CGI Story: The Development of Computer Generated Imaging

Alex Schure founded the New York Institute of Technology (NYIT) in 1955 to take advantage of the wave of students studying under the GI Bill. NYIT was initially a correspondence school awarding technical certificates. It distinguished itself from the other technical schools by sending graduates a mess of electronics parts supposedly taken from Soviet warehouses during World War II. NYIT was a huge success, eventually acquiring three campuses in Manhattan, Old Westbury, and Long Island.

Alex Shure

Alex Schure, 1920-2009

Walt Disney had died in 1966, and Schure watched as the studio Disney had founded entered a creative dark age. Films like The Aristocats (1970) hardly lived up to Walt’s legacy and received lukewarm reviews.

Shure hoped that he would be able to take advantage of his wealth and new technology to become the storyteller of the generation. He established a conventional animation studio at NYIT’s Westbury campus to produce feature films.

Schure was much more interested in the undeveloped world of computer animation, however. He established a Computer Graphics Lab to research the future of computer animation. Ostensibly, the group would work on a feature film, The Works, that would be produced by Schure, but he knew that it would be years in the offing, if not decades.

NYIT’s Old Westbury campus was straight out of F. Scott Fitzgerald. Huge estates owned by Rockefellers and Vanderbilts (nowadays occupied by Howard Stern and Victoria Gotti) neighbored quaint cottages and servants quarters. NYIT’s Old Westbury was actually just a few blocks of mansions retrofitted into classrooms and labs. The computer graphics lab was housed in the garage of one of the mansions, which actually had a bear pit in the backyard.

The Manhattan campus was also expensive; it was adjacent to Lincoln Center.

Ed Catmull

Schure’s interest in computer animation was sparked by a video he received in 1974 (videos were still quite rare in those days) of a digitally animated hand and face by Ed Catmull, who was finishing his PhD at the University of Utah. Catmull had developed texture mapping. He made a model of his hand by plotting points on notable features. He then “coated” the hand with digital skin, the first example of texture mapping.

Schure was immediately taken with the sophistication of the video and appointed Catmull the head of the Computer Graphics Lab, where he had an unlimited budget and no release dates.

Catmull used his budget and connections at the University of Utah to buy six framebuffers. Framebuffers are video output devices with memory for a single image. These were initially quite expensive. Film resolution, six megapixels, would cost tens of millions of dollars, making it unfeasible even for NYIT. As a result, the researchers had to settle for less than film quality, which had the added plus of being easier to manipulate while still good enough for television.

Xerox PARC and SuperPaint

Meanwhile, history was being made on the other side of the country. Xerox’s skunkworks, Palo Alto Research Center (PARC), was in its heyday. The world’s first workstation, the Alto, and the world’s most prevalent networking standard, ethernet, were being developed in the early 1970s, potentially making Xerox the standard in office systems.

These innovations would make billions for other companies (Sun, Apple, Microsoft, etc.); Xerox usually failed to capitalize on their inventions.

One such product was SuperPaint, which was developed by the eccentric Dick Shoup. On April 10, 1973 – the same day the original Alto was being booted up for the first time – Shoup was testing the world’s first digital cameras and one of the largest framebuffers ever created. The minicomputer and framebuffer that Shoup used to manipulate his images cost the company hundreds of thousands of dollars, and the product didn’t have an application in the office market. Shoup’s supervisors were not all that impressed with SuperPaint. Still, he was permitted to press forward with his project and make major improvements, including upping the resolution and adding color.

Alvy Ray Smith

Dick Shoup invented SuperPaint, but Alvy Ray Smith made it famous and even created practical applications for the project. Smith was a hulking hippy-esque computer scientist in the same vein as Richard Stallman. His PhD thesis on Game Theory was featured on the front cover of Scientific American, and he served as a professor at NYU.

A serious skiing accident that put him on his back for a few months caused him to rethink his life and career, and he moved to California, where he studied and taught a few classes at Berkeley. One weekend he had to visit Stanford for a conference and had dinner with Shoup, who was a friend. Shoup told him about his work on SuperPaint; Smith was immediately enamored with the work and was given a tour.

