Larry Cuba on Choreographing Form

Larry Cuba, the computer film and animation legend featured in textbooks and known for his work on the original Star Wars, has long inspired coded animators, from the demoscene to today. One such artist, Andreas Gysin, joined Peter Bauman (Monk Antony) to delve into Cuba's illustrious career, its challenges and the progression of his techniques in the context of abstract film.
About the Author
Larry Cuba, Calculated Movements (Still), 1985. Courtesy of the artist

Larry Cuba on Choreographing Form

Larry Cuba, the computer film and animation legend featured in textbooks and known for his work on the original Star Wars, has long inspired coded animators, from the demoscene to today. One such artist, Andreas Gysin, joined Peter Bauman (Monk Antony) to delve into Cuba's illustrious career, its challenges and the progression of his techniques in the context of abstract film.

Peter Bauman: Let’s start at the beginning with the iconic First Fig (1974). You wrote the algorithm on a mainframe computer on borrowed time at NASA's Jet Propulsion Laboratory (JPL), where you got your start in computer filmmaking. What were the challenges of working with those machines in a creative way in a place filled with scientists?

Larry Cuba: It was brutal. I approached John Whitney to see if I could apprentice with him and learn computer animation. But he was between computers at that point so he wasn't working on something. I was fortunate to find out that this professor at CalArts [California Institute of the Arts] had made an arrangement with the Jet Propulsion Lab to get time on the systems for her students. But there weren't a lot of students who were interested in programming at that time, not in the art school but I jumped in. The Jet Propulsion Lab was quite a distance from CalArts, where I was living. And there was also this thing called funny money. When they give you time, they put “money” in your account (which is just a number). They charged you things as you use them. So when you use up time on the mainframe, it deducts from your account. If you print something out, it deducts from your account and so on. It was like telephone calls at that time; it depended on time of day and day of week.

The cheapest time was on the weekends and at night. So to maximize my funny money account, I would optimize when I used the machines. And I really couldn't do that from the Jet Propulsion Lab. I just couldn't commute in the middle of the night and so on. So I got a teletype and a modem and worked from my home at CalArts. The teletype was text-only. So I was trying to debug my graphics programs from a text-only machine. I would work during the week and then on the weekends, I would do the long runs where the animation was plotted to magnetic tape. Then On Monday, the operators would come in and run the tape through the Stromberg Carlson SC4020 microfilm recorder and plotter, which makes the film. They process it and then the film would be ready on Wednesday. So on Wednesday, I would drive out to the Jet Propulsion Lab to pick up my film. Then I'd drive over to John Whitney's house, which is in Pacific Palisades, to look at it because CalArs didn't have any 35-millimeter equipment and the film coming off of the computers at Jet Propulsion Lab was 35 millimeter. John let me use his Moviola and I’d go out there, look at it and then come back and start the cycle again.

Cuba's room in 1974 when making First Fig, showing the teletype, modem and printouts taped to the wall. Courtesy of Larry Cuba

Basically, I was getting one roll of film per week. This was my feedback. So that's why I say it was brutal, both because of the turnaround time and because I was dealing mostly with text. They had graphics displays at Jet Propulsion Lab so when I was there, I could see if the programs were actually drawing correctly. But for the most part, I was looking at the numbers and seeing if they made sense. Once I had all the black-and-white footage from the computer, Whitney also lent me the use of his optical printer to rephotograph it and add color. This was his standard operation, too. He got black-and-white material off of the computer and then he rephotographed it on his optical printer to add color. But one thing I learned from making First Fig was that I didn't really want to use an optical printer to add color to the black-and-white imagery because that process wasn't algorithmic. It all of a sudden became a completely different process.

At that point, I decided that I would only work with parameters that I could control from within the program so that I could work algorithmically. That's why the rest of my films turned out to be in black and white. They were also in black and white because I was interested in the choreography of form—the movement, the motion. And I felt that the black and white emphasized movement.

Andreas Gysin: First Fig is only six minutes long. How many minutes did you produce to get to the final film? As the process was very slow, were you able to explore many rolls to get feedback for the direction?

Larry Cuba: I can't exactly remember how many rolls I generated but I think it was maybe about an hour's worth of film, at least. And that was typical of the other films, too. I generated an hour or two's worth of material, which was essentially a series of experiments to find my way through the material. Then, at some point, I would have to stop generating new experiments and put together what I had.

