In this series of episodes, we talk to many of the scientists at Blue Sky Studios, which created the Ice Age series of animated features, including the recently released Ice Age: Dawn of the Dinosaurs. In episode 1, we hear from company founders Carl Ludwig and Eugene Troubetzkoy and senior research associate Hugo Ayala. Web sites related to this episode include www.blueskystudios.com and www.iceagemovie.com
Steve: Welcome to Science Talk, the weekly podcast of Scientific American posted on July 10th, 2009. I'm Steve Mirsky. And I've got a special multipart podcast for you about the very smart people over at Blue Sky studios. A few weeks back, I got a phone call from
our[a] podcast listener named Hugo Ayala. He said he worked at Blue Sky and they had made a movie called Ice Age: Dawn of the Dinosaurs, and that maybe I'd like to come by and see what the production people were up to. At first I thought it was a small documentary house or something. But as we spoke, I realized I was talking to somebody from the outfit that makes the widely successful Ice Age series of animated movies, featuring the voices of, well, I'll let the deep-voiced guy tell you.
(Trailer for Ice Age movie)
voice over: Ray Romano...
Romano: He'll bounce back. It's one of the advantages of being Sid.
voice over: John Leguizamo...
Leguizamo: I'm a mommy!
voice over: Denis Leary...
Leary: For the record, I blame you for this.
voice over: And Queen Latifah.
Queen Latifah: Talk to the trunk.
voice over: Scrat. And introducing... Scrat.
Steve: So, I went up to Blue Sky Studios in nearby Greenwich, Connecticut, spent the whole day there and spoke to a lot of very interesting scientists. In this first part of the podcast, you'll hear two of the founders of the company, Carl Ludwig and Eugene Troubetzkoy, and then we'll talk to Hugo Ayala. In upcoming parts, we'll talk to all the science guys on the R&D team and the director of the third part of the Ice Age series, Dawn of the Dinosaurs, which is now playing everywhere. In part 1, the first voice you'll hear is Carl Ludwig.
Ludwig: We've been around since 1987, and we worked together at a company called MAGI, and they were one of the pioneers in computer animated movies; we worked on TRON. So at the time when Eugene had been there for quite a while, they did nuclear research right?
Troubetzkoy: That's correct.
Ludwig: MAGI, [was] where the foundations of ray tracing began; there was a seminal paper by Goldstein and Nagel that was written back in MAGI.
Steve: Ray tracing—you'll have to explain [that] to me.
Ludwig: Well, ray tracing is a rendering technique, which simulates the way light really works, except it's in reverse. Rays are fired from the camera and intersections are calculated with the geometry. Once that intersection is calculated, the normal is computed and then you know, based on material characteristics and stuff and locations of light sources, a computation is made to determine how much light is coming from that object. As opposed to the standard technique, the other technique, which is a Scanline render, which you know, basically calculates the intersection of a line in plane with various patches, mostly polygons. So it came from the idea that you guys were doing nuclear shielding studies, right?
Troubetzkoy: Yes. Yes, we were doing nuclear shielding studies again applying a ray tracing approach, following neutrons that originated either from a nuclear reactor or from an atomic bomb and calculating the effects. Now the similarity with computer graphics is very clear—instead of following neutrons we can follow photons, from the camera to the scene and computer response; in this case a visual response.
Ludwig: Yeah, Phil Mittelman, who was the president of MAGI, I guess, he was the one who had the concept to visualize the stuff: “ "Why don't we pretend these are photons?" I don't know if he was the one or not, but that's the story I heard. Somebody had the idea.
Steve: So this was Louis de Brogli, coming to life to make cartoons.
Troubetzkoy: Well, I'm a student of Louis de Brogli.
Steve: Were you literally his student?
Troubetzkoy: I'm literally his student in Paris.
Steve: Wow. Prince Louis de Brogli, Nobel laureate.
Ludwig: And by the way, this is Prince Eugene Troubetzkoy.
Troubetzkoy: Well, I was at the University of Paris, and I took a course given by him on quantum mechanics, and it was a very small class of maybe a dozen people and extremely interesting discussions.
Steve: If I remember correctly, de Brogli said, "If you can treat photons in a certain way, there's no reason you can't treat electrons in the same way." That's the most basic picture of his work.
Troubetzkoy: Well, his main contribution was breaking up with the classical approach—the Newtonian approach—and he was just introducing quantum mechanics, [wave] mechanics, as he called it.
Steve: My other recollection is that his professors didn't quite know what to make of his doctoral thesis, and they asked Einstein to look at it and he said, "Yeah, this looks pretty good", and he got his doctorate.
Troubetzkoy: Yeah, yes. I didn't know him that at that time, he was [a] student, but I heard stories that people were not trusting him in the beginning.
Steve: Well, my mind is sufficiently blown for the day.
