Pixar, a company that revolutionized the feature film industry, had an obscure origin. A group of researchers from the most elite research institutions in the US eventually gathered at a former diploma mill and later defected to Lucasfilm. Their division was purchased by Steve Jobs and became Pixar, which created Toy Story and is now owned by Disney.
Dick Shoup, who would later pioneer digital video, had a PhD from Carnegie Mellon and worked at the Berkley Computer Corporation (BCC). One of the early players in the timesharing market, BCC went belly up, and Shoup emigrated to Xerox's Palo Alto Research Center (PARC), where he got a reputation for being eccentric. In order to make the most use of his time, he developed a technique for riding his bike from his house to the desk without dismounting once.
On April 10, 1973, while many of his colleagues were at the debut of the Xerox Alto prototype hardware displaying a picture of Sesame Street's Cookie Monster, Shoup was obsessing over digital video. A floor below the other engineers, Shoup was seated in front of a black & white video camera hooked to one of the lab's minicomputers recording the first ever digital video. He had created a simple frame buffer, revolutionary in its day and indispensable today.
The early systems that he used had hundreds of thousands of dollars of memory, making the machine much more expensive than any of the workstations in the Alto family. Xerox didn't exercise much discipline at PARC, so even though Xerox would never release digital video equipment, Shoup still received his funding.
From there, he expanded the frame buffer so it could handle color and grab footage from any video source, such as LaserDiscs or videotapes. Once the computer had the video in memory, he could modify it as though it were a standard image. As Shoup added more and more editing features to the frame buffer, he named his program SuperPaint.
The head of the computer scientists at PARC, Bob Taylor, wanted to make the Alto into an integrated office system, something that Xerox would be able to market to its existing customers. He envisioned entire offices running on ethernet networks of Altos, laser printers, and file/email servers.
SuperPaint didn't fit into this system. Even though it did have a market (video production and design), it didn't fit into the market that Xerox wanted to focus on.
A year later, Shoup was making progress in optimizing SuperPaint, but he had little support for the project until Alvy Ray Smith arrived at his house.
Smith was a brilliant mathematician and economist who wrote his PhD thesis on automata theory. He became an associate professor at New York University, but he lost interest in teaching after a serious skiing accident left him immobilized for a few months.
After he recovered, Smith started traveling the country in his Ford Torino and ended up in southern California, where he eventually got a teaching job at Berkeley. The Berkeley library didn't have a book he needed, so he went to Stanford for the weekend, where he bunked with his old friend, Dick Shoup, who gave him a tour of PARC. He saw the first workstations in history, the Xerox Altos running SuperPaint.
Smith was floored by SuperPaint. He was an amateur artist and was immediately enamored with his newfound ability to create full color images with a computer, something that was impossible (to his knowledge) outside of PARC.
"I realized this is what I had come to California for" ( Dealers of Lightning, 234), Smith said of SuperPaint.
Shoup was enamored with Smith. Not only would he have a companion in the graphics lab, he would have a tester (with a fresh pair of eyes) and an advocate for SuperPaint. PARC's personnel department refused to hire a resident artist like Smith, so Shoup got a purchase order, as though Smith were equipment, to make sure that he was paid for his efforts.
Shoup was able to construct incredible video sequences that were impossible to create in any other medium. He took an image of his girlfriend's face and transformed it into a kaleidoscope of colors that was still very organic in nature. He created a gecko's hide that cycled through colors and modified Star Trek episodes with new psychedelic effects.
A small subculture developed around SuperPaint as Smith invited his friends to come and watch his more and more abstract video clips. After Smith and Shoup submitted their work to local PBS affiliate, KCET, as a segment to a program called Supervision, Taylor lost it. He demanded that Shoup delete every frame that referenced Xerox in any way, which took hours on end.
Days later, Smith was dismissed and SuperPaint was dismantled. PARC would focus squarely on black and white. However Shoup had the last laugh - in 1983 he was awarded an Emmy for his work with SuperPaint.
Without SuperPaint, Smith started looking for another outlet for his creative energy. He heard from one of his former associates at Berkley that the University of Utah had recently acquired a frame buffer, the core component of SuperPaint. When Smith arrived in Utah, he was dismayed to find out that he was too late. A millionaire from New York, Alex Schure, had already bought every computer graphics component and tape.