Andreas Gysin: This absence of color is something that is really interesting for me. I understand that also because color, as I see it, is like an extra dimension. You already have the shapes and then you have the motion. Color adds one more dimension to the composition—something to consider and something to include in the choreography. I find this choice pretty radical because what remains is only form and movement. What did you learn about working with form, color and movement from John Whitney?

Larry Cuba: As I had mentioned, Whitney was not working on a film when I met him but he was working on getting access. When I was nearing the end of First Fig, he called me to say he got access to the computers at III [Information International, Inc.] and was ready to make a new film. But they hadn't provided him with a programmer like they had with the previous film that he made there. He inquired whether I would come and work with him to program his new film. That's when we started working on Arabesque at III. Now, the situation there was much different than First Fig because we were working directly on the film recorder itself and we had film processing equipment, too. At the end of the day, while working on generating film, we processed it and would have a roll of film to take home. John would have a roll of film to take home to work on his optical printer. By working on John’s film, I got access to the III computers to make my own film. That's when I made Two Space. It was a much more productive situation than at the Jet Propulsion Lab because I had a roll of film to take home each night.

It went from one roll per week to one per day.

Initially, I was introduced to computer graphics because I learned about John Whitney’s films. I went to LA to go to CalArts and study filmmaking there and to work with John because, literally, he was the only person I knew making computer-animated films at that time, certainly as an artist. So it was natural for me to seek him out. He had an elaborate theory of digital harmony. He wrote this up in his book, Digital Harmony. I personally didn't really subscribe to that theory so I wasn't exploring it as he had described.

What we both had in common was the primacy of the visuals. We both approached the visuals and the temporal composition of the film first.

Then music was added later to complement the composition that we had already designed with the visuals.

Larry Cuba, Two Space (Still), 1979. Courtesy of the artist

Andreas Gysin: As you brought up the music, can you elaborate on your motivation to include it? Is the motion itself not enough? Or would it then not be considered a complete animation or film without the musical aspect?

Larry Cuba: That was aspirational. It would be ideal if you produced a film that would stand on its own without music but that goal had not been reached. So it was better just to add music at that point. It's about the context as well, because when you make a film, it's shown in a certain context. In the theater, people are sitting down, focused on it, in a dark room. It's shown from beginning to end, linearly, as opposed to something that appears in a gallery, in which people can come and go and take as much of it or as little of it as they want, as it just runs all the time. It certainly might be easier to do something without music in that context than it would be in the typical film context of a theater. There are expectations based on the context.

Andreas Gysin: So they were meant to be seen at the cinema with people seated and watching from beginning to end?

Larry Cuba: Yes, projected on a large screen from beginning to end—titles and credits, just like a traditional movie. That was the paradigm at the time. More recently, I've started to wonder about that because we now have many more platforms for distributing art. I think that a lot of these experiments that are in the production process of making the final film are interesting entities in and of themselves, even though they're maybe thirty seconds or one minute long. But now we have streaming platforms where a minute of animation is a viable work. In a theater, they don't really have an outlet for a one-minute film.

Peter Bauman: The way that we consume media these days has entirely changed with social media, especially where it's catered to brief encounters. You don't get to engage as deeply as you once could.

Larry Cuba: Well, it's a different format and it's made me rethink some of the process I go through to reduce a large collection of experiments down to a six-minute or eight-minute film.

Peter Bauman: May we explore what that process looks like today? What languages do you use? Have you been working on anything new?

Larry Cuba: Well, the language has always been an issue with me. First Fig was in Fortran and Two Space was programmed in a language called RAP. RAP has an interesting history. It was a special language for generating graphics, specifically and it was based on another language from Caltech [California Institute of Technology]. That one was developed because John Whitney went to do one of his films at Caltech and they made this language for him based on the specific kind of work that he did. When I discovered it, it turned out to be perfect for what I wanted to do.

Then I lost access to III and went to Chicago to work with the GRASS [Graphics Symbiosis System] language. We need to distinguish between GRASS and Z-Grass. GRASS was the original language Tom DeFanti wrote for the mini computer and the Vector General, which was a large interactive vector display for real-time animation in the '70s. Later in the '80s, when they developed a microcomputer, it was Raster scan, low-res and slow. They ported the GRASS language over to it and it became Z-Grass because of the Z80 processor in the computer.