Smith managed to get a job at PARC, where he worked on the rapidly improving SuperPaint. PARC management wasn’t willing to hire him outright, so he was paid using a purchase order. He spent the bulk of his time creating psychedelic videos demonstrating the resolution and color of SuperPaint.

Eventually a subculture developed around SuperPaint, which attracted the attention of Bob Taylor, their boss. In his mind, Alto and the office were Xerox’s future, not psychedelic videos and hippies. Taylor canceled Smith’s purchase order and disassembled the framebuffer.

Smith had heard about the work being done at the University of Utah. To his surprise, he found that NYIT (still a very obscure school) had bought all the framebuffers and set up a lab. He talked to Schure personally and secured a position on the team.

The Glory Days of the CRG

Smith and Catmull were the two “adults” in the team. Both men had earned PhDs and had managed groups of students and researchers. The Computer Research Group (CRG) wasn’t really different, since they weren’t working on any specific project. The Works, Catmull estimated, was still decades off.

The team puttered away at the problems of computer animation (mainly low resolution and limited computer resources) and acquired quite a reputation. Academics at heart, Catmull and Smith encouraged their team to at least review journal articles – and to contribute to the articles if they made a breakthrough. NYIT had some of the nicest equipment in America and was producing very attractive and somewhat realistic animations, which was unprecedented in the small field of computer graphics researchers.

Tubby the TubaThe researchers at CRG should have been content to keep working with their unlimited budget and benevolent patron, but their confidence in Alexander Schure was short-lived. Schure had founded his conventional animation studio at the same time as he founded the CRG, and over three years they produced the Schure-directed Tubby the Tuba, which was distributed by MGM.

Alex fancied himself the new Disney, but Tubby was abysmal, both technically and creatively. The animation was dull and even muddy at the action scenes. The story was depressing and proved to be unpopular. Smith and the rest of the CRG attended one of the early showings and were disgusted. At the end of the show, Alvy turned to one of compatriots and said “This guy doesn’t have it.” (The Second Coming of Steve Jobs, 93)

It had always been the goal of some of the researchers to produce an animated feature film using computer generated animation. Catmull determined that such a movie would require 130,000 frames, which would require billions of dollars worth of memory and processing power. Using Moore’s Law, Catmull estimated that the same technology would only cost $15 million by 1992, which was about as much as Disney typically spent on an animated feature.

Once the technology was ready, it was clear to the CRG that Schure was not the one to produce their movie. Even if it was visually stunning and technologically advanced, a movie with a poor story would only discourage other film makers from producing CG films.

There were rumors in the budding computer graphics community that a young animator at Disney, John Lasseter, was making mockups of a hybrid CG/traditionally animated film based on the popular children’s book, Where the Wild Things Are. Perhaps a major movie studio would be interested in the CRG’s research and expertise?

Unexpectedly, a Disney employee had dropped by the lab in 1978 to look at the painting system being developed there. Feature animation was suffering at Disney. Costs were rising, and experienced animators, many of whom had been with the company since the 40s, were retiring. Using a computer to ink cels would be a huge boon for Disney. Fewer experienced animators meant smaller budgets and shorter production times. The technology was still too coarse to use in production work, and Disney didn’t make any commitments.

Leaving NYIT

Unexpectedly, Francis Ford Coppola made contact with Alvy Ray Smith through his company, American Zoetrope. Coppola was in the throws of producing Apocalypse Now (1979) in the Philippines. Traditionally, a director would have each day’s film developed the next day so he or she could watch “dailies” and find out if any scenes needed to be reshot. This was impossible for Apocalypse Now, which was two weeks away from the processing labs in Hollywood and Europe. If a scene needed to be reshot, Coppola might find out about it weeks after the set was knocked down or after the entire team moved to a brand new location.

A new technology called video was beginning to approach film quality. A standard used for Japanese TV production, called HDTV, offered nearly the same resolution as ordinary film and could be stored on magnetic tape or Laserdiscs that would be watchable almost immediately after being shot. Coppola wanted to have the CRG work on creating a video-KEM, allowing him to edit the film immediately after it was shot.

