Next Generation Digital Studio2/15/2012 2:32 AM Eastern
TDRL Pursues Paradigm Shift With Foodfight!
TDRL’s plan includes the CG feature film Foodfight! (above) and high-end CG theme park film work like (top left) Hershey’s The Really Big 3D Show.
At first glance, Larry Kasanoff appeared to be asking his dog for approval on a shot from the upcoming, computer-animated feature film, Foodfight!
In the midst of a maelstrom of activity swirling around Santa Monica's Threshold Digital Research Labs (TDRL) on a recent afternoon, the sight of animators waiting anxiously by Kasanoff's door while the aforementioned pooch, Kiley, stared intently at Kasanoff's laptop seemed surreal, to say the least. Eventually, Kiley nudged the screen with her nose, after which the producer promptly asked the artists to make a couple of minor changes to the shot.
As the animators trudged back to their workstations — Kiley following — a chuckling Kasanoff, TDRL's chairman and producer/director of Foodfight!, refused to discuss the canine's exact role at the studio. But he did concede that the Foodfight! project and his company's five-year effort to become a global, digital animation production studio are both unorthodox endeavors, to say the least.
Foodfight! is the crown jewel in TDRL's master plan, and it is, indeed, a rare development in Hollywood: an independently financed, CG film that is already deep into production outside the studio system. Company executives say the film will include 138 main characters with speaking parts, 6,254 secondary characters that will be glimpsed in groups throughout the film, and 174 sets that include almost 5,000 buildings and 12,000 lights.
If true, all this would make Foodfight! one of the most complex digitally animated feature films ever produced, and yet, Kasanoff insists the movie will be made for as much as 40% less than recent, high-end, studio CG movies. He also claims the movie will soon reach the public via a studio distribution deal currently under negotiation.
Foodfight! tells a story about what happens in a supermarket when the lights go out and all the people leave. At that point, the store turns into a wild city, where various product icons — both fictional icons and animated characters based on icons associated with real products, such as Mr. Clean, Chester Cheetah, and others, all licensed by TDRL — battle the nefarious schemes of the mysterious Brand X.
To illustrate the project's viability and to separate it from other so-called independent CG movie projects, TDRL recently gave Millimeter an exclusive, big-screen viewing of a 7-minute clip from the movie — a clip which had been transferred to film and includes fully rendered and richly colored portions of the film's opening sequence, and several scenes featuring impressive character animation. TDRL also showed Millimeter several computer models and scenes still under construction.
Kasanoff, former president of James Cameron's company, Lightstorm Entertainment, and producer of the Mortal Kombat entertainment media franchise, claims the movie will be “proof that $120 million dollar CG films can be made independently for half that price, at the same or higher quality as studio CG films, and attract the same audience.”
This is noteworthy, considering only a handful of domestic companies has ever attempted to self-finance and produce a feature-length CG film that could merit entry into the major studio distribution chain. In fact, the animation landscape is littered with failed projects (see “Cautious Hope,” page 22), making TDRL's efforts seem all the more bold.
On the business side, TDRL proceeded into production on Foodfight! before getting anywhere near a studio distribution deal. For a high-end, computer-generated film, this is an exceedingly rare and risky approach.
“We had to put a huge chunk of our own money into the project at the very start,” says Kasanoff. “This does break every rule in the producer's handbook, which says, ‘Never risk your own money.’ But we felt, three years ago, that we had to invest our own money to prove to the world we could make a digitally animated feature independently.”
According to Kasanoff, TDRL invested millions of its own money to launch the project, and has since added enough from private investors to permit the movie to be made on a $60 million budget.
The goal of all this, however, is far more than the mere production of a CG feature. The ultimate plan, Kasanoff ambitiously admits, is to move TDRL to “a level beyond Pixar.”
The comparison is made with respect since Pixar remains one of the few independent computer-animation companies ever to make multiple, successful, high-end feature films and retain its independence, although the company had to use its exclusivity deal with Disney to accomplish that goal. Its peers — PDI and Blue Sky — were sold outright to major studios (DreamWorks and Fox, respectively) in order to get there, and few others have even tried.
TDRL, on the other hand, enters the fray with no allegiance to any major studio, by offering a product that, creatively and financially, fits distributors' requirements “like a glove,” according to Kasanoff. He believes that fun CG movies produced economically represent an enticing low risk for distributors.
