An Interview with John Lasseter
(From Animato #19, Winter 1990)
By Harry McCracken
SIGGRAPH, the annual convention of' the Association for Computing Machinery's Special Interest Group on Graphics, is a mammoth convergence of the computer graphics industry that lasts a week but would take years to explore fully. Seminars and panels discuss the science and art of computer animation; acres of exhibition space display products that push graphics technology to remarkable limits.
But in many ways the heart of the show is the Computer Animation Theater, a show of the most outstanding new works of computer animation, ranging from sophisticated technical exercises to increasingly--films with characters as real as those in more traditionally animated films. And for the past few years, a new film by John Lasseter and his collaborators at the computer hardware and software company Pixar has been among the most eagerly-awaited works in the show. In 1986, it was the groundbreaking Luxo Jr.; last year, it was Tin Toy, the first computer-animated film to win an Academy Award This year, the Lasseter film that premiered to a wildly enthusiastic reception was Knickknack, an ingenious, very funny cartoon which gives us some idea of what Chuck Jones or Tex Avery might have donewith computer animation. Luxo Jr. established standards for computer character animation that have inspired many of the best computeranimated films made since then; Lasseter's own subsequent films are among the finest of those films, and each one has shown us more clearly what tremendous potential this new art form has.
I interviewed Lasseter at SIGGRAPH in Boston in August, 1989
HARRY McCRACKEN: / should start by asking how you got interested in animation in the first place
JOHN LASSETER: I got interested in it when I was really quite young, as I guess most animators do. I used to get up very early on Saturday mornings and watch all the cartoons until the golf matches came on, or the football. And I used to go out and see all the Disney films.
When I was a freshman in high school, our library had a copy of the Bob Thomas book The Art of Animation, the one all about Sleeping Beauty. I got that, and I read it. And it was sort of funny, I realized that people actually did the job of making cartoons. And I thought, "That's what I want to do."
I can tell you exactly when I realized that I wanted to be an animator. It was at a screening of The Sword and the Stone at the local theater. I don't know if in your town there's a theater that, if a movie is playing there, you know that's it; after there, the movie's gone. It was the end of the release. Forty-nine cents. It was the Wardman Theater in Whittier.
So I saw Sword in the Stone, got out, my mom picked me up, and I said, "I want to work for Disney. I want to be an animator." And luckily, my mother was an art teacher at a high school for thirty-eight years, and she was always supportive of being an artist as a profession.
I wrote to Disney and all those things through high school, and took figure drawing courses. And when I was graduating from high school, Cal Arts was forming their character animation program as a separate program from the film graphics program. The next year I went there; it was the first year of the program. I went there for four years, then went to work for Disney.
What did you work on at Disney?
When I first started working there, I did a little bit of animation on The Fox and the Hound. Then I worked in the story department for a while, on a number of projects that didn't get off the ground. Then I worked as an animator again, on Mickey's Christmas Carol.
About that tune, Tron was being made, and that's when I got interested in computer animation. Bill Kroyer and Jerry Rees were doing it, and I saw some of the early work on that and got real excited about computer graphics. I was able to get Tom Wilhite, who at that time was head of production at the studio, interested in combining character animation with computergenerated animation. I worked with Glen Keane, who's a brilliant animator, and we did a thirtysecond test called the Wild Things test, which combined handdrawn animation with computergenerated backgrounds.
After that, I went up to Lucasfilm, and started working with their computer animation group. The first thing I worked on there was The Adventures of Andre and Wally B. which was a short animated film we did for SIGGRAPH in 1984 when it was in Minneapolis. Then I worked on Young Sherlock Homes. I've sort of done a project a year while I've been up there. In February of '86 we spun off and became Pixar and that year did Luxo Jr, and then the year after that did Red's Dream, and then Tin Toy, and this year Knickknack.
Did you leave Disney because you wanted them to get more involved in computer animation then they were at the time?
Yes, sort of. At the time, the expense was so much, and there was very little that had been developed. It required a lot of development in order for it to be usable. The Wild Things test proved to be really successful, I think in proving that it could work, but also it was quite expensive at the time. They were concerned that it was just too expensive.
There were still some people who stayed dedicated to computer animation, and since they've done some great work with it, of course.
Was there a point when you felt you reached a breakthrough with your work in computer animation? There's a much bigger difference between Andre and Wally B. and Luxo Jr. in style and approach then between Luxo and the films that have followed.
