt in animation as well, from the Taylorization of animation studios initiated by John Bray, to the use of techniques like rotoscoping.
When I looked more closely at this drive to mechanization, it struck me that in many cases, whether I was looking at examples from live action or from animation produced at the turn of the 20th century, the drive to digitization was already apparent. In other words, the mechanization of cinema in the 20th century and the digitization of cinema in the 21st are related drives, acting on live-action cinema and animation in related ways. So it is not a matter of "improvements in animation technology make it impossible to tell animation from live-action, [and] improvements in special effects have made it impossible to tell live-action from animation," as Mark Langer said in the article already mentioned, but rather, that we ever saw the two as separate at all.
I will illustrate my point by taking three cases from early cinema: cases of early motion capture, early rotoscoping, and early digitization as represented by trick films, in the work of film and animation pioneers Etienne Jules Marey, Emil Reynaud, and Georges Méliès.
Marey, and his associate, Georges Demenÿ, were peers of Edweard Muybridge and pioneers in motion studies. In the pursuit of a better understanding of how the human body moves, Marey used an early version of motion capture: Demenÿ or other test subjects would wear black body socks marked with white dots so that only dots were recorded by the camera as the subject moved. When filmed, all that was visible were the white lines and the white dots that marked the joints, creating a skeleton dance version of the movement.
Emile Reynaud, better known as an early animator, also applied a form of digitization. In 1896 he adapted Marey's proto-motion picture device, the chronophotographe, to make a motion picture camera-projector and made a handful of films. The first of these was a classic vaudeville act by two clowns, Footit and Chocolate (black or in blackface), loosely based on an episode of William Tell: Chocolat has an apple on his head (and takes bites out of it) and Footit shoots it off with a water rifle, soaking Chocolat in the process. Once Reynaud had the film (shot at 16 frames a second) he took a few frames from one part and a few frames from another. These short selected sequences were then reproduced on the transparent celluloid, improved by drawing and coloring applied by hand and then strung into a sequential loop by joining them within in a perforated flexible metal band. Reynaud repeated this process with two other early digitizations-in-a-mechanical-format, one entitled Le Premier Cigare (Mimodrame Comique) in which a university student tried his first cigar and found it comically sickening, and another vaudeville act featuring a pair of clowns, called Les clowns Prince (Sc¸ne comique) made in 1898, which was never shown to the public. Unfortunately, none of these early efforts survive.
A similar method was used in 1899-1900, by the Brothers Bing of Nuremberg, along with other German toy firms, Planck, Bub and Carette, and the French Lapierre Company, all of who made cartoons for use in toy viewers based on live-action films. These toy cartoon animators invented a form of rotoscoping, tracing from early live action films such as the Lumi¸re film L'arroseur arros¸e, a Méliès trick film, The Serpentine Dance (Loie Fuller), 1901, Skiers (two films from 1900), Jumping Clowns, Clown and Dog, and Rider all by Ernst Planck, all from 1910. Rotoscoping continued to be important in animation films until the advent of digital motion capture. (Films such as Disney's Snow White (1943) relied on rotoscoping extensively, though the animators were reluctant to admit to it.)
A closer examination of trick film techniques shows that they also can be considered a mechanical version of modern computer simulation techniques. These techniques were then applied to animation using puppetry and models.
Trick films made before 1908 by artists such as Méliès working in his own studio and Zecca and Segundo de Chom—n working for Pathé included processes such as stop-substitution (stopping the camera and replacing a beautiful princess with an old hag, or a horse with a toy), filming in slow motion so that when projected at normal speed the film would appear speeded up, combining such fast-motion though superimposition with a regular speed sequence so that some characters moved at comically fast speeds and others at normal speed, cutting alternate frames out of a sequence to speed it up, shooting with the camera hanging upside down so that the film when projected normally would play the action backwards, fade in and fade out of a figure in superimposition to simulate the apparition and disappearance of a ghostly figure, and the use of props such as removable limbs, miniature sets, and miniature props. The list is much longer but this gives an indication of the creativity of the film manufacturers working in live action cinema before 1910.
