Bill Desowitz talks technology, convergence, The Force Unleashed and Star Wars: The Clone Wars with Richard Kerris, who recently left Apple to become CTO of Lucasfilm.
Bill Desowitz: It must be fun and challenging to return to 3D graphics, given your former stint as director of Maya Technologies at Alias|Wavefront.
Richard Kerris: It is. I told my friends it's like coming home. I've spent 20-plus years working with a lot of the people here, building tools for them, and now I'm part of it. And it's been a great welcome and it's been a wonderful experience.
BD: And you picked a great time.
RK: Yes, this is a big year for us. There's so much going on here, starting with our facility here at the Presidio coming together and then the projects that are happening.
We have all this stuff coming on the horizon, and to be a part of the team now and to talk about the tools and technologies that are necessary to create the next level of where we want to go with the next projects that are happening. It's a great opportunity with all the different assets of the company now in games and TV and film and animation and sound all figuring out ways of doing things together.
BD: So what can we expect this year in terms of raising the bar?
RK: I can tell you that the path on how we get there is really where the innovation is taking place. We are working hard to find ways to use collaborative technologies across our different divisions so they can both influence and inspire each other in the work that's being done. And what we're seeing as a result are newer approaches to things or new ways of getting things done, [such as] having a film discussion about how they can achieve a realtime environment for prototyping it or previsualizing it that's derived out of our game group. Or [conversely], taking the influence of the talented visual effects supervisors for determining how a game looks. We have this great work that's happening with ILM [Indiana Jones and the Kingdom of the Crystal Skull, Iron Man, Star Trek], we also have LucasArts going to be releasing The Force Unleashed, which is this amazing game that is going to come out on the top platforms, and it looks spectacular. And a lot of the influence to how it looks and feels comes from this collaboration, which sometimes is two guys having lunch and talking about their different approaches: one guy being from the film side and the other being from the game side. Now we are exploring how to take that collaboration further by taking technologies from one to the other.
BD: Zviz, for instance, which utilizes the game engine for previs in film and animation.
RK: Absolutely. The animation group is using Zvis for storyboarding and storytelling to mock up and map out how they want things to go. Those assets then drop right back into their pipeline and they can go off and do that. But we're also going to see it happen as a non-linear choreographed system where directors can sit and do shot blocking, composition and stuff and be able to drop it back into the Zeno pipeline or a motion control rig, but really bringing those technologies together. And we think we have the most advanced system in the world for doing something like this, primarily because of the blend of people that we have tackling these problems.
You know, if you're in film effects all your life, and you think film effects for whatever challenge you have, it can be groundbreaking. But if you're in film effects and you talk with someone that's been in game building for their entire life, and you discuss what your challenges are, you naturally come to some new ways of thinking. And this can be really exciting, because we have people thinking about how to do lighting more like a film in a game. So I tell the people here that collaboration happens with the discussion you have over lunch all the way down to the technology that you're going to use to do it.
BD: Let's talk about realtime rendering, which is now being touted as the holy grail.
RK: We are absolutely down that path of exploring what it would take to have the game engine be the core rendering tool for Zviz, which would then allow you to tell story as if you were playing a game. Now it sounds simple on the surface -- just put the game engine in there -- but the reality is that there's a whole lot of stuff that has to be thought of and interacted with to make it work. So we're actually bringing in talent from UI design, because you really want to have a new conceptual way of doing this without having to be a scientist. So we're really thinking about UI in as much an innovative way as what's going on the screen. So how do you interact with this tool and what's the new paradigm for non-linear choreography? I think what we'll see in this next breakthrough are the results of this merged technology. George's vision of getting everybody under one roof and great things will happen [is starting to become a reality].
BD: Where are you currently with Zvis?
RK: We're working on the 3.0 version. We have taken a lot of technologies from various groups and reviewed them and implemented where necessary. And our goal is to have a tool that is being used by our divisions. And we will have that out internally within 45 days that will get used in production. There are a bunch of new technologies that we have put in, but, again, we're looking at the whole way that you use it: the user interface approach to it. I think what we will see is a tool that can be used by directors, by storytellers, by animators and by previs. It's the word processing of storytelling, if you will.
BD: How far are we from seeing the game engine being used as a final render solution?