A video camera would be integrated into the film camera so the director would have a video of the day’s shoots immediately. A computer would record the frame numbers and create a report after the film was edited on where to cut the negative. Ideally, video would eventually achieve film quality, and the negatives could be replaced altogether, but that would be decades away.

The project was exciting and would prompt a revolution in the industry in less than ten years, but Coppola was not nearly as reliable a patron as Schure. Zoetrope’s fortunes ebbed and flowed with the release and success of Coppola’s movies. Sometimes the studio would be flush with cash, and other times it struggled to make payroll. Add to that Coppola’s supposedly erratic personality, and the CRG was not interested.

Star Wars

Star Wars was released in 1977, and it was a smash hit. Unlike other science fiction films (such as 2001: A Space Odyssey, which had been released almost ten years earlier), outer space in Star Wars wasn’t cold and inaccessible. Space ships were small, gritty, and accessible, like the Millennium Falcon and the X-wing Fighters, not like 2001’s Aries lander, Discovery One, or Leonov, all three of which were designed in consultation with NASA engineers.

Star Wars was a smash hit, and George Lucas was loaded with money and ideas for the film industry. He started work on a new headquarters for his film making company, Lucasfilm Ltd., in the outskirts of Marin County. The property was to embody the “blue jeans Victorian” aesthetic of the company. Even though the ranch was to be state of the art and completely new, it was built in a variety of old styles. The main house, Lucas’ home and headquarters, was a huge Victorian decked out with a wraparound porch and stained glass.

Besides his ranch, Lucas was making investments in the future of Star Wars. The dogfights in Star Wars were not only tedious to make, but also very expensive. Since every ship had to be manually composited on top of every frame, a very large film format was required. Lucas had stumbled upon VistaVision, a film format last used in Bing Crosby films. Most movies were shot in 35mm, but VistaVision was 70mm, capturing far more detail. Dealing with the film was a pain, one that could be alleviated somewhat by computers. George Lucas was also subject to the ebb and flow of movie work, but Star Wars had blinded the CRG, and they didn’t care.

George Lucas

Nobody at Industrial Light and Magic, the company that Lucas had started to create the effects for Indiana Jones and The Empire Strikes Back, thought that computer animation could replace the models of space ships or even create characters, but they recognized that computers could eliminate some of the tedium of compositing space ships on top of one another. Bob Gindy, Lucas’s real estate developer, was tasked with finding a group that could make video editing a reality, especially in compositing. When Gindy asked around University of Utah and other universities known for computer graphics, Ralph Guggenheim’s name kept coming up.

When Gindy called Guggenheim, he introduced himself as George Lucas’s “head of development” and asked if the technology at the CRG could help Lucas produce his special effects. Guggenheim suggested that he talk to Catmull and Smith. As an afterthought as he was hanging up, Gindy asked, “Can you make a spaceship fly around on the screen?” Ralph replied, “Sure, we do it every day!” (The Second Coming of Steve Jobs, 133)

Schure didn’t lay out a timetable for any new products and didn’t hesitate from outfitting the lab with state of the art equipment. The only thing he expected in exchange for his generosity and the independence he granted the lab was loyalty. Schure once found Jim Clark’s résumé and letters to employers on a CRG mini and went ballistic. The two men, according to onlookers, almost came to fisticuffs.

Despite Schure’s generosity, Catmull and most of the other researchers at the CRG were resigned to never creating a compelling animated film. George Lucas had the creativity and experience in film making that Schure didn’t, and more importantly, had created Star Wars, which had a small CG sequence. It was decided (at the team’s official residence, not in the office) that the CRG would move to Lucasfilm, where it would work on video editing and creating small special effects for Lucas’s movies. Catmull typed out a letter on a borrowed typewriter (they didn’t dare use a computer that Schure might be able to search) to George Lucas and explained that they would move.

In order to avoid the ire of Schure that Clark had weathered and to avoid hurting his feelings (the team, for the most part, genuinely liked Schure), the twenty or so CRG members who would leave for Lucasfilm but would not take jobs immediately. They would quit the CRG over the course of several months find “intermediary” employers, and after a year or so would move to Lucasfilm, eliminating any of Schure’s suspicions.

