The year was 1982, and computers were still an enigma to most people. In fact, so were computer graphics, particularly outside university and research settings. Nevertheless, Disney built a movie around these concepts, devising a story line that unfolds within the mysterious setting of a computer mainframe, with many of the characters assuming the identity of components: User, Program, and RAM to name a few. It was a time when the term “hacker” was not widely understood, nor were the concepts of rendering, shading, modeling, and so forth.
While the live-action TRON was not the first feature film to use computer graphics, it was the first movie to make extensive use of computer animation—at a time when few companies or individuals knew how to create CG effects. Yet, a handful of brave pioneers from fledgling start-ups accepted the challenge to put CGI on the grid. One such person was Frank Vitz, who served as a technical director at Abel and Associates, which, along with MAGI, Triple-I, and Digital Effects, created the CG for TRON. Today, Vitz is a CG supervisor at Elec-TRON-ic Arts, where he continues to set new boundaries in computer graphics in next-generation video games.
Here, Vitz talks with CGW chief editor Karen Moltenbrey as he takes us back to the early 1980s and gives us a unique perspective of what it was like to work on the groundbreaking movie TRON. (For in in-depth look at today’s cutting-edge effects used to create the sequel TRON: Legacy, see “Inside Job” in the December 2010 issue of CGW.)
How long were you at Abel and Associates?
I worked at Robert Abel and Associates for almost seven years, after Bob Abel hired me as a technical director in 1980. My titles ranged from technical director, to director of research and development, to VP of production. It was a glorious run, and I consider myself extremely lucky to have worked with the elite group that passed through Abel’s doors during that time…many of whom have gone on to become visual effects mega stars.
What were some of the memorable projects that you worked on at Abel and Associates?
There were so many great ones. I worked with Kenny Mirman on a series of commercials for TRW that earned us several Clios. I was in charge of R&D on the raster graphics system that we debuted on the short film “High Fidelity” directed by Randy Roberts, and which we later employed on a series of groundbreaking commercials, including the so-called “Sexy Robot” commercial for the Canned Foods Council. That was one of our first fully rendered CG commercials. And of course, one of the biggest highlights for me was our work on TRON.
Was that your first foray into feature-film visual effects?
Yes, in the sense that it was the first big movie that I worked on, but we created visual effects on all the other projects we did at Abel, as well. These included music videos, commercials, and special film projects. We had a full complement of visual effects tools and techniques at our disposal, most of which we had invented ourselves, including models and miniatures, motion control, cell animation, live action, rotoscoping, optical compositing, and the brand-new computer graphics systems. The Abel style, if it could said that there was one, was always a synthesis of multiple techniques combined in new and imaginative ways. Bob liked to say, ‘We create visuals that remind you of something you have never seen before!’
Specifically, what work did Abel do for the original TRON film?
My team created a teaser trailer for the movie first, which we parlayed into being awarded the creation of the opening titles and the ‘Real World to Game World Transition Sequence,’ or ‘Flynn’s Ride,’ as we called it. We also did a lot of the tedious compositing work on the live-action sequences inside the game world; these were shot on high-contrast film, transferred to animation cells, and composited on special computer-assisted animation camera stands. So, you actually took us on the first visit to the Grid?
Yes, our main digital effects sequence takes the viewer from Kevin Flynn sitting in front of the giant laser, getting disassembled and sucked down through the energy matrix, to the internal world of TRON, where he arrives and is rezzed up as a game world avatar of himself on the Game Grid. (It’s fun to use those archetypal words to describe a movie made decades before the films that brought them into the mainstream! But TRON really was there first.)
What kind of equipment did you use back then to achieve the effects?
The core of our suite of tools was the Evans & Sutherland Picture System (PS2). This was a high-resolution vector graphics machine capable of drawing beautiful, smooth vector lines in real time on a CRT screen. We had several of these machines, each hosted by its own PDP-11/60 minicomputer (from Digital Equipment Corp.). The original use of the Picture Systems at Abel was as an interactive planning tool; designers used them to create and preview motion-control camera choreography. As I mentioned earlier, Abel’s reputation was built on our ability to seamlessly combine imagery using a variety of different techniques. We had our own motion-control camera software, called ‘Camcon,’ which could simulate and interface with any kind of camera system. Using Camcon on the PS2, we could preview what a shot would look like on any of the many camera systems in the building. We could also constrain the choreography such that moves created on the PS2 and then used to drive the various motion-control cameras would result in elements that fit together perfectly in an optical composite. It was an amazing system.
Even before I had arrived on the scene, the artists at Abel had also realized that the images on the Picture System were beautiful in their own right. We built computer-controlled 35mm cameras that interfaced with the Picture Systems and recorded the images that they generated right off the monitor screen. This development led to one of Abel’s signature looks, which comprised smoothly animating glowing vectors…archetypal and wildly unbounded; the art directors at Abel exploited it and produced an explosion of creative new styles. And when used alone in its pure computer graphics-only form, it was a perfect fit for TRON.
What was the mind-set back then when you took on that first project? This was all so new....