Schure made his fortune at the New York Institute of Technology (NYIT), a rumored diploma mill he created to take advantage of the wave of GI Bill students after World War II. NYIT eventually became an accredited school and occupied a building near Lincoln Center in Manhattan. The school still didn't have much of a reputation, though, and the college's bookstore stocked car parts instead of textbooks.
Word of Smith's arrival in Utah found its way back to Schure, who invited Smith to come to New York to help him with his computer research. When Smith walked into Schure's garage (the computer research group operated independently of NYIT), he was greeted by Ed Catmull, who had a PhD in physics. In his teens, Catmull had created the first computer generated film, a rotating model of his hand, and he was recruited by Schure shortly after he bought the University of Utah's equipment.
Catmull explained that Schure wanted to create the first computer generated feature film and was willing to spend as much money as it took. The researchers thrived in Schure's environment and made huge leaps in the field. There were fewer than a dozen people in the group, but they were already able to create very complex computer generated shapes, especially compared to what Smith had been able to do at PARC.
Schure's researchers were pretty much given free reign over the work they chose to do. Some began work on a system that would accelerate the creation of conventional, hand-drawn animated films by digitally tracing the lines and then filling them in with virtual paint. The system would eventually turn into Disney's CAPS.
The rest of the researchers worked on a vague project called "The Works". With no outside help or guidance, the researchers accomplished little. Most of their efforts were directed towards creating sample animations to define the characteristics of the characters. It was unclear to many of the researchers if there was even a script for the project.
At the same time the researchers were working to create animation for decades in the future, Schure was financing a group of conventional animators to create his own animated feature, Tubby the Tuba. Schure fancied himself as the next Walt Disney, but he lacked the taste to create films like Disney did; Tubby was filled with technical flaws and an incomprehensible story line.
The researchers were aghast at Schure's effort and decided that they would have to find another sponsor for their work, someone who could bring their ideas to fruition. Almost by accident, the members received calls from George Lucas and Francis Ford Coppola.
Francis Ford Coppola was at the zenith of his career. He had been awarded academy awards for The Godfather, The Godfather Part II, and Apocalypse Now, and he had produced American Graffiti (which was directed by George Lucas).
Coppola wanted to start integrating computer generated effects and computer based video editing (which required frame buffers to display frames) into films produced by American Zoetrope, his company. Catmull traveled to San Francisco for a number of meetings with the director, but he thought that Coppola's profuse drug use would impede any progress the researchers would be able to make.
The researchers had all played hooky from NYIT on May 25, 1977 to watch Star Wars, and they were thrilled with the film. The special effects were not spectacular (though the compositing required for space dog fights was impressive) but the story was great. They believed that George Lucas would be receptive to adopting computer generated effects to eliminate the tedious work of drawing in light saber blades and moving flotillas of spaceships.
Production of the dogfights was very labor intensive. Each model of ship had to be built with meticulous detail, since 70mm film would catch any imperfections. The models were placed on tracks and would make a single motion. The process would be repeated for every shot and every ship, and then they would all be composited (cut and pasted) on top of a background shot. The process could take weeks or months for a few minutes of footage, though the results were very good.
Catmull was correct. That year Lucas contracted with a computer company, Triple-I, but he decided to use conventional methods instead. The company succeeded in producing a digital model of an X-Wing fighter, which made its way to the cover of an influential computer magazine. Carmull and Smith were incredulous however. The model had 'jaggies' ,or pixelated, edges that made it look unrealistic. Even though the effects never made it into the movie, Triple-I proved to Lucas that computers could be useful in film production.
Smith was tapped to go to Lucas' home town of Modesto, California, to make the pitch to Lucas' real estate developer, who was receptive to the group. All of the researchers were invited to come to Lucas and start a brand new computer division which would work on digital video and digital models.
In order to avoid angering Schure by a sudden exodus to a possible competitor, the researchers slowly migrated to the computer division after several months at temporary jobs. Catmull was the first to arrive at the computer division and was given control of the team. By 1980, the rest of team was reassembled and helped produce the special effects for Return of the Jedi.
