We do a lot of work with very fast turnaround schedules - including work for episodic TV and music videos. [The shop will commence commercials projects this month.] We're able to do a lot more now than we ever were. When I started doing this nine years ago, a lot of things just weren't possible. Starting two years ago, the biggest revolution has been the switch from an SGI-based platform to an NT platform. That's by far the most dramatic change we've gone through. It means that there is a lot more competition among the people who make the hardware to make things faster and cheaper. In the beginning, you had to buy these machines: when I started using Alias [before the company became known as Alias/Wavefront, now a unit of SGI] the software they made, Poweranimator, cost $70,000 for the software and then it was $100,000 for the box it ran on. The stuff we're doing now - the boxes we're getting - we're putting them together ourselves. We're able to order exactly what we want, exactly the processors we want, the exact motherboards, the exact video cards, the exact 3-D cards, etc., for $6,000. And the price of the software has come down to match, so a product like Maya complete, which is so much more powerful than the original package, is $2,000. So there are two fronts, the hardware front and the software front, and we're able to build these custom systems for a lot less, yet they're so much more powerful. We're using dual 2.2 ghz Pentium processors and because the software price is so low we're able to utilize so many more packages: we have compositing software, we have multiple 3-D packages, we have real-time NTSC in and out on the system, which was previously unheard of. Before, you'd have to export your frames to a really expensive DDR, which was a special real-time playback system, or export them to an Inferno to even see your shots and the cut or what they might look like on a TV screen. Now we can do that right from our end; the effect of all this is that we just have more power for less money.
How does this change what you're being asked to do?
Going hand-in-hand with all of this is that the client's expectations of what we can do in a tight turnaround have become much higher. What you used to see was a kind of slow trickle from films - an effect you'd see in a film - a year later you'd see that in a commercial. Now when a movie comes out, like Blade 2 or Lord of the Rings, the tech we're seeing there for digital doubles, for example, right after the release of those we're doing digital doubles for our TV shows and our commercials and videos, too.
So for us it's exciting, it's all becoming one thing. And now that we're on the cusp of being able to produce HD in the same time frame, which is almost the same resolution as film, so within next two years easily we'll be doing the same schedules on images that are four times larger than the images we're doing now - and that's because we're able to buy more processors, faster processors, better software.
It's a big challenge for us; already this year half of our shows are going HD; by next year the plan is all the shows we're currently in production on will be HD.
Talk about your craft and how it has facilitated new areas of creativity. Discuss this in context of a cutting-edge project you've worked on recently.
We've been doing a lot of "digital double" work. What we do for [Fox series] Angel is take, say, David Boreanis, the lead actor, drive a Winnebago to the set and inside that is a full body 3-D scanner. He stands in middle of it and a laser passes over his body and translates that info into a 3-D model of him. We take that back to our shop, we clean it up, we texture it - because we paint textures based on high-res digital photos we take at the same time. Then we put a skeleton in it, which is a system that allows us to animate the model realistically and we are able to replace him in specific shots; shots that are impossible to do, like a stunt that's physically impossible or too dangerous. We put him in plates he was never meant to be in or in all-CG environments. You saw it in Lord of the Rings; they're doing it for The Matrix and we're doing it on Firefly and Angel. We also used it on the Linkin Park video we just did for "Points of Authority." The coolest thing about that was the creative involvement we had with the director, [band member] Joe Hahn. He came to us with a very ambitious concept - something they took to a bunch of other companies in town and something that about six months ago I would have said was impossible. They came to us with the idea for a four and a half minute all-CG video, which was a war between a tribe of robots and a tribe of demons - a Ben-Hur-style battle with awesome epic shots. I worked with Joe on the "In the End" video as lead animator and visual effects supervisor, along with Patrick Tatopoulos, the designer - so, fortunately, they decided to bring the core of that team back together for this video. We did the video in 12 weeks - from four minutes of black frames to four minutes of high quality, approaching Final Fantasy-style animation. An all-CG production like that wouldn't have been possible without the advances I talked about earlier.