Ludwig: This is where it all began, you know with these ideas. So the work on TRON was taking place and then the company had these great ideas of expanding the business, but the industry wasn't really ready for it yet, so that part of the company folded and it was bought by another company and then that folded. But by this time we had all these good ideas of what we wanted to do, so one day we got together at Eugene's house and we decided, "Lets give this a shot ourselves." There were six of us that began and we started off in a little three-room office in Briarcliff Manor and started writing software. We bought a couple of Sun 350s at Carnegie Mellon, because one of the guy was teaching there, and started writing the software. And in the fall of 1987, the stock market crashed. So, our ideas of getting money from somebody else disappeared, so this was all our own money, what we had; and we decided we were going to run out and we had to find a way to make some money. So we started doing commercials. And we kept ourselves barely alive, doing commercials, but slowly the business began to expand. We hired a couple of people, and we built it up and we had a commercial business, where we were doing commercials, and you know, it was always very well received. And then we wanted to get into the movie business, but always with animation in mind, because one of the founders, Chris Wedge, was an animator and he wanted to push that. So, we started our first movie, I think was Joe's Apartment, which was an MTV film. We did all the roaches and [it] was kind of [a] fun project and it looked great. And then we started getting movie work from L.A. We started doing things like we had all the Aliens and Alien: Resurrection and stuff like that and we started doing all of that. And one of the advantages we had is we could go to the set and put up a couple of spheres on the set, one chrome sphere, one white sphere, see where the lights were coming from, come back to the studio, and set up the lights just the way they were set on the set, and because it's a very physical way to render, it would match perfectly. So that began to take [off and] we started building the company; and we were about what 60 people or so, right? Something like that. And then we got to the point where Chris wanted to do the short, so we did a short with a help of at the time, [Deck,] they lent us some computing equipment. We did an animated short, completed rendered with radiosity and we showed it at the Academy presentation it was like something no one had ever seen, plus it was a great little story and there was great animation, and we won an Oscar. And when we won the Oscar, that caught the attention of Twentieth Century Fox and they began to express interest in the company. So little by little, you know, we got together, and we didn't know whether we wanted to [sell it] or not, but [we'd] hanging by our fingernails for so long, our hands were getting tired, so we decided we would do it and [it] would give us an entry into making animated movies.
Steve: And the short that own the Oscar was...
Steve: Right, which is included in the DVD of the first Ice Age Movie, right.
Ludwig: Yes, I think it is.
Troubetzkoy: Yes, that's correct.
Steve: It's a very moving little cartoon.
Ludwig: Yes, its absolutely incredible and it was a real, real high point for us. So that's
Steve: Will it be pejorative if I refer to it as a cartoon?
Ludwig: No, no. You can refer to it as a cartoon if you like, you know, some people will call it animated short or whatever, but only if you want to differentiate between the standard hand-drawn stuff and you know, it's kind of like calling Fantasia [a] cartoon.
Ludwig: There's a little bit of difference.
Steve: Okay, and now in addition to the Ice Age films, what are some of the other big projects that people may have heard of?
Ludwig: Well, there's the Ice Age films, there's Horton Hears a Who! there's Robots and that's about the size of it, and there's one we we're working on now, and a couple of more down the road that are little different.
Steve: Is this mostly hard work and a little bit of fun or is it mostly fun and a little bit of hard work?
Ludwig: I'd say it's actually mostly both. It is hard work, but when you have passion for what you do, which most of the people in this building do, even though it's hard work, it's fun, it's enjoyable. And there's always a challenge, and we love challenges and we love solving problems.
Steve: And is it the very similitude that you are able to achieve because of the ray tracing that was never possible in animation before, is that what makes it just sort of jump off the screen?
Ludwig: Well, it makes it easier for us to enrich the imagery. If we have a concept or we want to render something and give it a richer feel, it's very easy for us to do that; and there's lot of things that we have at our disposal because we do work, it behaves very similar to the way light really manifests itself. So a lot of the physical characteristics, like diffusion or things like that, you can really carry forward effectively.
Steve: And the practical outcome of that is, when you say richness, there's more density of information in a given area?
Ludwig: Not really, it's not the density, it's the quality; you know, the way something looks. You know, I have this little cube here that's somewhat translucent, and its got these beautiful characteristics. When you look at it, you can see that, you know, not only [does] the light diffuse through [it], but the corners are dark. And then in a cloud, you have the same sort of thing where the edges, even though it's very bright white, always has a little darkness to them, because there's not enough material around there to diffuse the light. So, capturing images like that and being able to render them correctly is far easier for us, I think.
Steve: What do you like about this as opposed to nuclear physics?
Troubetzkoy: Well, nuclear physics, I was applying it to [either] nuclear reactors or to the bomb and [neither of them] is popular these days, whereas computer graphics is popular. And I enjoyed my work in nuclear physics, and I'm enjoying my work here in computer graphics.
Steve: I'm sure, especially the children, but even their parents, when they go to see these films—you know, I really enjoy them as a semi-adult myself—but they have no idea that the technology originated, I had no idea, with the science that you need to understand to do nuclear physics.
Troubetzkoy: Yes, that's the way we started, but other people are computer graphics specialists who started [a] different way, I mean, they just wanted to do computer graphics and the approaches are rather different. So we've learned from our experience, and I think we learned the right things, like the importance of ray tracing.