Well, everyone at Abel was a pioneer, and we knew it…we were known for pushing the envelope and trying things that had never been done before. Richard Taylor (visual effects supervisor on TRON) had worked at Abel before me, and had a reputation for creating beautiful and stunningly elaborate multimedia extravaganzas. As I recall, when he brought TRON to the company as a potential project, most of the guys were leery of getting involved because it seemed overly ambitious, even crazy. But Kenny Mirman and I were fearless and stepped up to work on the teaser. We did not quite know what we were getting ourselves into, but we knew it was cool, and we already had visions of high-resolution computer graphics dancing in our head.
Were you confident the work could be done?
I knew we could do it. We had the digital framework of the PS2 system already in place. I knew that the PS2 system had lots of untapped potential. And we had some of the smartest people I had ever met working with us.
Was there one thing or moment when you knew it could be pulled off?
I think that for me it was when we first put together a rough optical composite of the pullback through the TRON logo. It combined a vector graphic representation of city lights at night with the word “TRON,” and seemed to encapsulate the promise of what we might achieve.
Approximately how many of you worked on the effects for the film?
The team was actually pretty small; there were a lot of other big projects under way in the studio at the time. I don’t remember everyone who worked on it, but the core team consisted of Kenny Mirman, myself, Tim McGovern, Richard “Doc” Baily, and several others. And of course, my boss and mentor, Bill Kovacs (founder of Wavefront Technologies), provided software support and inspiration.
How long did it take?
I think we were actively at it for about a year. Most of the time we were working around the clock and often I would crash on the floor at the studio or at Kenny’s house in Hollywood, rather than make the commute back and forth to Orange, where I was living at the time.
What was the R&D like?
It was exhilarating and frustrating. We did not have time to do everything that we wanted to do, so we focused on being able to first get the complex images we were creating out onto film at high resolution and reasonably glitch-free, and second, on creating new ways to make the vector objects animate and move.
We developed a special 70mm film recorder with computer-controlled bi-pack traveling mattes. The PS2 systems were monochrome, so we developed computer-controlled filter wheels and various filters to create optical effects. The main production technique was similar to modern CG in that we used the PS2 to create low-resolution versions of our animation that would run in real time on the monitor; then we would rez up the models to the point where it might take the PS2 several minutes to create a single frame–we would record them frame by frame in multiple passes onto film.
We also developed a lot of special software tools: instancing and mirroring tools for the mandala section, wave propagation algorithms for the surface of the world Game Grid, filling and culling algorithms to help the vector system approximate solid surfaces.
Ultimately, the frustration was due to the limitations of vector graphics in general. We were trying to push the system into displaying more data than it was capable of, and it would often just crash in the middle of a long shot.
Did your team at Abel interact much with the other groups working on the film, like MAGI or Triple-I?
We were pretty much in our own little world. Richard Taylor saw how different the techniques that each company used were from one another and wisely avoided any complicated overlap among the styles. While it would have been cool to collaborate more with them (and, in fact, I got to know many of them after the fact), during production we hunkered down and focused on our own part. Richard Taylor worked directly with us and kept us well insulated from any outside distractions or politics.
What were the biggest stumbling blocks to achieving the work?
There were some fundamental limitations to the vector graphic approach that we used in our work and in trying to create film-resolution scenes—we ran smack into them all headfirst. Of course, we knew about the alternative to vector graphics; what we consider to be standard computer graphics, or CG, today was referred to as ‘raster graphics’ in the early ’80s, and we had our own raster system under development at Abel. Triple-I and MAGI were ahead of us in that field at the time. In 1982, our vector system at Abel was way ahead of our raster system, so that was the way we went for TRON. Our PS2-based vector graphics system could create beautiful glowing lines, but it could not do hidden surface removal well, and we had to design elaborate matting strategies and complex sorting algorithms to help create the illusion of solid surfaces. Most of these techniques involved incredibly long-running multiple passes on the film recorder that we had to progressively scan onto the same piece of film.
A perfect example of the difficulties this created can be seen in ‘Flynn’s Ride’ in the section inside the tunnel and approaching the I/O tower near the end. At one point, our film recorder developed a light leak right in the middle of a 30-hour shot. We ended up sealing me into the film recorder room and duct-taping shut the door and all the cracks around the door in the hope of stopping any light leaks from ruining the film. I sat in there in the dark overnight tending the bi-pack film magazines, changing filters, and eating cold pizza; but it worked, and the shot was a hero!
As an interesting aside, we did later complete and commercialize the Abel Raster Graphics system by 1986. It was a complete modular package with components that most CG artists would recognize today, including modeling, choreography, rendering, and compositing tools. We even had an object-oriented user-extensible rendering package that supported global illumination and radiosity, courtesy of graphics genius Roy Hall. In 1986, no less! The entire Abel Image Research (AIR) tool suite went on to become incorporated into the core of Wavefront Technologies’ software, which, in turn, led to Alias Wavefront and Maya (software used by the majority of studios working on tent-pole films today, including TRON: Legacy).
What is the most memorable thing about that project, as you look back?