Joining Lucasfilm

LucasfilmComputers played a role in some of Lucas’ first Star Wars movies. The most expensive scenes were the space dogfights, with tens of spaceships in a single frame. In 20 years, these fights would be produced almost entirely on computers, but in 1976, they were produced using very large film and models on tracks. What would become Industrial Light and Magic shot each spaceship separately, compositing them together manually by layering the frames on top of each other.

Death Star from Star WarsA computer was used to record and recreate the movements of the camera across the scene so the shots could feature non-static camera positions. The earliest computer generated effects also made an appearance in Star Wars. Larry Cuba created the vector diagrams of the Death Star featured in the rebel briefing before Luke Skywalker destroyed the space station.

These vector animations paled in comparison to what was being done in the cradle of computer animation, NYIT.

Skywalker Ranch would not be done for several more years, so Lucasfilm was sprinkled throughout California. Industrial Light and Magic (ILM) was based in Los Angeles, since ILM was expected to work with outside studios. Lucas was fond of Northern California and did most of the post-production work in a complex of anonymous bungalows outside San Francisco. Catmull had left for Lucasfilm directly and had set up an office in San Ranselmo, an anonymous industrial town in Marin County, and shared the space with George’s wife and lead film editor, Marcia.

Lucasfilm was still very much dependent on the cycles of show business. Lucas didn’t have enough money to finance Empire Strikes Back on his own. He was forced to find outside financing, though he insisted on complete independence. Despite the money issues with the rest of Lucasfilm, the Computer Division (the relatively uninspiring name that was given to the scion of CRG), was relatively well funded, though certainly not as well funded as the CRG. Instead of the half dozen state of the art frame buffers capable of creating film resolution images, the Computer Division was stuck with relatively low-end equipment. For example, an early competitor turned defense contractor, Triple-I, used a Cray to render images while the Computer Division used a VAX.

In the course of a year after the exodus from the CRG, which still existed under new leadership, the Computer Division had over 40 employees and had a number of projects. The most ambitious – and perhaps most important to the film industry – was Guggenheim’s groundbreaking work on video editing. Other projects included the computer graphics tools that had to be rewritten after the departure from NYIT and a project management system for ILM’s complicated special effects process.


The animators were not complacent after their move. Though Lucas would not include any significant computer animation in his upcoming Star Wars releases, Catmull and Smith would not be deterred. Shorts like “The Road to Pt. Reyes” and stills from The Works demonstrated the technology of computer animation but they were not examples of good film making. To impress Lucas, a shot would have to tell a story and not be hindered by technology.

Genesis device animationILM was working on creating the visual effects for Star Trek II: The Wrath of Kahn (1982). The villain, Kahn (pronounced KAHN!!!!!), captures a weapon called Genesis that “resynthesizes” all organic molecules on a planet and starts a new life cycle. A computer simulation of Genesis took place on a desolate, uninhabited planet. The scene would not be feasible or realistic using matte paintings and scale models, so ILM came to the animators.

The scene was ideal because a high degree of realism would not be necessary. The simulation was taking place on a computer, so lower resolution and lower frame rates would actually be desirable and would also make the project a lot more feasible. Smith directed the project and integrated as many of the obscure skills as the team had. Somebody on the team was interested in gametes and was responsible for creating a missile that looked like a sperm cell. Another worked on creating a camera track that would be impossible in the real world, one that would grab Lucas’ attention.

Besides the composition and visual effects, the very act of transmitting the images from the computers to film was a feat. A special laser scanner was developed by a team headed by David DiFrancesco to capture the contents of a very high resolution (greater than 2000 x 3000 pixels – 6 megapixels) CRT one frame at a time.

The final shot was less than a minute long and was a tour de force of all the techniques the computer division had developed. The actual resynthesis of organic matter into new life was accomplished using the particle system created by William T. Reeves, originally to make random and realistic looking flames. The final result was a little artificial, however. As rivers and oceans formed, the scene transitioned into a matte painting of a jungle scene.

Star Trek II netted good reviews (Roger Ebert gave the film a 4/5), but Smith received an impressive endorsement from Lucas. After the movie had been screened for him, Lucas stuck his head in Smith’s door and said, “Great camera move.” (The Second Coming of Steve Jobs, 99) and walked off. The continued survival and funding of the Computer Division and the animation team was suddenly much more secure.