Steve: Next up is Hugo Ayala. Hugo, tell me about what you do here, and how you're M.I.T. training prepared you for it.
Ayala: I spent a lot of time with the effects group. My background is in fluid mechanics. So I work a lot on the tools that they use for doing simulations. Every once in a while [though]
they'll, you know, to really get the simulation out of the computer, you need tools to input the data, to output the data; and my background is also in creating work flows to speed up the amount of work that one person can do, simplified user interfaces, studying everything from the time that data is born till its finally rendered and tried to minimize the steps that a human has to intervene. So I do a lot of, sort of, work flow analysis and repair throughout the studio.
Steve: You have a doctorate from M.I.T. in mechanical engineering?
Ayala: Yes, I do have a doctorate. I worked on bulldozers and seals for bulldozers.
Steve: That's really interesting. Your thesis, your doctoral thesis was?
Ayala: Soil ingestion of elastomeric seals.
Steve: And this had to do with the way that, you know, big Caterpillar equipment winds up picking up soil particles and how it gets, how it gums up the works really.
Ayala: Mostly, bulldozers are big large pieces of equipment that are used in mines and the dirt and soil gets ingested into their tracks, that causes the tracks to wear, and the primary cause of the tract's failure is [t]hat the tracks basically break apart. And this all of course, starts when the dirt gets in. My contribution to this field was figuring out that it's the clay. You don't really think of [it] as an erosive particle, but the problem is that the clay particles are so small, they sneak in through the seal barrier and they gum up the oil; and that causes then the oil or the seal to start separating and dry up and then letting in other more aggressive or abrasive particles in. My contribution was to design a way to disturb the way that the clay was getting into the seals so that it wouldn't gum up as quickly, and so that the seals last longer because it takes longer for the seals to dry out and for other material to get in and for the oil to leak out, which really is the failure mechanism.
Steve: Your discoveries were put to practical effect.
Ayala: Yeah, Caterpillar is manufacturing, you know, seals that have these ridges which we designed, you know, me and my advisor and co-advisor and other research people in my group, those are being built by Caterpillar.
Steve: And you told me earlier they last much longer than the seals.
Ayala: Our tests show that they can last up to eight times longer.
Steve: That's a pretty big deal in their world.
Ayala: They're happy with it.
Steve: That's good. So, you wind up working here, you wind up going to Hollywood.
Ayala: I always wanted to. I think, you know, for a long time, I wanted to be in the storytelling business, but I knew that I wanted to take a science part of it. I always, every time I went to movies, you know, my interest, maybe since I was a little kid, was in fluids and there was one class in school that I never had to study for, it was fluid mechanics. But every time I go see a movie, I will see some effect that just seemed off, because in my head it just seemed wrong. And I knew that that's the kind of thing that I wanted to do, and so that's how I ended up here.
Steve: So, your specialty here is, in addition to the whole workflow is, to figure out ways with software, to design software, that makes the kind of fluid dynamics—cloud motion or water flow—look more real.
Ayala: Originally, that's why I came here. Right now, I don't design as much of that software because there are versions of that software available in the market that do a lot of that stuff. Most of my role now in the area of fluid mechanics is really to be as an advisor for people who use the software, who run the simulations; my favorite example is, someone came to me and said, "I have this scene; all these bubbles are coming to the surface; it looks wrong." And I tell them, you know, after looking at it for a second, "I think you need the bigger bubbles to move faster, going up, because they displace more water, they have hard buoyancy, they go up." So that's the kind of thing that I'm sort of trained to see, and I can advise a lot of
what effects people on how to improve their shots, a lot of times. And I will show you for this next coming movie, for Ice Age: Dawn of the Dinosaurs, I implemented a software for all the lava that you see in the film. Not only did I [do] the software, I actually did the artistic side of it in terms of putting it together, me and a group of, you know, with the collaboration of other departments. So, sometimes I do get to write the software and use it, but most of the time, I'm just really consulting with other effects studios, with other people.
Steve: Did you go back and study the way lava flows? Because it doesn't flow the same way as water flows—that surface tension is completely different, all that kind of thing.
Ayala: I think the viscosities are like the primary thing, like water when it goes around an obstacle, you would get the water nearby, doesn't get displaced as much, but in lava, because this is has high viscosity, a lot of lava, sort of, moves when its going around an obstacle, so that was one of the things that I worked on; and I also did the footprint software. There wasn't any good software that we could use that, from beginning to end, would cover the ability to make footprints. There are a lot of very important plot points in the movie where, you know, the fact that you see footprints is important, so I created, you know, the simulation code that makes the footprints, that makes them look realistic and allows all the effects to actually create them quickly from beginning to end. There's a place in the movie where the characters come upon some footprints and the sculpting department was going to sculpt in those footprints, but they found the footprints that were made by myself looked more realistic, so they just put one of my footprints in there.
Steve: That's cool.
Steve: That's it for part 1. Stay tuned for the rest of the series on Blue Sky Studios coming your way very soon. I'm Steve Mirsky.