Looking back, I find it amazing that today we can create computer animation in video games at far better resolution and quality than what took hours per frame on big computers for TRON, and we do it in real-time rates of 30 or 60 frames per second on a $300 console. We could easily re-create the original movie in real time on a PlayStation 3 or Xbox 360, but it would not have the glitches and pops that give the movie its quaint (by today’s standards) look.
Have you seen the new TRON yet? What are your initial thoughts about revisiting the Grid?
I have not seen the film yet. But I am very excited to see it. The trailers have been encouraging because it looks like they have come up with a visual style that is a logical extension of the original movie. TRON was a breakthrough movie on at least two levels: It opened the door to digital visual effects, and it predicted the future of video games. If TRON: Legacy comes across as the visually exciting next chapter in the TRON story, it will be a success for me on the first score. Whether it can rise to the challenge of predicting the next 30 years of video game evolution is the big question! This is important stuff to me personally; these days I am actually creating video games—I have to revisit the Grid every day!
What are some of the milestones you made in your career after TRON?
Gosh, I have worked on so many cool projects! After Abel folded, I went on to become an independent visual effects supervisor and technical consultant. I worked extensively with my good friend Jeff Kleiser (former president of Kleiser-Walczak studio and owner of Synthespian Studios) developing visual effects for a wide range of projects. I worked on the original Stargate movie and developed the ‘Stargate effect’ with Jeff Okun (current VES chair). I developed techniques for some of the first mocap-driven digital stunt doubles on Judge Dredd. I developed the so-called ‘squinching’ algorithm that we used to create the stereoscopic 3D effects on ‘The Adventures of Spiderman,’ a 70mm theme-park ride in Orlando. I supervised effects work for Carrie II – The Rage. I was a VFX supervisor on both X-Men 1 and 2, where I led the team that gave Mystique her 3D shape-shifting abilities. I have also worked on a variety of other CG projects for museums, film, television, and industrial clients.
What are you doing now?
You’re going to love this part...I am working at Elec-TRON-ic Arts in Vancouver, BC, making video games. In 2003, I was recruited by Glenn Entis (one of the founders of Pacific Data Images) to join EA. I had been yakking at SIGGRAPH and other places over the years about the convergence between digital visual effects in movies and the video game industry. I saw how the slow, compute-intensive techniques used in film were being gradually refined, optimized, and moved over into the real-time domain of video games. I predicted that the game industry would eventually supersede the movie world as the hotbed of new graphics development. Glenn offered me a job at EA setting up a group to help make it so. I accepted his challenge, and we founded a research group at Electronic Arts that developed quite a few advanced techniques that are just now making it into our current games. Last year, I worked on the critically acclaimed game Fight Night Round 4. Currently, I am working on a re-invention of the popular EA snowboarding franchise SSX that will feature advanced, real-time, physics-driven visual effects and procedural animation. So I have come full circle from working on TRON, a movie that predicted the future of video games and the Net, to working on the actual video game technology that is bringing those predictions to life. Pretty heady stuff!
Anything else you want to add about that momentous time?
Working on the original TRON movie was a defining moment for me and, I believe, for the visual effects industry. For the first time, we clearly saw that the future lay in the digital realm. We set sail and never looked back. Now I have been working in digital visual effects for almost 30 years! While I would have loved to have had the chance to work on TRON: Legacy (there would have been some cool symmetry in that!), I did not get a chance to go after it. I have been really busy at EA, working on bringing the world of cinematic digital visual effects into the real-time game domain. So far, I have been able to survive, and even thrive, on the game grid. Of course this is part of what TRON was about, so I guess in some sense I have been involved at least in spirit. Anyway, I am happy to be working on the cutting edge of real-time graphics.
I remember a conversation I had with Kenny Mirman somewhere near the end of the TRON production; we were sitting outside the film recorder room at 3:30 in the morning, bleary-eyed and waiting for a shot to finish so we could rush the film to the lab. We were almost done, and we knew that the look was set—nothing more that we could do to reach those unattained places that still existed in our imaginations…at least not on this movie. Kenny wondered aloud just how well the effects we had worked so hard on would withstand the test of time. Well, of course we knew even then that what we had accomplished, while very cool and groundbreaking, would soon look primitive. And it does; we ourselves being amongst those who would make it so. (sigh) That has been a recurring problem with visual effects. The technology improves and audiences become more sophisticated along the way; what looks cutting edge today will probably be outdated by next year. I think that perhaps with the level of photorealistic rendering we have reached today, this phase is finally coming to an end. Digital visual effects may become a mature craft. After all, it took 50 years for the physical film and camera technology of movies to mature. TRON certainly launched us down the road of digital visual effects, and it is amazing to see in TRON: Legacy how far we have come.
The idea of worlds within worlds is another interesting theme for me, and it’s a central aspect of TRON. The prospect of being able to build real-time game worlds that fully engage all our senses and are indistinguishable from reality—that is what I am working on now. It is still just beyond our reach, and yet something to aspire to. That is what keeps me interested and why TRON is still relevant to me.