The gift of the MAGI. (Mathematical Applications Group, Inc.) John Andersen.
Before the letters MAGI came to be the acronym for the Mathematical Applications Group, Incorporated, they formed the plural of the singular magus, which came from the Latin meaning "sorcerer." From this root came the word magic. This is entirely appropriate, as we shall see.
Magic is not a bad word to describe what is going on at the MAGI facilities in Elmsford, NY, and Santa Monica, CA. With the publicity that has accompanied the film Tron, it is unlikely that you are entirely unfamiliar with the work of MAGI. If, in addition, you have seen commercials for the arcade game Zaxxon, or the home video game Worm War I (which, by the way, was written by my predecessor at Creative, David Lubar), you have seen the work of MAGI. What you have seen is vivid, computer generated imagery, real enough to captivate.
I recently had an opportunity to visit the Elmsford labs of MAGI, and to speak with some of the people who are responsible for these remarkable effects. Crossing the Tappan Zee Bridge on a misty morning, I made the turn into Elmsford. I fought back the urge to look for the tower of the Master Control Program, and followed the directions I had with me.
The first person I spoke to when I arrived was Chris Wedge, a young artist and animator. He led me through a labyrinth of corridors. Computer animation is only a small part of what MAGI does. With over 150 employees, they also provide computerized services for the direct mail industry. They developed a business slide-making system now marketed nationally by Xerox Corporation. The company is also involved with computer aided design and manufacturing (CAD/CAM). Chris led me back to the section of the building where computer generated animation takes place. 3-D Ferrari
He had what looked to be the gutted remains of a toy car. Stripes of fine tape mapped grids across the contours of red fenders, hoods, and doors littering his desk. "I'm modeling a Ferrari," he explained. "The process of inputting its three-dimensional coordinates is somewhat tedious. "He directed my attention to a Chromatics terminal nearby.
There on the screen was a rudimentary rendering of the car, as if it had been fashioned out of luminous guitar wire. With the click of a few keys, he made the image spin slowly. "This is the first step, "he said, with a smile. "Once we have the shape fully defined, we can make it do anything we want it to do."
We were joined by Larry Elim, who manages the place and had been kind enough to invite me for the visit. I had recently read his piece in Artist and Computer, and enjoyed it. I asked him to provide a bit of background about the company.
MAGI, he said, was founded in 1965, by Dr. Phillip S. Mittelman, a nuclear physicist and noted computer scientist. He has degrees from Harvard as well as Rensselaer Polytechnic Institute.
The company at that time had absolutely no intention of becoming involved in computer graphics. They were involved, however, in three-dimensional modeling and simulation. In fact, under the direction of Dr. Mittelman, who remains president and chairman of MAGI, the company conducted shielding studies of theoretical nuclear blasts.
MAGI is no longer involved in any way with nuclear studies or in any government contract. It may seem a far cry from nuclear shielding studies to the graphics of Tron, but jump was really not all that great. The tool for three-dimensional simulation had already been developed. In 1967, the first graphics using the systems were initiated, as a research and development tool. The first picture generated was of a helicopter. A System Ahead OF Its Time
The system had, by 1972 reached the point where it was decided that a commercial animation service could be embarked upon. MAGI was thn too far ahead of its time. In 1973, when Larry Elin, in his own words, "married the boss's daughter," a rethinking of the project took place. It was determined that the audience, capability, and acceptability of the system had been misjudged.
They did not give up, however; the project was kept alive. The software team, consisting of Dr. Mittelman and many scientists still with the company today, continued to improve the software. The idea was to wait until the time did come.
The dubbed the system Syntha-Vision, and waited. And their time finally came.
In 1978, competition began to appear--Robert Abel Associates, NYIT, and other groups. Computer power was getting cheaper. The creative possibilities of computers blossomed, as did interest in their potential as artistic tools. In early 1979, Disney Studios approached Magi, expressing interest in computer animation techniques. At roughly the same time, the business slide system was improving the cash flow fow research and development in computerized animation techniques.
A group of test films was created to explore the possibilities of the Syntha Vision system. Steven Lisberger, the motivating force behind Tron, was extremely pleased with preliminary results. The impetus was there now to spur dramatic improvements in the software. Thus the process came to be one of the leading technologies in the field of computer animation.
Chris took me into the cold, humming computer room, and showed me hundreds of reels of magnetic tape. On each tape, there are five to eighteen final film frames, depending on how much data is necessary to construct a specific frame. Chris underscored the parallel between the Syntha Vision process and traditional stop-motion animation, in which he was trained. However, instead of an artist doing rough drawings on paper, going from storyboards to meticulously executed cel transparencies, and then to the camers, this whole process takes place in the computer. It becomes the animator's job to define that "whorl" will, as a final step, be photographed. Defining A Camera
The objects, whether lightcycles or Ferraris, must first be defined. Once they are, they need never be redefined--they will then be generated and regenerated as three-dimensional data sets. They can be depicted from any angle and at any distance. The next task is to define a camera, explicity setting parameters for its focal length. The animator may desire for a given shot a wide angle or a more narrowed perspective. Of course the camers position and angle must be determined. Inputting the background and defining the intensity as well as the positioning of the light source follow. The complex software takes over from there--to create entirely ficticious but utterly believable pictures.