How have the demands of creatives and producers changed?
It's just a lot more realism. For example, there are different kinds of rendering techniques: there is ray tracing, which is a high quality rendering technique, or radiosity. These are things that a few years ago were theoretical. Now people are expecting radiosity-type renders from their CG elements, meaning that the reflected light from an object that's next to another object in the scene casts light onto that first object. They are expecting things to look completely photo real, that's the thing. Before it seemed you could get away with things, like there were projects that were more CG projects, like your chrome logos and your super-reflective robots and your vehicles that didn't look all that realistic but that were still cool, like the chrome dolphin thing. People got sick of that, and now we're doing things that are way more transparent. We do tons of effects here that you wouldn't even know are effects. We're manipulating, we're changing the camera moves - we're giving the director a chance to manipulate the camera after the objects have been shot. So for the video for "In the End," when it's supposed to look like we're photographing a 500-foot statute from a helicopter, we shot that on stage where we barely had 20 feet of room from the back to the front of the stage. I was able to take that camera data and extend it in CG and make it look exactly the way the director wanted it. There are things that are on that transparent level so that it's becoming less of a novelty and more of a tool filmmakers use to make their creative vision possible.
Are the demands realistic?
No! But that's good. Once they see it one time, they want it 10 times better the next time. But I don't see it as a problem, because it keeps us on our toes and it keeps us pushing the envelope. That's what's good for the industry and that's what's good for us as a company and that's what's good for the artists.
For example for the series Firefly (a new Fox series from Buffy creator Joss Whedon), first of all it's the sheer volume of it and the quality of the shots. To do 100 shots for a weekly show that are of the quality you'd see in a motion picture - that's a big accomplishment.
We talked about digital double tech - we're doing all-CG environments that look completely photoreal based on reference photography and stock plates that have been shot on location. For the first episode, we're taking a bunch of shots that were taken in the desert in Sedona and creating a completely photorealistic desert scrub brush environment that a CG train goes through and that you're hard-pressed to tell wasn't at least shot with a real back plate. There is the sheer volume of shots, and there is the complexity in the modeling, texturing and the lighting, which is much higher than anything we've seen before. Even a year ago the producers of the show may have wanted to go a different route by making the models, making physical scale models and photographing them in motion control, but we're not doing that anymore. We're going all CG.
Is it cheaper?
No I don't think it's cheaper, it just allows more flexibility. Directors in particular love it because, in my opinion, in talking to the directors I work with, for them the biggest thing is the ability in CG to have complete freedom to move the camera wherever they want. These guys are always looking for the newest crane and the coolest rig so they can make these shots that are really dynamic and exciting. In CG you can go wherever you want and they find it really liberating.
How are the skill sets for effects artists changing?
I'm also an instructor at [ the L.A.-based] Gnomon School of Visual Effects. I see more and more kids who are interested in it every year because they have access to the software now in their homes. We're getting a fresh pool of talent almost every month, and these are people who are really, really good, who grew up on this stuff. A guy like me, I'm 32, I was on the cusp of the whole computer thing - but we're getting people now who could use a computer before they could read. It's a whole different type of computer literacy that these people have that nobody's really had before.
I did everything from traditional animation to cinematography to film history. It's been really helpful, because those are skill sets a lot of these kids don't have. It's kind of more of a videogame generation - they don't really understand so much the way things are shot, things like lighting and cinematography that go into creating the images they're trying to match. That's the downside of it It's rare that people know anything about traditional photography or traditional filmmaking - they're so removed from how images are created, and it's really important to know. Lighting for film and TV is all about tricks - there are certain tricks the cinematographers do to create things like the ambient light that comes off a campfire - and when you're creating a CG element that's supposed to match exactly, you have to know what the trick is so you can replicate it in 3-D and put our 3-D lights in the same place, and when you don't know that you're at a disadvantage. That's the key to the future: the software is going to be so cheap its going to be essentially free, so the big thing is going to be educating these people in how to use it effectively.