Not all of the projects the computer division embarked on were as widely seen as the Death Star map. The computer division's most significant contribution was actually project management software that ran off their minicomputer, named Dagobah. The software helped automate the intricate process of shooting scenes in space.
Other than the light sabers and space ships, the team did very little until Paramount Pictures came to ILM (Industrial Light and Magic, the new name for the computer division) and requested that it create a scene for Star Trek II: The Wrath of Kahn. Meyer wanted to create a scene that depicted the destruction and reintegration of a planet called Genesis (named for the weapon used to destroy the planet). ILM assigned Catmull's team of computer researchers to generate the sequence using computers.
Genesis was simplistic by today's standards. The space ship that destroys the planet looks like a sperm cell, and the planet is little more than a plain mountainous rock (not coincidentally, those two effects were particular strengths of the researchers). The interesting part was the camera work, which would have been impossible with conventional filming techniques. Genesis was one of the most popular parts of the entire film and made an excellent impression with George Lucas, who began utilizing the team for other special effects.
Disney also took note of Genesis. Michael Eisner, who had recently been hired from Paramount to rebuild Disney's struggling animation studio, asked that Catmull develop a system that would trace and color each animation cel and replace the team of women who did the boring process by hand. Catmull's team enthusiastically embraced the idea and developed the Computer Animation Production System (CAPS) for Disney, though the company would not admit that it used computers for almost a decade.
CAPS was not a trivial development. Not only did the researcher's have to create a SuperPaint-like interface to allow the animators to fill in the hand drawn cels, but over a hundred thousand high resolution images would have to be stored for a single feature film, impossible on all but the most powerful computers of the day. Disney paid millions of dollars for the storage system itself in addition to the researchers' painting program.
The system used for CAPS was the precursor to the Pixar Image Computer, which until 1987 was the primary breadwinner for the computer division. The PIC wouldn't be formally released until later; it essentially acted as a large frame buffer that was capable of handling film resolution images (roughly 2,000 x 3,000 pixels) and rendering simple models. Each PIC cost over $100,000 and was dependent on a Unix workstation.
Even without the outside attention, Genesis was a huge accomplishment for the researchers. RenderMan, the software that turned the simple models the animators worked with into high resolution images, was written by Loren Carpenter and was originally named Motion Doctor. RenderMan would form the basis for every project that the researchers would undertake for the next two decades and allowed traditional animators to begin working with computer animation with relatively little training.
Another key element to the researcher's future materialized in 1984. John Lasseter, who had worked on Disney's Tron (a film with almost 30 minutes of computer generated animations) and a failed project to create a computer animated version of the storybook Where the Wild Things Are, joined Lucasfilm at Catmull's urging to experiment with computer animation for at least a few months.
Lasseter immediately dived into creating a computer generated short film for his son, titled The Adventures of Wally and Andre B. The plot centered around a toddler who is stalked by a bee in a large forest and eventually stung. The forms were fairly complex for computer animation. The human face was capable of conveying simple emotions, and Andre had complex movements.
The real advance in the film was its backdrop, which featured a great deal of detail. The programmers had created a program that randomly generated the movement of every leaf, blade of grass, and branch in the film, eliminating the need for animators to define every element in the film.
Though the completed film terrified his son, The Adventures of Wally and Andre B was a boon. The film garnered much recognition for the computer researchers and would rekindle Disney's interest in computer animation after the failure of Tron and the early death of Where the Wild Things Are.
Catmull's researchers would have been satisfied to stay at ILM creating special effects until the technology required to create a computer generated feature film became affordable, but Lucas was hard up for money. He was in the midst of a divorce, and by California law he and his wife held all property jointly, so the couple would have to split the family's net worth evenly.
Instead of giving her any of his lucrative movie companies, Lucas decided to give her cash. The researchers seemed to be decades away from being profitable. Outside of Disney and small video clips, there seemed to be no market for its expensive products.
As news of Lucas' pending divestment of the research division got out, buyers started stepping forward. A consortium consisting of the Ross Perot-founded and GM-owned EDS and Phillips got as far as drafting the contracts, but Ross Perot was pushed out and the deal died.