Right. It's pretty obvious that Luxo Jr. was a real breakthrough, not only for me and Pixar but for the industry as well. When we were spun off and became Pixar, they said, "For the first year of Pixar, we want to have a film in the film show at SIGGRAPH. You guys do it."
Bill Reeves, Eben Ostby and myself didn't have a film we wanted to do. So we all sort of did a little something we wanted to. Bill was working on some interesting research on waves, so he did a little piece with waves. Eben was doing some procedural animation; he did something with a beach chair. And I was interested in doing things with lamps. I had done some student films with them, and they were kind of fun.
I started working on doing lamps. I modeled one Luxo lamp, and then a friend of mine came over with his baby. And then I went back to working on the lamp, and wondered what the lamp would look like as a baby. I scaled different parts of it down: the springs are the same diameter, but they're much shorter. The same with the rods. The shade is small but the bulb is the same size. The reason the bulb is the same size is because that's something you buy at the hardware store; it doesn't grow.
So I animated it, and the story developed as I went, and we premiered it at SIGGRAPH. I love showing the films at SIGGRAPH because you get such a great reaction. The reaction to Luxo Jr. was phenomenal; people had never seen anything quite like that before, and it got a really wonderful ovation.
The thing I wanted to do in Luxo Jr. was make the characters and story the most important thing, not the fact that it was done with computer graphics. As you see in the film show at SIGGRAPH, a lot of times it's computer graphics for computer graphics nerds. People get excited about it purely because it was generated with a computer.
I wanted people who had never even seen a computer before to look at it and enjoy it as a film. I did a couple of things: I locked the camera down, didn't move it.
There's so much stuff flying around in computer films.
Oh God, yeah; you get sick. They do it because you can do it. And people tend to have real bright colors, without thinking about the way things look.
After the film show, Jim Blinn, who's one of the pioneers in this field, came running up to me and said, "John, I have to ask you a question." And I thought, "God, I don't know anything about these algorithms; I know he's going to ask me about the shadow algorithms or something like that." And he asked me, "John, was the parent lamp a mother or a father?"
You figured you had succeeded then.
Yes, exactly. Here, one of the real brains in computer graphics was concerned more about whether the parent lamp was a mother or a father.
It's interesting; that question keeps coming up. A lot of people say it's a mother; a lot of people say it's a father. I always envisioned it as a father, but it's based greatly on my mother. To me, if it was a mother lamp, she would never let the baby jump up on that ball But the dad is like, "Go ahead, jump up on it, fall off and break your bulb. You'll learn a lesson."
What role do you play in making your. computeranimated films in comparison to the role the director or animator plays in the creation of a..traditional animated film?
I come up with the initial concepts. We bounce the idea around with the crew we have. Most of them have computer backgrounds, but over the years they've become quite savvy with animation and stories.
So we usually develop the stories together, and I'll do the storyboard. From the storyboard we define what needs to be modeled. We generally divide up the modeling task between the crew. I'll do some modeling, and then I'll do all the animation, generally. Some of the other people have started doing a little bit of the animation.
I also direct it as far as what it looks like, color decisions, staging it, doing the angles. It's sort of up to me to keep the storyline together in my head. And then Bill and Eben usually are the ones who render it, after I'm finished doing the animation.
Have you been particularly influenced by any artists in your work?
Yes. There's Walt Disney; his films are just brilliant in their staging and characters, of course. Chuck Jones is probably my next biggest influence. As a director, he has the greatest timing there is; I think you'll agree with that.
But also there's Frank Thomas and Ollie Johnston as animators. The reason I like their work so much is that they do such great characters. I love the work of Ward Kimball as well; he's a big influence. But his work, and Milt Kahl's work, are much more identifiable. You'll look at Milt Kahl's work and say, "Oh, there's a Milt Kahl scene." His stuff is brilliant, but I think Frank Thomas and Ollie Johnston do work where the character is it, is everything, and their stuff just comes alive.
Also Norman McLaren and the Canadian Film Board...
How has he influenced you--his use of color?
Yes, and also his series of films called "Animated Motion." They're such wonderful teaching tools.
You know, in going to Cal Arts, and being born and raised in Los Angeles, Disney and Warner Bros. cartoons were basically my sole influence. But since I left Disney and started doing short films, I've been going to animation festivals around the world. And that's like a whole new world that's opened up to me of short films: the Film Board, all the European films, and even things from the United States. Stuff that you never, ever see, and it's such great work. The work of Paul Driessen, the work of Bill Condie. Cordell Barker's The Cat Came Back is just fabulous. I keep getting influenced by these people.