Cuba at the University of Illinois at Chicago on the GRASS system used on 3/78 (Objects and Transformations). Courtesy of Larry Cuba

After my experience in Chicago using GRASS to make the film 3/78 [(Objects and Transformations)] in 1978, in the '80s, I worked with the Z-Grass machine at home in my studio here in Santa Cruz and made Calculated Movements. That was such a very long, tedious project. I got a little burned out on the whole thing and stopped filmmaking for a while in '85 when it was done.

But eventually I came back to it and started looking for a language I could use. Both GRASS and Z-Grass went away, like most research languages do. I didn't want to work in C. I had some experience with C but I didn't want to actually program in it. This was the early '90s and then Python came along. It was a brand new language that I had heard about and I dove into it. A few of the Python developers were working on an OpenGL Python library so you could do graphics. That enabled me to do a lot of experimenting in the mid '90s. I was invited to do a residency at the ZKM Karlsruhe. And I kept saying, “Well, what do you program in with these SGI workstations?” They said “C.” And I said, “Sorry.” [Laughs]

So when Python came along, in the mid-’90s, I went there for my residency and started experimenting with Python and OpenGL on the workstations there. But when I came home, I didn't have the workstations to work with anymore. After a couple of years of waiting, Python and OpenGL became available for Windows.

I started up again in my home studio on a PC and I've accumulated a great deal of experiments over the years without putting them together into a film.

But I wasn't quite getting what I wanted. It wasn’t gelling. Then I had a big computer crash and meltdown. I didn't lose everything but when I rebuilt the system and put all the software back on, it didn't work. I have all this material from my Python/OpenGL era and I put together some of it into an “animation notebook,” like the piece that you see on Twitter. I've shown it as an installation where it runs continuously. It was shown at a film festival in Wroclaw, Poland, and also at a big exhibition at the ZKM Karlsruhe.

Then I started working on my own language. I formulated this language roughly based on RAP from the Two Space era but modernized in a lot of different ways. I’ve been calling the language OSKAR for Oskar Fischinger. It's only recently gotten to the point where it's usable. So I haven't actually produced much on it yet. I'm just getting going on it. But it embodies all the things I wanted the language to do that I wasn't able to program in Python or other languages.

Larry Cuba, 3/78 (Objects and Transformations), 1978. Courtesy of the artist

Andreas Gysin: Is the language you mentioned for personal use or will you eventually make it open source?

Larry Cuba: I'm curious to see if there would be anybody interested in this language because it is pretty much tailored to do what I've been doing and wanted to do with animation. But once we get the manual to reflect exactly what the language does, then I think it would be available. It's not a subset of Python but it's programmed within a Python program. You can use Python and then you can write in OSKAR if they're separated by codes. Together, this makes your program part Python and part OSKAR. When it goes through the compiler, it comes out as a big Python program, which drives Blender when it runs. It creates animation in Blender. So I've replaced OpenGL with Blender just because I couldn't stand the idea of using OpenGL again. And Blender is pretty good.

Andreas Gysin: It's a very interesting workflow. I also use a custom-made tool for myself. Then, at a certain point, it's possible to enter a cycle where working on the tool becomes so predominant that it absorbs all the work instead of then using the tool to do the work it was meant to.

Larry Cuba: Yeah, that's always a problem because it's very seductive to work on the tools, add features and think of more things that they can do. But at the outset of this language, we nailed down pretty much exactly what the syntax would be and what the language would do. There would always be more that we wanted but if we got to that point, then at least it would be usable. I'm curious about your software, too, Andreas. What are you using for the ASCII animation you do?

Andreas Gysin: I use several programming languages, depending a bit on the context. But recently, I've been targeting the browser. So it's natural to write this in a scripting language like JavaScript. For this particular project, my goal was to create a tool for an audience because I started to receive so many questions and inquiries. But it can also become a problem because, as soon as you open the gates, you get feature requests, bug reports and many opinions: “Maybe instead of doing it like this, it should have been done like that.” Maintaining a piece of software is a completely different job.

Larry Cuba: I would release it with the manual “as is.” Here's the language; here's what it does. You take it from there because a language is in itself a different project from animation.

Andreas Gysin: I found it really beautiful that you named it after Oskar Fischinger.

I remember when I discovered this man's work that it hit me deeply; it changed me and it influenced me. I completely changed the way I did and saw things.

Did you have a similar experience the first time you saw one of his films, thinking it was something new you had never seen?

