Catmull Offers Tech Talk

[1]

With Ed Catmull [2] receiving the Gordon E. Sawyer Oscar for a lifetime of visionary CG achievement at the Sci Tech Awards on Saturday at the Beverly Wilshire, VFXWorld asked the co-founder of Pixar [3] and president of Walt Disney [4] and Pixar Animation Studios to talk about some of the latest advancements.

Bill Desowitz: Let's begin with your take on where we are at the moment with Pixar and Disney.

Ed Catmull [5]: Well, it's a very exciting time. We were, of course, just thrilled at what's happening with WALL•E [6] and the recognition and how it struck a chord with people. And, of course, while we were making it, we didn't know this was going to happen. Every film is risky but we like to take artistic risks.

BD: And what would you single out among WALL•E's many advancements?

EC: Heavy use of depth of field for artistic purposes. Getting all this right and having it look great was a challenge. And they pulled it off: it was extraordinary animation. And we feel good about the next film coming from Pixar, Up [7], and it's different from anything else.

[8]

BD: Can you give us a little sneak peek at what's new?

EC: It's a different kind of look to it. WALL•E was trying for quite a realistic look in the world and Up is not trying to match reality but create something new. It looks gorgeous.

BD: The balloons alone are mesmerizing.

EC: Yes, and I presume you know that the look of Bolt [9] is very different from what they did before.

BD: Yes, I enjoyed speaking with Hank Driskill [10], the technical supervisor, about the rigorous painterly approach applied to CG.

EC: A great look to it and painterly in an appropriate sense in the 3D world, and the ability to do some extraordinary animation. I thought the animation on Rhino was just phenomenal.

BD: And what has been the impact of Bolt at Disney?

EC: Bolt has dramatically altered the way the studio works... Also, I believe that the stereo involved is the best that's ever been done anywhere.

[11]

The addition of stereo 3-D to Bolt [11] made action sequences pop. © Disney Enterprises.

BD: I thought the action-packed opening was very effective.

EC: Really the whole use throughout the film, and I've seen a lot of stereo.

BD: What struck you?

EC: Well, what they did was make stereo an artistic element of the movie, so it wasn't about throwing things in your face all of the time, and they made it fit there. They used it, they changed the depth, they made it an extension, in the same way they used depth of field with WALL•E as an artistic tool. So, for me, it's best when it doesn't draw attention to itself. Like, most people wouldn't have said, "Oh, they used a lot of depth of field in WALL•E." They wouldn't notice but would just say, "Oh, they made it look real good." And that's the way stereo should be: a good addition that doesn't get in the way of the story, but, on the other hand, gives it real depth.

[12]

BD: Well, we can't wait to see how the Toy Story [13] franchise goes 3-D along with all the other upcoming Pixar movies. Now, overall, where do you see efficiency and productivity improvements occurring for the CG industry?

EC: I think that films are made best when there is some optimal size of crew that works on them, so a lot of the technical drive is to figure out how we [can achieve that]. A lot of that is mundane, but it does affect the way they think about the film and it affects the costs, so it's an important drive for us.

BD: And what interests you about new technology out there?

EC: One of the questions is, as the science and mathematics of simulation improves, then how do you take that technology and put artistic controls on it? And that's a fairly difficult problem. The thing about simulation is that while there is a great deal of work that is being done in other industries, they're trying to simulate reality. And in our case, we're sometimes trying to simulate things which aren't real, and, more important, we're trying to guide the simulation. So what that means is that we have to develop new technology to do something unique in our industry.

The Toy Story franchise will be re-released in 3-D in advance of a third installment, due in 2010. © Disney/PIXAR. All rights reserved.

[14]

BD: Are there any minor holy grails that you can pinpoint in this regard?

EC: I haven't been able to put them in those terms for some time. But before we could make films, we could grab on to a few holy grails. And now there is so much going on and so many smart people on it, that it's not a single, guiding force, but, rather a large number of creative ideas falling under the umbrella of providing good tools for the filmmakers.

BD: There's been a lot of talk lately about realtime rendering. What are your thoughts on where that's headed?

EC: The drive towards realtime and fast turnaround in the creative loop is still very strong. And we're still a long ways away from that. So, there certainly are pressures to use the hardware and the graphics boards out there in smarter and smarter ways. Those are clear and strong pressures. And people will be creative about that. In addition, the availability of computation power continues to increase, but in different dimension than it used to. For many years, it was just the processors got faster. And now, of course, the hardware vendors are coming out with large numbers of processors. And while the desire to have large numbers of processors has been around for a long time, it’s only when it becomes early enough a mass market item, that I think people will be able to come up with tools to adequately use it. So the result for our industry, of course, is that we will see continued increased compute power and graphics power and people trying to come up with clever ways of applying it that give us breakthroughs in our turnaround time for the artistic loop.

BD: And quality is the key issue.

EC: Yeah, and it's clear that the games' chips are already pretty extraordinary. But then there's this gap between what the best of them can do and the fine quality you can do for a film. And when you try to bridge that gap, you basically have a problem. Then the question is: "How close can you get?" There are some that might accept lower quality. Or you might be able to use it in your decision-making loop and then have it as a pass afterwards to have the processors go and put in what you need to finish it off.

Bill Desowitz is senior editor of AWN and VFXWorld.