To offer those movies, TDRL had to become a global, digital production studio, designed to keep overhead low by working remotely with animators, all of whom are plugged into the company's network. Then, to make sure all the company's finances do not depend solely on Foodfight!, TDRL uses that same infrastructure to power a robust production services division that creates high-end CG theme park films, short films, and animated TV shows for outside clients, generating revenue and valuable R&D breakthroughs that are, in turn, applied to the feature film division.
TDRL executives emphasize that the recipe for this effort is multi-tiered, including strong stories and a reliance on a lean, veteran management team with extensive experience as film and TV producers. But the technical center of this effort is the studio's digital network.
For Nickelodeon,TDRL’s production services created Berkeley Breathed’s Edward Fudwupper Fibbed Big.
That network is built around PC-based hardware, off-the-shelf software, the development of select proprietary tools, and the use of high-speed communication technology to permit routine remote collaboration.
Kasanoff hired industry veteran George Johnsen five years ago to serve as TDRL's chief technology officer and build the pipeline's infrastructure.
“The strategy was designed to use off-shelf technology, but often in ways it was not originally designed for,” Johnsen explains. The first step in this process was choosing a uniform hardware platform, and TDRL partnered with IBM for this purpose.
“IBM had a number of advantages in terms of digital production,” says Johnsen. “First, their base hardware platform, the IntelliStation, is used on a global level, which makes it easy for any artist around the world to work within our equipment specs — an important requirement in building a virtual production network. Second, in addition to IntelliStations, IBM has vast resources at its disposal, and has particular expertise with moving, storing, and tracking tremendously large files. They were equipped to help us move rendered frames of animation around the world. They already had software and database systems that we could modify for digital film production. And finally, IBM is a big supporter of Linux.”
According to Johnsen, IBM's worldwide Linux initiative was important because TDRL felt its next technology step should be a logical migration toward the Linux platform.
“At some point, you need to make the operating system invisible if you want to maximize efficiency,” Johnsen explains. “Linux makes sense because the interior architecture of the operating system does not slow down with any of the functions, it's a universal upgrading system, and although Windows has improved greatly in terms of making the operating system transparent, Linux is even more transparent. Why is this important for us? If you are dealing with large, high-resolution pieces of art, making a movie, you don't want to see the operating system on your screen. All you want to see is the art. Also, the system works well in terms of our plan for data management and control.”
The goal of this approach, according to Johnsen, is to decentralize production. Rather than keeping an entire job under one roof in Santa Monica, or working with a single offshore collaborator, TDRL was committed to the notion of being a virtual studio, and so, Johnsen called upon Sprint's communications network.
“Obviously, we want to utilize talent wherever it exists, and we want to keep overhead low, but we also need to have everyone working with uniform tools to manage the data better,” says Johnsen. “This [Sprint] network lets us do that. Each of our animators anywhere in the world can have a common hardware and software suite, which we designed with IBM. We have a 400-node rendering system here at our facility, which we'll be growing as Foodfight! moves further along, so we can get scene files from our artists anywhere in the world and render them here into finished frames.”
In addition to using off-the-shelf components, TDRL also writes its own software and plug-ins, all dedicated to the goal of producing animation more cost-effectively.
“I'm not a big proponent of writing proprietary systems from the ground up, but I do believe in writing pieces of systems to work with installed technology, so we developed our own artificial intelligence system, to let us animate crowds quicker, for instance,” says Johnsen. “We have written other code that enhances our ability to do the sort of ‘classic animation’ look that you will see in Foodfight! But the software and plug-ins we develop work directly with off-the-shelf packages, so there is no additional time wasted trying to get people up to speed. We also use off-shelf software in combinations you don't often see — like LightWave and Maya together. We do that mainly because both have specific advantages, and both have a highly trained user community out there, so as we hire animators to work remotely around the world, we don't have to start from scratch in training them on the technology we use.”
Veteran industry visual effects supervisor and producer Alison Savitch runs TDRL's growing production services division. That division, however, has great synergy with the movie production side because it brings feature-film production procedures to the world of theme park and ride movies, a place where TDRL has found solutions and shortcuts to many technical challenges.
In the last year, the division created Hershey's The Really Big 3D Show, a stereoscopic, feature-length, 3D film for visitors to Hershey's Chocolate World in Hershey, Penn., Edward Fudwupper Fibbed Big, a theatrical short for Nickelodeon, directed by Bloom County comic strip creator Berkeley Breathed, and the company is also in production on theme park films for Disney and Paramount.
TDRL chairman Larry Kasanoff.
“For the Hershey's job, we developed a way to generate 3D stereoscopic imagery without the traditional time constraints associated with the complex mathematical algorithms normally required to create such images,” says Savitch. “In other words, we took the headache out of it for the animators and the client through our development of this technology. We've also developed code to make our LightWave render scheme more efficient, and those are only a couple of examples.”