RK: I don't know. Everybody's always talked about games looking so great that they're going to be done in film, but then what's done in film and TV pushes the envelope further. I think that in our world we use it as a tool for storytelling and previsualization. We're not looking at it as a tool for final render, because we're always going to push the envelope of what can be done, if it's a stylized render, which is being done with Clone Wars, or the break new ground [photoreal] kind of thing that's being done in the ILM space. So I don't see a machinima-type approach at this point. But certainly we're excited about some of the recent developments in the industry, with mental images being acquired by NVIDIA. We think that will bear some wonderful fruit down the road that we'll all take advantage of. But I don't think that we'll see movies from a realtime game engine that people will want to watch. Remember, people are getting educated about the quality they've come to expect from watching movies and TV, and that's a rendered quality. So to pull them back into a realtime thing, they're going to think: game. But the technology's blending one another in the process and this is really where the magic is: that pipeline of creativity.
BD: What is happening with regard to improving performance in both hardware and software?
RK: Certainly we have a great relationship with AMD. And I'm real pleased with the roadmap that they've got and very pleased with the support that we've had on our render farm and desktops. So from a hardware standpoint, we're really doing great. Same goes with NVIDIA as well. We use NVIDIA cards all over the place. Again, I look at what's taking place and we're going to see some interesting things, with AMD taking on their graphics subsidiary [ATI], and a graphics card company, NVIDIA, taking on a rendering company, mental images, [as well as a physics processing company, AGEIA].
With regard to the speed and iteration of things, there's a lot being done in the software side of the world, both internally with Zeno and Zviz, but we're also very happy with the progress that's being made with our partnering with Autodesk. Their projects being done with Maya and 3ds Max and the acquisition of Mudbox has us very excited about what that has to offer. And I think we're uniquely positioned to take advantage of this merging within the industry of hardware and software… We're the longstanding brand and customer that can push these companies to deliver the tools that we need that can be used across these different mediums.
BD: Let's talk a little about current R&D at ILM.
RK: The depth and breadth of our R&D staff is world class. And what they're doing in the areas of Mocap and rigid and solid body dynamics and fluid dynamics, I can't say exactly where you're going to see it, but it's going to be out in the very near future, and you're going to be able to see the results of some of their work in projects at ILM, LucasArts and the animation division. I think we're also pursuing papers about some of these technologies that will be presented at the upcoming SIGGRAPH. It's definitely part of our DNA.
BD: What are the next holy grails?
RK: Well, in the convergence area, the holy grail is figuring out the blending of game and film where appropriate so you can take advantage of different toolsets like previsualization using a game engine or the game using lighting that's been derived from film.
The holy grails in the area of performance is just being able to optimize as much as you can out of the graphics chips that are out there and the performance of the CPUs, so that things that we're pushing them on, the fluid simulations and so forth, can happen faster and faster. I look at some of the stuff that we're doing for the films being worked on right now and it's just incredible how real-looking this simulation is. So whether it's fire or water -- stuff that we've seen in the past -- but now being given to an artist so they can almost interactively place and manipulate and do the things that they want with it, so they'll be able to visualize more what the end rendered result will be like.
And, again, with the graphics that are happening in our games division, and the approach that they're taking that has that influence coming from the film group, you are seeing some visually exciting results. The Force Unleashed has a lot of depth to it and you feel like you're in the environment with the characters. And a lot of that has to do with the lighting and the texturing and the development going on. And I think the holy grail there will be to continually push the film approach and get that high quality.
An interesting metric that we're looking at here is how many iterations are necessary before a shot is going into the final path In the past, it was 30 or 40% beyond your specific effects shot to test it out, see how it looks. And that's being brought down dramatically because the artist can interactively get a better understanding of what they're going to have as an end result because we're putting more processing power, graphics power and, in some cases, engine power into the mix.
BD: Let's talk about Clone Wars, which debuts as a feature in August through Warner Bros. and launches as a series in the fall on Cartoon Network and TNT.
RK: I'm a huge fan of the Clone Wars series here because stylistically it's beautiful, it's really nice on your eyes and the amount of work being done is unprecedented for an animated series. We are seeing more work going through that pipeline than ever before in a time period that's shorter than ever before. You have a very stylized look to it, it's got a beautiful quality as far as texture and lighting and, because the pipeline has been built to be so optimized, the artists are pushing through minutes of animation per week so that when it's ready to go, it's going to be one after the other after the other. You won't have to wait another year for a thing to happen. That's an example of a pipeline using both third-party off-the-shelf technologies and our own stuff, including the game engine approach to visualization for things that are being storyboarded. It's a beautiful mix of things and I'm very proud of it.
Bill Desowitz is the editor of VFXWorld.
Bill Desowitz, former editor of VFXWorld, is currently the Crafts Editor of IndieWire.