The first Pixar product was born around the time of Genesis. In order to eliminate the need for a film scanner, a system needed to be created that could output computer generated images directly to film. Smith started talking up the device, which was tentatively named the Pixar, at the 1983 SIGGRAPH conference. The Pixar would be a framebuffer and rendering farm in a single box. The Computer Division finally had a potential product on their hands with a possibly huge market. Not only would filmmakers be interested in such a device, so would physicians, engineers, and any other film where 3D visualizations would be useful.

John Lasseter

The transformation of what would become Pixar from a research and development group into an animation group began with the debut of Star Trek and was solidified by the hiring of John Lasseter from Disney. Disney’s animation studio was in the throes of a personality struggle between young animators and executives. The studio had largely lost its golden touch of the 30s, 40s, and 50s seen in films like Snow White and the Seven Dwarfs (1937) and Cinderella (1950). In the 70s and early 80s, movies like The Rescuers (1977) and The Fox and the Hound (1981) received lukewarm reviews and paltry box office returns.

tron-1982Lasseter joined Disney as a character at the Disneyland theme park in Anaheim, California. Eventually, he worked his way into animation, where he played a role in Tron (1982), the most notable use of computer graphics (even though Star Trek II predated it). Disney had long used a multiplane camera where partially transparent cels were stacked on top of each other and moved at different speeds, providing a sense of depth. Lasseter believed that the same technology used to create the computer world in Tron could be used to create 3D backgrounds that would supplant the multiplane camera.

Lasseter presented a proof of concept short film based on the popular Maurice Sendak book Where the Wild Things Are that used traditional cel animation in the foreground but a computer generated background. Lasseter became active in the computer graphics industry. During a computer graphics conference on the berthed Queen Mary in Long Beach, Lasseter met Catmull and Smith. They were united by the same goal: using computer graphics in a feature animated film. Due to political infighting, Where the Wild Things Are was never pursued, but Lasseter would retain his contacts at Lucasfilm.

After Disney passed on Where the Wild Things Are, Lasseter, Glen Keane, and Thomas Wilhite began work on The Brave Little Toaster using the same computer generated background technique. Unfortunately, The Brave Little Toaster was canceled, though it would later be revived as a standard cel animation movie.

Disgusted by the project’s cancellation, Lasseter quit the studio, intending to come back in less than a year, and started playing around at the Computer Division at Lucasfilm.

The Adventures of André and Wally B.

Adventures of Andre and Wally BLasseter was not an engineer and offered helpful advice on making some of the Division’s software more intuitive. Especially Tom Duff’s Motion Doctor, which had turned into something of an animation studio. Objects could be modeled and then manipulated or bent, the basis of computer generated animation today.

After a few weeks, Lasseter had become a presence at the Computer Division. He was the type of person who would actually use the software that the engineers were working on. On top of that, he was a sociable guy who was fun to be around.

Alvy Ray Smith began work on a short totally computer generated animated film and recruited Lasseter to animate it. Eventually named The Adventures of André and Wally B. after the popular 1981 movie, My Dinner with André, the short centered around André, who was awoken by a friendly bee and then flees. This upsets the bee, Wally B., who pursues André and stings him.

The story was less important that the technology it showcased. The same particle system that was used to create the random looking flames in Genesis was adapted to create thousands of trees and blades of grass that moved independently of each other and were unique. The characters themselves were modeled out of teardrops, cones, and spheres, a technique that would be used with later films.

David DiFrancesco was working on transmitting computer images to film while avoiding jaggies or rough edges. This not only involved working on the laser scanner but it also included the task of “smoothing” edges. The solution that he had implemented involved randomly sampling a single point on the edge of a shape and coloring the entire pixel. This made the border fuzzy, which was not noticeable in film, since it refreshed 24 times a second.

DiFrancesco had won a lot of industry attention for his solution, which he called motion blur, and he periodically had guests over at the Computer Division offices to see the rendering and scanning in person. Several Disney executives had been invited and were impressed with the process and the actual animation (even though DiFrancesco only had a few frames to work with).