And I loved Roger Rabbit. It's almost like an animated film for animators. All the gags they pulled from history. They just went nuts, and it's really fun.
What are some of the stumbling blocks in using computer animation things you'd like to do but can't? Is that in your mind much?
In computer animation there are a lot of limitations that traditional animation doesn't have. And vice versa, actually. As soon as I started working with computer animation, I realized that the easiest thing to do in handdrawn animation are the most difficult to do in computer animation. An example of that is organic shapes, like Dopey and all the great animation of the Dwarfs. You see that and it's just so fluid, and yet it seems connected. That's so hard to do with computer animation; it's virtually impossible. It's easy to make a sphere or any object scale in X, Y. or Z, but to make something move around and keep the same volume is so hard. We keep doing research in that area.
But then trying to animate a room with a moving camera shot in hand animation, is also virtually impossible. And also, the shadowing, and shading, and lighting, and reflection, refraction...all that stuff you get in computer animation is virtually impossible to do in hand animation. To me, it's really important for animators to understand the medium they're working in, whether it's sand animation, clay animation, cel animation, or computer animation.
Traditional animation is one cheat after another. It's always an illusion of depth, or illusion of this or that. When I work with the computer graphics guys, they seen much more to be purists. They really want it to be truly refractive, truly this or that. So I've introduced a lot of traditional animationlike cheats into the computer animation we do, and it's really broadened their perspective a little bit.
We keep pushing the boundaries out, and now I know exactly what areas are very important to me but difficult to do, and those are the areas are the kinds of places we focus in on. If you've seen procedural animation, like Chris Wedge's Balloon Guy, where things are just kind of blublublublubluba. (Flops around loosely in imitation of Balloon Guy) I love that kind of thing. The way Chris did it with Balloon Guy is great, because he as an animator defined the initial stuff, and then let the computer do it.
There are a lot of people who are just letting the computer do the animation. You can just type in "Character Walk," and it'll walk someplace. That takes the fun out of it for the animator. So what we've done is always keep the animator in initial control, and then let the computer do some of the more mundane stuff. The first use of procedural animation was in Luxo Jr., with the the ball rolling. Making a ball roll on the ground is actually quite difficult, because you have to match the translation with the rotation, and the size of the ball and so on. And I sat there with a calculator figuring all this out, and I realized, "What am I doing? Computers should be able to do this."
So Eben wrote this whole procedural animation system we have that does that. In Red's Dream we did it with the unicycle: the wheels turning, and keeping the pedals flat, and all that. All I did was to do a pass, with the timing of it and the character moving around. The snow (in Knickknack) is another good example. I just animated the character, and played with a few parameters, and the computer did all of the snow floating around. So as we go on, more and more tools are being developed. It's getting more and more power, but the animator still has the initial control, and we can still tweak it after the computer is done.
Which of your films or characters do you think has been most successful so far in achieving what you want to do?
Luxo Jr., without question. Tin Toy won the Oscar, but I wish the baby had been a little more cute. But the story was to the point where it was a baby monster, so it worked. It worked really well, in fact; it may have been better, since the baby looked kind of bizarre, than it might have if the baby was really, really cute.
I like the sad ending in Red's Dream. Knickknack I think works really quite well. It's surprising the reaction that it's getting. But generally, most peoples' favorite is Luxo Jr., because it's just this little simple thing, and it's complete on its own.
Knickknack seems more cartoony than your other films.
Right, it was a very conscious decision.
It's more of a gag cartoon.
Right. After Tin Toy, we really wanted to do a cartoon. I went back and looked at my collection of Chuck Jones and things.
Another thing I wanted to ask you about was the sound effects in your films. They seem more important than in most animated films, and I was wondering if that was something related to the fact that you're working in computer animation, or if you'd do that no matter what.
I'd do that no matter what. It's the work of Gary Rydstrom, who works at Sprocket Systems, which is the post production facility at Lucasfilm. He's brilliant; we've become really close friends.
Sound has been very important to me. Actually back when I was a student and first began cutting sound effects to go with my animation, I had this scene where a lamp was falling from a shelf and breaking its lightbulb. I was trying all these big crashes, and nothing was working. And I accidentally synched up the wrong sound to it, which was this little tiny minute little "tink," with this big camera jar and everything. I just cracked up because it gave it a completely different feeling. And in a way, it was that moment that I realized how important good sound effects were.
