From...

March 17, 1999
Web posted at: 7:37 p.m. EST (0037 GMT)

by Stuart Silverstone

(IDG) -- For six years, Digital Domain has been the clear leader in digital visual effects for feature films. The company has received four Academy Award nominations. Some of its best-known credits are True Lies, Apollo 13 and, of course, Titanic. Scott Ross, Digital Domain's co-founder, president and CEO, originally wanted to be a rock-and-roll star and was in fact a jazz musician before working at Lucas Films from the mid-'80s until 1992. Ross recently spoke with Stuart Silverstone at Digital Domain's Venice, Calif., studios.

MORE COMPUTING INTELLIGENCE
	IDG.net home page
Computerworld's home page
Computerworld Year 2000 resource center
Computerworld's online subscription center
Reviews & in-depth info at IDG.net
IDG.net's personal news page
Questions about computers? Let IDG.net's editors help you
Subscribe to IDG.net's free daily newsletter for IT leaders
Search IDG.net in 12 languages
News Radio
	Computerworld Minute
	Fusion audio primers

How has your computing equipment changed in the past six years? When we started working in high-resolution graphics, there was only one platform: Iris hardware running [Silicon Graphics Inc.'s operating system]. Over these six years, there's been a migration toward NT, with lots more competition on the hardware side from DEC, HP, IBM. Now there's equipment with enough graphics capacity at a price/performance level far superior to what was available in 1992-93. For imaging computers, we have a dual standard: half SGI and half Alpha or Intel-based NT platforms — with SGI now migrating toward the NT-based solution.

On the NT side, we just brought in a boatload of IBM IntelliStations to supplement the DEC Alphas we acquired about two years ago.

How about application software? The acquisition of SoftImage Inc. by Microsoft provided the first major 3-D computer graphics ported to NT. Now other applications are moving to NT. But cost-effective solutions for compositing and real-time, high-resolution rendering aren't there yet; they're still too compute-intensive.

Also, we've written a lot of code in-house in C++, all object-oriented. One advantage of the time we started [our business] is that object-oriented programming came into fashion, allowing us to use building blocks in most of our software programs. Previously, computer graphics developers had written all their code time and time again, for different purposes, but using similar kinds of applications.

What about servers? For imaging computing, with its rendering issues of high-resolution, deep-bit-per-pixel images, we built render farms — SGIs and, on the NT side, racks of DEC Alphas. I wish there were a cleaner solution. We run Linux, which gives us an opportunity to run Unix-based as well as NT-based platforms on our render farms. For real-time compositing, we have servers that are multi-CPU, parallel-processor machines, big SGI Onyxes.

How do you stay ahead of the visual-effects competition that's emerged? There are no big surprises [anymore] in technology. We're aware of most everything; vendors show us products a year or two before they are productized, and we're involved in beta testing. We're up to speed on the next 12 to 18 months.

It's easy to talk about technology. But that's not what it's all about. It's really about artistry, the men and women who use the technology. That's what keeps you ahead of the competition. The real issue is retention of talent, recognizing artists, the persons who drive the boxes.

What kind of staff keeps the technology running? The imaging support group is 7/24 coverage. It's about six people. While we don't have a help desk, we're working on something like that, because as we bring in more and more artists, they are less and less familiar with computer technology — which puts more demand on our system administrators and IT people.

You must be constantly upgrading. We're lucky enough to be on the bleeding edge and to be a visible company, so vendors afford us the opportunity to see the latest and greatest. And in some cases, afford us the opportunity to design it with them.

What could corporate IT managers learn from Digital Domain? Fairly soon, the computer will no longer be a productivity tool, but a communications tool. The power base has moved from hardware manufacturers to software developers. In a very short time, it will move away from software developers toward the end user, and then the power really falls into the hands of the creative person.

The technology-adverse can now, using the more intuitive interfaces, create without writing algorithms [or] script shells. I can imagine a future of modeling and painting with a pair of virtual reality glasses: seeing the model in 3-D, using a digital paint brush or sculpting knife to model. It's not that far away; we've been doing some experiments with it.

Conversely, what would you like to learn from corporate IT managers? Security. We deal with images worth millions of dollars, images that are highly confidential. With secure distributed networking, we could send images for rendering to a central location. We'd like encryption-encoded security to be able to extend beyond our own firewall.

Any promising technologies on the horizon? Digital projectors. I'm starting to see projectors that come up to the quality of film projection — another step in a total digital process — allowing the acquisition of images on celluloid film, intermediary processing on digital and final projection with digital technology that looks like film.

Silverstone is a freelance writer in Los Angeles. His Internet address is stuarts@graphics.org.