If Foodfight! succeeds, Kasanoff predicts this will cause a paradigm shift in how CG features are born in Hollywood. “Look at the history of Hollywood,” he says. “People have made great movies for decades. Some of those movies, however, were made for the wrong price. If some movies that are now considered box-office failures were made at half their actual cost, perhaps they wouldn't be considered failures. So the requirement of a company like ours is not only to make great movies, which we are confident we are doing, but to also do it at the right price. We strongly feel that we have created the right approach to make this happen.”
Square Pictures’ Final Fantasy: The Spirits Within marked a breakthrough in R&D, but, even with a Sony distribution deal, didn’t deliver at the box office.
In early 2001, the gamble made by Honolulu-based Square Pictures, a subsidiary of Japanese videogame company Square Limited, had the entertainment industry taking notice. Square, in partnership with distributor Sony Pictures, pumped almost $140 million into the production of the near-photo-real, video game-based, CG feature film, Final Fantasy: The Spirits Within. Expensive R&D conducted by Square's engineers gave the film's characters eerily human characteristics. This, in turn, led some pundits to declare the effort a breakthrough in computer-generated feature films.
The movie, however, was critically panned and failed at the box office. As a result, Square's parent company got out of the movie business in a decided hurry, shutting down Square Pictures and abandoning its pricey, state-of-the-art Honolulu facility.
This cautionary tale illustrates the pitfalls of attempting to produce independently CG feature films, even with an advance major studio distribution deal. Without such a deal, the job gets even harder and usually results in a strictly limited production budget. Sinbad: Beyond the Veil of the Mists, for instance, was produced in the late ‘90s by India's Pentafour Software and Exports, Ltd. (now called Pentamedia Graphics Ltd.), using a unique approach of motion-capturing all character movement in order to hold costs under $20 million. That film, however, failed to penetrate American cinemas and quickly disappeared. Around the same time, a Hong Kong-based consortium of investors attempted to establish an animation facility in mainland China in 2001 to produce Through the Moebius Strip, based on drawings by French comic-book artist Jean Giraud for just $6 million. They failed to complete the movie, and the project was eventually abandoned.
These incidents, among others, might explain why attempts to ramp up existing facilities or start new ones specifically to make computer-animated feature films remain lonely efforts. All the big successes have required studio intimacy of one type or another. Pixar has retained its independence and flourished through an exclusive partnership with Disney; PDI was eventually purchased by DreamWorks to serve as its CG-moviemaking arm; and Blue Sky was purchased by Fox. Still, animation fans have yet to see a Digital Domain or a Rhythm & Hues feature, though both companies, and many others, say they have such projects in development.
Indeed, there are dozens of independent CG movies in development around the globe outside major studio pipelines. Few are in production and not likely to be in production any time soon.
Out of this fog, however, a new approach has emerged: the production of low-budget, CG movies out of existing television and home video pipelines at boutique facilities, distributed by major or minor studios. In the past year, this model has shown promise with two successful — in the sense of making modest profits at the box office — CG films. Jimmy Neutron: Boy Genius was produced by DNA Productions, Irving, Texas, from a concept developed at DNA and then sold to distributor Nickelodeon Films/Paramount. Jonah: A Veggie Tales Movie was produced by Chicago's Big Idea Productions, based on its line of family-friendly, animated home videos, and distributed by Artisan Entertainment.
Both of those projects currently have sequels in development, using the same formulas that made the first two successful. But the jury remains out as to whether those formulas are case-specific, niche successes, or the start of a new trend.
In December, Big Idea announced the layoff of 35 staff members, even as it was developing the next Veggie Tales movie. Executives say the layoffs were the result of the company needing to stay lean until the next Veggie Tales film goes into production.
To keep costs down, the Jimmy Neutron movie and television show were produced simultaneously.
“It's a matter of timing,” says Mike Nawrocki, Big Idea's co-founder, of the layoffs. “Part of our approach is to grow staff during feature film cycles and return to our pre-existing size in-between while we develop the next feature and concentrate on our core home video projects.”
DNA seems poised to churn out future theatrical CG films, not only for Paramount, but other partners as well. In addition to a Neutron sequel, the company is currently developing a CG feature called The Ant Bully for Universal. These opportunities have come the company's way largely because of Jimmy Neutron's success, which in turn, happened largely because DNA was able to keep costs under control while producing the film and a Neutron TV show for Nickelodeon simultaneously, sharing digital assets back-and-forth.