Camera movement and object movement is defined entirely by the animator. Chris therefore feels not in the least compromised by the technology with which he works. As graphics systems become more sophisticated, offering immediate feedback to the artist, animators will not need to be on the intimate terms Chris is with details of the system. Animation can be done poorly through the MAGI system, he stresses, just as it can be done poorly using traditional methods.
The animator is still ultimately responsible for the finished product. Without him, it will be impossible for a Recognizer or a sports car to roll around a curve or tip off a precipice in an acceptably realistic or stirring manner. While animation has undeniably been transformed, creative human input remains undiminished--is, in fact, extended and enhanced.
The current hardware consists of a Perkin Elmer 3240, hardly what one might call a micro, I would agree. A new computer room is now under construction; it will house an SEL computer, which is about seven times faster. This will help surmount the delayed feedback problem. A special camera is used to expose computer images on 35mm film. An image is projected on a flat screen CRT, capable of generating extremely high resolution, even though the screen is no more than six inches in diameter. In fact, there is no shutter in the camera--the film is exposed by the single scan line.
When the frame is completed, the screen goes dark, and the film is advanced another frame. Color is handled in what amounts to a Technicolor process--black and white separations are created initially. Color can then be created and manipulated in any way desired.
The Syntha Vision technique is a marriage between computer and existing film technology--Wedge evoked the term computer-assisted animation to describe the process.
Though the Tron script was entirely storyboarded, it evolved as time went on, as did the capabilities of Syntha-Vision. The casting of shadows, the depiction of reflections on the surfaces of objects, and the attenuation of light--that is, the drop-off in the intensity of light with the increase of distance--were added to the system.
Artist's renderings of each shot were laid out in detail so that everyone could clearly understand what would transpire in any given scene. After a shot had been entirely specified to the computer, it was studied on color terminals, then reworked as many times as necessary to perfect the shot.
For Tron, a terminal was installed at Disney Studios and a modem used to send shots for analysis directly. Disney in some cases called for subtle changes--Elin states that his crew learned a great deal about the mechanics of animation from Disney experts.
The animators at MAGI are guite seasoned. But they are only half the crew. The other half is composed of the programmers, the mathematicians, the physicists--they are not concerned in any way with the projects at hand. They work only to improve the software. Left Brain, Right Brain Gap
And so the MAGI group works in these divergent directions, while successfully bridging the supposed and celebrated "left brain, right brain" gap between scientific and artistic minds.
Larry states that the gap is not nearly as big as we tend to think it is. There is mathematics in art, and art in mathematics--the difference is more semantic than philosophical. A common vocabulary, like pictures, can bridge the gap. The MAGI artists and programmers embody this idea in their mutual respect for one another. They set an example well worth noting.
Chris Wedge said that bridging this gap might be essential to our evolution as a species. Larry stated the ideal quite succinctly, evoking the term "renaissance. "He hopes that technology will foster a real renaissance for the American animation industry.
Elin asserts that the computer can provide an indispensable tool for the animator, but that we should not expect too much from it. The computer can replace the "stoop" labor in animation, but never the inspiration or the artistry. He very fervently believes that computers will never generate acceptable computer animation of characters themselves, at least not with much more quality than the stuff currently littering Saturday morning television.
Elin's vision is one wherein the computer can clear the way for the artist to focus on truly caring character animation, taking account of things like expression and body language.
And what of the future? Better software and peripherals will make the process more flexible for the artist. Right now the artist works more or less blindly, waits, sees a few seconds of a result, and on that basis, modifies the effort. Efforts are underway to change this.
Christine Chang, another member of the animators staff, showed me a system now under scrutiny by Disney, wherein cels are colored using a sophisticated graphics tablet--rather than someone having to sit and color each cel meticulously by hand. This dramatically boosts productivity, and combined with other computer techniques, will contribute to the revitalization of the animation industry in this country. Certainly the only way the U.S. will ever regain preeminence over foreign concerns is to use its technology to make the process cheaper, without sacrificing quality for automation, as it has in recent decades.
As the state of the art approaches photorealism, computer assisted animation will be used more regularly in live-action films, and not just for special effects. Francis Ford Coppola has already foreseen the day when his "studios" will in large part, reside inside a computer. If he wants, for example, a ghost town in the Old West, has will simply have one generated. The day remains in the future, but inexorably approaches.
As for the possibility of Tron II, Elim would not be pinned down. He says Disney continues to ask for tests. Exploration is currently in progress for a new Maurice Sendak film to be produced by Disney studios.
It remains for MAGI to show the magic of a partnership between Disney's acknowledged masters, and the new Wunderkinder on the block. You can bet the results will be interesting--and revolutionary.