Luckily, a wealthy investor had heard about the researchers and was interested.
Alan Kay had maintained a relationship with Steve Jobs after he and his entourage from Apple had toured PARC. In 1984, Kay had told Jobs about Lucas' band of researchers, but Jobs was uninterested. He was still Apple's chairman and head of the Macintosh team, and he had no plans to leave - especially after the successful LaserWriter launch.
Despite his disinterest in acquiring ILM, he decided to take a tour and see what was going on - and he was blown away. He saw the most sophisticated computer graphics in the world, stuff that put the Genesis sequence to shame.
Unfortunately for Jobs, Macintosh sales were only a fraction of what he had predicted, and he was forced out in 1985 after his failed coup attempt against John Sculley (see The NeXT Years: Steve Jobs before His Triumphant Return to Apple). Jobs sold all of his Apple stock almost immediately after leaving and started a company to create computers for higher education, NeXT Computer Inc.
With loads of cash left to burn, Jobs was contacted by Catmull and Alvy about taking a second look at acquiring the division. They had a meeting at his Spartan mansion in the exclusive neighborhood of Woodside, and Jobs immediately agreed to buy the division, which he spontaneously named Pixar.
Lucas had originally hoped to get upwards of $100 million for the division and associated intellectual property, but Jobs played hardball and got the price down to $10 million. Around 5% of the company would be given to Catmull and Smith, but Jobs would maintain the lion's share of the company, though he would have limited control.
The biggest issue was geography. Modesto was almost two hours away from Jobs' home outside Palo Alto, and he was unwilling to drive down unless he had to (in fact, he visited Pixar's headquarters fewer than six times before 1994). Secondly, he was in way over his head with computer graphics. He was no engineer, although he had a decent understanding of the inner workings of the Macintosh, he knew next to nothing compared to the PhDs at Pixar.
The first product that Pixar shipped was the Pixar Image Computer, which was essentially the storage system of CAPS. The device worked as a peripheral to an SGI (which was being run by Schure alumnus Jim Clark) or Sun workstation and allowed images to be stored and displayed on a computer screen.
Priced at $135,000, the device was incredibly expensive and seemed to have limited appeal. Pixar decided to push the device outside of computer animation, to doctors who needed to review high resolution copies of x-rays and MRI scans, and opened sales offices across the nation. Unfortunately for Pixar, only about a hundred of the devices were sold. Most of them went to Disney and the NSA, with a few going to large hospitals in the Northeast.
At the end of 1986, Pixar's future appeared bleak, but it was about to improve tremendously. Lasseter had created another short film, this one meant to demonstrate the merits of the Pixar Image Computer. Titled Luxo Jr, the film featured a pair of anthropomorphic Luxo brand lamps. The larger was an exasperated parent watching the smaller lamp play with and eventually pop a toy ball.
Lasseter chose the lamps for two reasons. The first was that they were convenient. His small team of animators all had desks outfitted with the inexpensive lamps, and they were relatively simple to animate. The second was the fact that the form demonstrated the strengths of computer animation. The simple lamps were able to convey complex emotions without having to create the enormously complex human face.
Luxo Jr was incredibly popular. Not only did it impress computer scientists, the film industry was impressed, too. The short was nominated for Best Short Animated Film at the 1986 Academy Awards.
Pixar now had clout in the film industry and would use it to its advantage.
With an Academy Award nomination under its belt, Pixar released its second product, RenderMan. Based on the work of Loren Carpenter while Pixar was still part of Lucasfilm, RenderMan was a moderate success in the marketplace. Jobs hoped that products like RenderMan would spark a consumer revolution, empowering untrained people to create computer animation on their home computer, the same way the Macintosh allowed untrained people to create professional-appearing documents like newsletters and flyers.
Unfortunately for Jobs, that revolution didn't occur. Other companies created animation packages based on RenderMan, like PixelPutty for the Macintosh, but none were designed for consumers to use. Pixar sold slightly more than 100,000 RenderMan licenses - phenomenal for a product with such limited consumer appeal. RenderMan won itself wide publicity and respect when James Cameron became a loyal user with his films, The Abyss and Terminator 2.