Larry Cuba: Fischinger was a major influence. I first saw him when I got to CalArts and it was amazing. His widow, Elfriede Fischinger, was still alive at the time. I was able to meet her and become friends with her. She showed me a lot of his paintings in her house. That was a very extraordinary and valuable experience. I remember showing John Whitney one of my early experiments for Calculated Movements. He looked at me and said, “Ah! Oskar Fischinger!” I said, “Oh no.” Then that's too close. It’s not supposed to look exactly like him.

Larry Cuba, First Fig (Still), 1974. Courtesy of the artist

Andreas Gysin: Was it easy to access this type of content at that time? How were you exposed to these physical reels? Were there events organized around abstract cinema?

Larry Cuba: It's true; they're all on 16-millimeter film so you couldn't just call them up on the web. You had to actually go to a screening. I saw a lot of them at CalArts. There was a teacher there who started the experimental animation program, Jules Engel. He was an abstractionist himself so he showed a lot of it there. After that, there were some cinemas around Los Angeles that showed experimental films and abstraction was thrown in with the genre of experimental. There wasn't very much that focused on abstraction as a separate genre. And that's basically why I started a nonprofit with that purpose—the Iota Center—because there wasn't a central headquarters or clearinghouse for the history of abstraction in film, including the contemporary artists who were still working in the genre. We actually put together screenings of abstraction and traveled them. We had a series called Kinetica, which tried to combine the historical works with the contemporary work to make the connection and to distinguish it as a separate genre, separate from just experimental film in general. We had some success with that.

Another very important person was Jordan Belson. By chance, I saw a Jordan Belson film when I was still an undergraduate and that's what introduced me to this whole world of experimental film in general and abstraction in particular.

And because I saw that film, I pursued that interest by finding more information about it. And in doing so, I discovered that John Whitney was using computers. And that's when I realized I needed to do that.

Peter Bauman: A lot of these early abstract filmmakers were making these films based on spiritual visions, including Jordan. Was that ever a consideration in your work?

Larry Cuba: That's right. Jordan and also James Whitney, John's brother. I could relate to that.

I found that visions in the form of visuals that I could see in my mind's eye were not that helpful in making films because I wasn't aiming to produce something I had seen. I was aiming to discover something that I hadn't seen by working at the level of the software and the algorithm. 

I thought that if I had storyboarded it—or seen ahead of time what the film was going to look like—I would lose interest in actually making it. The imagery coming out of the production itself was what kept me interested in the whole project.

Peter Bauman: I’m intrigued by the quotation on your homepage from Plato that “geometry draws the soul towards truth.” Plato considered the making of art and poetry to provide for the soul as well. How do you see the relationship between geometry and poetry?

Larry Cuba: Well, geometry has a certain formal purity. Some criticism of computer graphics claims that it's too perfect. Everything is so perfectly arranged and drawn and maybe even too symmetrical. And that's lifeless. I'm attracted to that sense of perfection that we make. It's satisfying to make order out of chaos. And geometry is one way of doing that. But it can’t be too symmetrical or too predictable so it has to have the element of surprise in it if it's going to be effective. That, of course, comes from experimentation. By doing experiments with uncertain outcomes, you can surprise yourself.

I think that it's even more about choreography than it is about geometry. And to choreograph movement, motion and forms is tricky when you're looking for something with an element of surprise. Whitney often talked about the way music works—that it sets up expectations. It would set up something that you would anticipate—and you would anticipate it resolving—which it might resolve, or it might go somewhere that you didn't expect. And it would surprise you. I think that's still a problem you face when you're creating with code. How do you set up that situation?

Andreas Gysin: When you mentioned geometry and its perfection, the sacred geometry of Islamic patterns came to mind. The artists injected little imperfections here and there—small errors—because only God could achieve perfect shapes. The artists themselves had to sabotage their work as a sign of submission. The work wasn't finished until they added a little flaw.

Peter Bauman: They were drawing the soul too close to truth.

Larry Cuba: Can't get too close to the sun.


Larry Cuba has been making code-based animation art since 1974. His abstract films have been screened around the world including at New York's Museum of Modern Art, The Hirshhorn Museum and The San Francisco Museum of Modern Art. In addition to working at the forefront of computer animation at NASA's Jet Propulsion Laboratory and University of Illinois at Chicago, Cuba collaborated with John Whitney and George Lucas in his storied career.

Andreas Gysin: Art. Code.

Peter Bauman (Monk Antony) is Le Random's Editor-in-Chief.