On Andre and Wally B. the sound was done by Ben Burtt, who's won numerous Oscars for Star Wars and Raiders and all those things, and he had so much fun doing it. And Gary's done all the sound effects from Luxo on. I bring Gary in even. before die storyboard is complete;: I'll show him the initial ideas, and I always leave lots of openings in my animation for Gary to do stuff. Tin Toy was probably the peak, because he did all that: wonderful stuff. Just the fact that it was a one-man band was for Gary, because I knew he would have a lot of fun doing it.
On Tin Toy, he was cutting the sound effects for Cocoon II at the same time. Cocoon II took him about a day, day and a half todo the sound.effects for one reel. Eight or nine reels make up a complete film. Gary took six weeks to do the sound effects for Tin Toy, because he was so into it. He was so into it because he loved it. He was doing it on his own time, and he kept layering and layering sound after sound. There must have been twenty different tracks for Tin Toy, and it really shows, be cause it's so rich.
Also, I've found that when I do animation, it's very important for me that you get a sense that the character is made out of a particular something. I wanted the feeling with the lamps (in Luxo Jr.) that their bases were very heavy, so when they land it's with a thud, and so on. In Tin Toy it was very important to get a sense that the character was made out of tin, and that the baby was fleshandblood and much more massive. Sound really helps.
Is there a reason why all your films are basically in pantomime?
You noticed that. I've done two student films one called Lady and the Lamp, and the other called Nightmare. Lady and the Lamp won a student Academy Award in 1979 for animation, and then Nightmare won the same award the next year. Lady and the Lamp, my very first film, is the only one that has any dialogue.
Each film, I want to give myself a challenge, to make it interesting. If you keep doing the same old thing, it's "Ho, hum." With Lady and the Lamp, it's the story of a lamp shop where all the lamps are alive, and this one little lamp breaks its light bulb and goes blind. It feels around trying to find another light bulb and ends up screwing in a gin bottle and getting drunk, and destroying the lamp shop. And it was very important that that I wanted to do this character that didn't talk. The lamp doesn't talk; it's the shopkeeper that talks. I wanted to get the sense that he was a character without doing the typical thing of sticking a face on an inanimate object. And I think I succeeded.
The next year, everyone at Cal Arts was doing things with dialogue. I wanted to do Nightmare without any dialogue, to just let the film play by itself. It was a challenge at the time to do it without dialogue. And then when I went back and did Luxo Jr., I just went on from there.
So you were thinking that way before you got involved with computer animation.
Oh yes. One thing that Chuck Jones said that always has been in my mind I guess it was a comment towards Saturday morning cartoons is that with really good animation you should be able to turn the sound off and still know what's going on. That's something I've always taken to heart, and it's been the foundation of my stories in a lot of ways.
I think maybe soon I'll start experimenting with doing dialogue with computer animation. Generally, the dialogue I've seen with computer animation has been pretty weak. There are all these principles and things that over the years people have developed with animating dialogue. At Disney, they teach you certain things, and I'm real interested in applying those to computer animation as well, like I've applied the other principles of animation, like stretch and squash, and anticipation, and timing and so on.
Are you interested, when it becomes economically feasible, in doing longer computer animated films, like features?
Yes, that's what we're working towards. The goal of our group is to eventually do a feature film. Ever since I've been with the group, we've been researching and developing computer animation systems, and with my influence it's very important to have computer animation systems that are developed for traditional animators to use. It takes quite a lot of training, but the tools are there that people are used to. And we want to get into longer forms of animation
When will that become feasible?
It's hard to say, but not long. We're starting to develop some longer format stuff.
Where do you see computer animation and yourself being in ten years or so? Do you see an end to handdrawn animation?
Never. Never, never, never, never. Computer animation is different than handdrawn animation. One of the misnomers that a lot of people think about is that computers go into other industries and replace hand workers. It's not that way at all with computer animation; it's a very different look.
Where I see the future, to be honest, is something I want to do more of: a combination of character animation done by hand, and character animation done by computers, and backgrounds done by painting and computer combined together. The technology we're developing is going to make it a lot more feasible to do that sort of thing, so it blends together better than in the past. Cel animation looks so different than computer animation, but I think with developments like what we did in the Wild Things test, and like in Roger Rabbit the shading that they achieved you'll be able to make cel animation look a little rounder, more like you can do with computer animation.