Revenues from RenderMan and the Pixar Image Computer alone would not be enough to make Pixar profitable, so Pixar began producing commercials and eventually a television show. Its most popular ads were for Life Savers candy and Listerine.
Pixar was still far from profitable, especially after the company began acquiring the workstations and servers required to making a television special. Pixar was so hard up for money that animators were reduced to begging for secondhand computers from their friends at Sun and SGI.
With a questionable financial future, Pixar pushed through the end of 1989. Lasseter's third short film, Tin Toy, was released and was awarded an Oscar for Best Short Animated Film. Tin Toy centered around a group of toys in an infant's room and was much more advanced than earlier productions. James Cameron also released The Abyss, which featured a RenderMan rendered character.
Disney was doing much better in the late 80s. The Little Mermaid was incredibly popular, the biggest release for Disney to date. To top that, a portion of the film was produced using CAPS.
Lasseter maintained contact with Disney, and contacted Jeffrey Katzenberg, the head of Disney's film division, about creating a Christmas television special. Katzenberg was so enamored with Pixar's shorts and software that he responded by offering to produce three films.
The staff at Pixar were floored. Catmull and Smith predicted that computer generated feature films would not be feasible until the mid 90s. Pixar was almost four years ahead of the curve. Even better than that, they had a deal with Disney, the undisputed king of animation.
Because of the changed focus of the company, there was a massive reorganization. All of the sales offices around the nation were shut down, and the Pixar Image Computer was discontinued. The remaining sales staff would focus squarely on licensing Pixar's RenderMan to other software makers.
The bulk of the company's resources were devoted to the creative division headed by Lasseter. Before that, there were fewer than a dozen animators working with Lasseter; that number doubled and then tripled during the production of the first film, Toy Story.
The bigger staff and investments in equipment made it necessary for Toy Story to succeed if Pixar expected to survive. The company had lost $60 million of Jobs' money at a time when Jobs was at an emotional low. NeXT was dying, with its loyal customers jumping ship to Windows NT and Sun workstations.
It seemed that if Pixar was unable to turn a profit soon, Jobs would sell it to a company like Microsoft (Alvy had become a fellow at Microsoft after leaving Pixar).
Steve Jobs flew to Anaheim to join Smith and Catmull during negotiations with Katzenberg to finalize the three film deal. Jobs had gotten into a heated argument with Katzenberg during a sales call months before, when he was trying to sell Disney a lab of his NeXTstations for the animators, so Catmull and Smith were nervous how the two would get along.
The initial deal gave Pixar a small percentage of box office sales in addition to compensation for the production of the film. Jobs demanded a bigger percentage on box office sales, a percentage of video sales, and unlimited software rights. Katzenberg was offended, and the two sparred for the better part of an hour.
Katzenberg eventually gave Pixar a bigger cut of box office sale (10-15% depending on how successful the film was) and rights to the software. Disney retained the rights to the films, which meant it could create sequels without Pixar. Disney also retained its hold on video sales.
The biggest boon for Disney was Pixar's production compensation. Katzenberg told Jobs that he never paid more than $15 million for an animated film, and it would be no different for Pixar. Jobs acquiesced and accepted the amount. (Katzenberg had lied - Beauty and the Beast had cost over $32 million to produce. It was only later that the Pixar executives found out that they had been snookered, but by then Katzenberg had left Disney for his own startup, the Paul Allen financed Dreamworks SKG.)
The employees at Pixar were optimistic about the future as Lasseter began writing the screenplay for Toy Story, but that would soon change. Jobs had never gotten along with Smith, and the relationship grew even worse during the negotiations with Disney. Smith was terrified that Katzenberg would walk out on the abrasive Jobs.
After Jobs snapped at Smith and yelled at him during a meeting at the Deer Park NeXT headquarters, Smith resigned from Pixar and started a brand new company, Altimira, which produced computer animation software. Altimira was eventually bought by Microsoft, and its technology was used in Microsoft's line of multimedia CD-ROMs.
Work on Toy Story was slow. Lasseter had no experience in writing screenplays for feature films and was struggling to create endearing characters. The story centered around a toy cowboy, Woody, and his conflict with a new space doll, Lunar Larry. Lasseter eventually put together a 30 second trailer to show off the characters and the animation technique.