Smith wanted to submit The Adventures of André and Wally B. for consideration in the shorts contest, so the team only had a few weeks to render the short (which was created using wireframe models without lighting or texture). Such a task would have been impossible just using their VAX minicomputer, so the Computer Division glommed computing power from anybody they could find. All five minicomputers at Lucasfilm (named for planets or moons from the Star Wars universe) were churning away.

Cray had long tried to sell a Cray to Lucasfilm. Even if the Computer Division would be able to rent it off to other groups when it wasn’t being used, the Cray would have been too big an investment to justify. Catmull was able to convince the salesman to let the Computer Division to use a “floor model” for free. The Division also used all 50 computers in the brand new Project Athena network, which was supposed to create a network spanning the country.

Andre and Wally BThe Adventures of André and Wally B. was completed in time to exhibit at SIGGRAPH, and the entire team was hoping for a splash. The Computer Division had no new scenes to exhibit at the 1983 SIGGRAPH conference (one year after they had shown the groundbreaking Genesis scene), so they showed the wireframe model of the Death Star from Return of the Jedi and were widely derided for their weak showing.

Young Sherlock HolmesThe Adventures of André and Wally B. received standing ovations from the attendees. It was the darling of the show and once again buoyed the position of the Computer Division within Lucasfilm. John Lasseter joined the Computer Division permanently and started work on the Divisions second major special effect for the film Young Sherlock Holmes (1985).

Smith was once again the director for a Paramount film. ILM used standard stop motion techniques for all but one scene. After a priest is given a hallucinogen, he sees a knight emerge from a stained glass window and follow him into the street. The scene was complex and difficult for two reasons. First, the monster interacted with other characters and the set. Not only did it walk around, but it was partly translucent. Secondly, beside the translucency, because the monster would need to look as realistic as possible, the scene would have to be near-film quality. This would be taxing at the lavishly equipped NYIT, but even more so at the somewhat more austere Computer Division, even with the growing number of Pixars at their disposal.

The solution to the resolution problem was using more than a single frame buffer to render each frame, effectively dividing each frame into more than one image. The final result was a highlight of the film, and it outshone many of ILM’s effects. The film also received good reviews and won attention from people around the budding graphics industry through the Computer Division’s frequent contributions to SIGGRAPH, which published a peer review journal and in addition to the annual conference, in the development of the effect. Especially noteworthy was the motion blur, which made its debut in the effect.

The software used to create the effect and The Adventures of André and Wally B., MotionDoctor, was also becoming a star. Lasseter had crossed a milestone: He had become adept at using the software, which would later be renamed Marionette. Ordinary animators who understood the motions of animation (organic looking movements) would be able to use the software with a minimum of training, quite an improvement over just a few years before.

Leaving Lucasfilm

Lucasfilm was in something of a funk. It’s resources were taxed heavily by the construction of the high-tech Skywalker Ranch and research projects like the Computer Division, but also SoundDroid and EditDroid (a sound and nonlinear video editor using Laserdiscs, respectively) and lots of films that George Lucas was producing (he was a hands-off producer; he mainly financed production). At a company meeting in 1983, Lucas announced that he and his wife and long term editor Marcia were divorcing. In California, this meant that she would be entitled to half of the family’s net worth.

Money was tight at Lucasfilm, and even though Lucas had been impressed by Genesis, The Adventures of André and Wally B., and the monster from Young Sherlock Holmes, the Computer Division did not come close to covering costs, much less turning a profit. Lucasfilm CEO Robert Greber was instructed by Lucas to start looking for a buyer for the animation group in the Computer Division.

Pixar’s Expanding Vision

Smith had postulated that medicine might be a lucrative market for the Pixar. Images could be modeled in 3D and used to teach or even as a prep to surgery. Bob Drebin, one of the engineers in the Computer Division, got his hands on a CAT scan of a pelvic bone from a medical journal. He then modeled it in a relatively simplistic way. He assigned different colors and textures to each shade in the CAT scan (which are essentially cross sections). Then he defined each region as a layer that could be peeled away to reveal lower regions. The resulting model won attention for the Computer Division and the Pixar from the conventional press, quite a step up from SIGGRAPH and Creative Computing. Medical journals even ran papers on the Pixar and the pelvic bone.