Let's look at some of these techniques more closely. First of all we have stop-substitution. There is an early film, for which no title has been found, in which a man is run over by a car. The camera is stopped before the man is actually run over, and a real cripple with dummy legs is put in his place. After the car runs over the dummy legs they are separated from the cripple's body, leading to the humorous conclusion of the film, where a doctor who was in the car replaces the leg and instantly the man (through another stop-substitution) is able to rise and walk.
The effect achieved through stop-substitution is thus similar to that achieved through digitization in the recent film Forrest Gump (Robert Zemeckis, 1994), in which Gary Sinise is shown to be legless from the knee down. In both cases the goal is the same: to simulate an amputee when in fact the principle actor is whole-bodied. The difference is the means to achieve it: in 1904 the means was mechanical; in 1994 the means were digital.
Matthew Solomon, in his essay "Twenty-Five Heads Under One Hat: Quick-Change in the 1890s", has made a connection between the turn-of-the-century illusions of the quick-change artistry type, such as "...the rapid alteration of character through costume changes; chapeaugraphy, the manipulation of a piece of felt to form different hats; and shadowgraphy, the use of the hands to create human and animal figures in a beam of light," and digital morphing.
Placing metamorphic performance within a longer history of transformation that includes not only the emergence of cinema but also the contemporary proliferation of digital media.... Foregrounds a significant set of continuities. Viewed from the late twentieth century, one hundred years later, when the cinematic is being increasingly replaced by the digital, quick-change, chapeaugraphy, and shadowgraphy take on added significance, appearing not so much archaic as visionary.
One example of such a performance is the film Untamable Whiskers by Georges Méliès. This film is prescient, in that it is not simply a record of a quick-change performance - the transformations are too detailed for that - nor is it simply a series of a repeated cinematic trick, but rather both combined, much in the way that morphing combines performance and digital trickery today. As Solomon concludes, cinema abandoned these early attempts at morphing, though the tradition could still be found in certain animated films, but it reappears now with the possibilities of digital media.
Most of the tricks I listed above are based on some kind of stop-motion technique used in live-action films, and it might seem a stretch to talk about trick films in the same breath as 2D animation. In fact, most film historians generally focus on the influence of early trick films on animation films in terms of content. But we must not forget that animation itself is a product of stop-motion animation, as each drawing is substituted by the next, shot on another bit of film, until the whole gives the impression of movement.
There is no question that the paths of cinema and animation were joined at the beginning of their history. It appears that their paths diverged after 1907, when live-action trick films using stop-motion such as those made by Georges Méliès fell out of favor, though live-action films to this day rely on the techniques of animation for their special effects sequences. Trick sequences in live action films are an area of overlap between cinema and animation, an area where we can see that what we thought were two mediums were really one.
We can discuss Blackton's Haunted Hotel, for example, as live action film and as a trick, or animated film. The mechanical techniques for making film move, animating objects or imbuing line drawings with life were essentially the same until 1907 or so. As trickality lost its popularity, live action cinema and animation went separate ways, though we can still find areas of overlap, such as Ray Harryhausen's stop-motion sequences in otherwise live-action films.
Many of these effects are still used in live-action cinema today, but increasingly, the effects are achieved digitally, by manipulating the image as part of a computer sequence. Because digitization was difficult and expensive, it was only used sporadically from around the mid-1980s until 1993, when Spielberg made the switch for Jurassic Park. Since then other Hollywood directors, even those deeply attached to stop-motion animation like Tim Burton, have followed suit. And with films in the 21st century, it is becoming increasingly difficult to see where the digitization or animation ends and the live action begins. Think of all-digital films like Final Fantasy and films with almost-all digital sets like Star Wars: Episode II - Attack of the Clones (George Lucas, 2002). Even films we think of as live-action dramas like A Beautiful Mind (Ron Howard, 2001) have numerous digital or animated sequences; (as mentioned previously in A Beautiful Mind these include the montage of the passage of seasons and the desk that goes out the window).
Digital movies, cinema and animation, mostly separate since around 1907, are coming back together, and this forces us to reconsider the true nature of both mediums. I have argued here is that animation has been the dominant paradigm since the advent of motion studies. If we give the animation paradigm its due we should be able to develop a better understanding, not only of moving picture media, but also of the drive to mechanization and digitization that our culture is caught up in today.