When Lasseter showed it to Disney executives, they were stunned. The footage was unlike anything they had produced. Even the critics in Disney's animation division, the ones who had the most to lose if Pixar succeeded, were enamored with the technique, if not the story line. Disney approved the script, and real production began on January 19, 1993.
The animators began work immediately, creating the models that they would manipulate in the other scenes. Woody had over 723 motion controls to animate his actions. In his mouth alone, there were 58 such controls. That type of detail required an incredible amount of computing power. It took hours for 117 Sun and SGI servers to render each frame, and the film had hundreds of thousands of frames. To accelerate the process, the animators turned their high powered workstations over to the renderfarm at night to accelerate the process.
Toy Story's voice talent was phenomenal. The untested firm had no trouble signing Tom Hanks and Tim Allen for little more than union voice actors after they were shown the short trailer. For the soundtrack, Lasseter personally courted the popular Randy Newman to write and perform the songs.
Everything seemed to be going well. Even Jobs, the absentee owner who continued to devote his attention to the flagging NeXT, took an interest in the film's progress. He wanted to be kept abreast of Toy Story's production, so Lasseter and producer Ralph Guggenheim made a biweekly trip to NeXT's headquarters (which had recently moved from the lavish Deer Park offices to lower cost space nearby) where they would show him clips and pretend to accept his advice.
Ten months after production moved forward, disaster struck Pixar. The Disney executives stopped production and threatened to cancel the deal if Lasseter didn't fix the script. The problem was that Woody came across as caustic and mean compared to the renamed Buzz Lightyear. That day, November 19, became known as Black Friday at Pixar.
The employees were scared that the halt meant Disney was reneging on its agreement, but executive producer Bonnie Arnold knew better. She was an industry veteran and knew that distributors like Disney were likely to shut down production for the smallest problem. She enrolled Lasseter in a screenwriting class, and, in the interim, devoted some of Pixar's 110 employees to commercial projects.
By April 1994, Lasseter had reworked the script to make Woody more endearing. Lasseter added scenes to the film before Buzz Lightyear arrived that showed Woody as a sympathetic, paternal leader of the toys.
Katzenberg started production again, and Pixar started working towards a Thanksgiving 1995 release date.
The finished film filled over 1,000 CD-ROMs with 110,000 individually rendered frames. Each frame had taken hours to prepare and render, but the film was still less expensive than Disney's other big animated film that year, Pocahontas. What's more, Pixar had only 27 animators on staff for the entire company, including Toy Story and commercials. Comparably long traditional animated films required at least 75 animators.
Jobs still wasn't terribly interested in what was happening at Pixar, but as November 22 neared, he perked up. At a Disney gala in New York, hosted by Rudy Giuliani, Jobs was witness to the dramatic reactions to Pixar's stunning work. He soaked up praise as the night wore on.
Several weeks later, Jobs took the title of president from Catmull. Catmull had never been interested in being president, and it signaled Jobs emergence at Pixar.
Disney spent $100 million on promoting Toy Story. Buzz Lightyear cutouts were in every Burger King, and trailers littered television and preview reels around the nation. The staff at Pixar was nervous of the amount of cash Disney was sinking into the project, fearing that it might flop.
They had little reason to worry.
On opening weekend, Toy Story earned $39.1 million, enough to recoup the production costs. By the end of its theatrical release, the film netted over $200 million in box office receipts, an incredible number for an animated feature.
Pixar became a household name and had proved that computer generated films weren't just possible, but profitable.
One week after opening day, Pixar went public with a stock price set at $22. The price soared to $50 on the first day of trading, turning Jobs into a billionaire. Pixar went on to renegotiate its contract with Disney and released six blockbuster films that made well over $1 billion.
In 2004, Pixar declined to renew its contract with Disney and appears to be poised to distribute its films on its own. On January 24, 2006, Disney announced that it was acquiring Pixar for roughly $7.4 billion in Disney stock. This made Jobs Disney's largest shareholder and turned Pixar into a wholly owned subsidiary of Disney.