Around this time, the Computer Division found its largest Pixar customer, Disney. Catmull started holding meetings to discuss possible uses for the paint systems he had helped develop, including feature animations. The Disney demonstration in 1978 had not been forgotten. Disney was convulsing under a huge leadership change with the installation of Michael Eisner as CEO.

Disney was now ripe for computer animation. Catmull and Smith started drafting a proposal to start using a Pixar-based paint system to ink hand-drawn cels. Disney was very interested and the system was named Computer Animation Production System. After two years of development and negotiations with Lucasfilm, a deal was reached and Disney bought dozens of Pixars.

Greber started looking for interested parties. His initial price was $15 million for total ownership of the Computer Division and all of its patents. After the investment, the Division would need another $15 million to develop the Pixar into a marketable product. Nobody bit. The fact that Lucasfilm wanted to sell its stake in the Division was hardly a ringing endorsement.

Steve Jobs had been given a tour of the Computer Division’s offices but balked at the price. However, he indicated that wanted to know if it dropped below $10 million.

Talks began with GM and Siemens over turning the Computer Division into a joint venture between the three parties. Phillips was one of the largest medical equipment suppliers in Europe. Presumably, the Pixar could be integrated into CAT scan machines or other imaging devices. GM could use the Pixar in its engineering division. Ross Perot was one of the major champions for GM buying a stake in the Computer Division. His EDS, now a subsidiary of GM, would be able to sell the Pixar as part of an engineering workstation in addition to using it internally.

The talks got as far as a contract being drafted between all three parties. Each would have a $10 million stake in the new company. Unexpectedly, there was a palace coup at GM that removed Ross Perot from its board of directors and from all meaningful involvement in the EDS, which he had founded. GM wasn’t interested in Perot’s pet project and ended negotiations. Phillips wasn’t going in alone and also left.

Desperate Times

COO Doug Norby returned to Jobs hat in hand. It was now nearing the end of 1985, and Lucasfilm had spent a year on a deal that fell through and was still spending money on the unprofitable division. After conceding rights to all of the short films that the Computer Division had created to show off their hardware, including Road to Pt. Reyes, The Adventures of André and Wally B and Luxo Jr. The final price for the Division, which was renamed Pixar, Inc., was only $5 million, a quarter of what GM and Phillips were willing to put in, plus another $5 million to keep the company going until the Pixar (later renamed the Pixar Image computer) was released.

The documents were signed, and Jobs became chairman and CEO of the new company. Catmull and Smith occupied the only other chairs on the board and were the only two employees to receive stock options.

On February 3, 1986, Pixar became an independent hardware company. The Pixar Image Computer was released shortly after Pixar was incorporated and was tepidly received. Eventually, Pixar would have to lay off most of its sales staff (which went to medical conventions around the country).

It wouldn’t be until 1988 when Pixar started producing award winning commercials to offset the losses in the hardware market. CAPS would be used in 1988 to ink a scene in The Little Mermaid, and in 1994 Toy Story was released as the world’s first totally computer generated feature film.

For more on the evolution of CGI, see The Pixar Story: Dick Shoup, Alex Schure, George Lucas, Steve Jobs, and Disney.


  • DroidMaker by Michael Rubin
  • The Second Coming of Steve Jobs by Alan Deutsch
  • Dealers of Lightning by Infotrac Reference Center Gold (for periodicals)
  • Wikipedia

A Word on Sources

Since I had no interviews or other first hand accounts of the doings at NYIT, PARC, Lucasfilm, or Pixar, I have relied on the works of others – especially DroidMakers by Michael Rubin. Not constrained to computer animation, Rubin chronicled the research that Francis Ford Coppola and especially George Lucas funded into nonlinear video editing and digital sound. DroidMakers is a must read for anybody interested in George Lucas or video.

Keywords: #cgi #pixar #disney #lucasfilm #stevejobs

Short link:

searchword: originofcgi

One thought on “CGI Story: The Development of Computer Generated Imaging