“We developed Jimmy Neutron [then called Johnny Quasar] as a CG short in the mid-‘90s, and shortly thereafter, [producer] Steve Oedekerk saw the piece and cold-called us to partner with him on developing a TV series,” says John A. Davis, DNA's co-founder, creator of Johnny Quasar, and director of the animated feature. “We finally did a pilot in 1998. The studio told us they were interested in the TV series, but also in a feature film, and the whole thing snowballed. It quickly became clear that we would have to make the movie very financially desirable for them to commit. We convinced them we could make the movie for less money with desktop tools and by tweaking our existing pipeline. We eventually sold them the rights to the character, and then set about making the movie, the TV show, and TV interstitials on their behalf.”
DNA kept its pipeline simple — PC-based hardware, and a combination of LightWave and Project: Messiah animation software for the feature film. The studio has subsequently dropped Messiah in favor of Maya, and has created proprietary plug-ins to allow it to move files seamlessly between the two packages.
“We had some ramp-up optimizing the TV pipeline to handle the feature film load and then to further update it for Maya,” says Davis. “We added dozens of computers [from Boxx Technologies] to grow our render farm. The approach worked quite well, and one reason for that was we did not attempt to R&D ways to make the character have the kind of detail you saw in Final Fantasy. This is a kid's film, and a central aspect of making it affordable was picking and choosing where to detail things, and where we could keep things simple. A mistake others have made was in having creative aspects of their story encumbered by technical limitations. If you need days to render a single hair, then obviously that is expensive and time-consuming.”
According to Davis, the movie was made for just under $30 million, a far cry from what typical, studio-made CG features cost. Its success finding an audience, however, was directly aided by having a studio partner with the wherewithal to heavily promote Jimmy Neutron on Nickelodeon long before the film debuted. Davis explains this is why, for the foreseeable future, it makes sense for companies like DNA to sell controlling rights on properties they develop to major studio partners in return for a slice of the pie and the opportunity to produce the film themselves.
The Veggie Tales movie Jonah was an even smaller effort in terms of budget — costing just $13 million to produce, and earning a domestic gross of almost twice that amount. Like Jimmy Neutron, however, Jonah was a property with a built-in audience, since Big Idea had been distributing Veggie Tales videos for years.
Dan Philips, the company's former vice president of production, was among staff laid off in December, but he remains confident that, under certain circumstances, small companies can succeed with low-budget, CG films.
“At the time, Big Idea had a nice pipeline in place for making videos, so we chose to make the feature film independently and go with Artisan to distribute it,” says Philips. “Artisan had a modest risk since Big Idea primarily financed production from its video sales, and eventually, we made what I'm sure is the least expensive, domestic CG theatrical feature release on record.”
Big Idea made the entire film using a single animation package — Maya. Keith Lango, currently the company's 3D director, says Big Idea worked closely with Alias/Wavefront to improve the Maya render module for Jonah, but eventually, Big Idea invested in a large Hewlett-Packard/Linux-based rendering infrastructure —” a horsepower rendering solution,” according to Lango.
For the Jimmy Neutron feature film, DNA Productions optimized its television pipeline.
“We made some rookie mistakes, thinking our capacity for a feature would only be a couple times more than our capacity for making half-hour videos,” says Lango. “But capacity actually goes up more like a factor of 10, so to get it done in time, we had to invest in a greater render capacity. We hit the market at the right time, and in that, we were very fortunate. Linux had just become mature enough and PCs sophisticated and powerful enough to handle this kind of rendering load. We started the project with older SGI technology, but as it progressed, we migrated to the less expensive PC/Linux world, and we are continuing that migration right now.”
Philips adds that other shortcuts worked for Jonah. “We used the Alias compositing package, Composer, which is now obsolete, because we bought it with Maya and stayed in that world the whole way through,” Philips says. “We also relied on video dailies, which cut costs tremendously, and Henry Vera [Big Idea's director of technical development] developed a calibrated color lookup table to make sure the images matched the way we were viewing them on a benchmark monitor. These steps, along with things like locking in the story reel with an edit in the preproduction phase, conforming to HD, and a few other things, allowed us to shorten or skip steps.”
The experience convinced Philips that small creative groups can have success churning out feature-level productions, under certain circumstances. His subsequent layoff has not altered that view.
“I think the tools are already there,” he says. “Hardware is no longer prohibitively expensive, off-shelf software is very sophisticated. Now, it's a matter of having the business savvy to obtain independent financing early on, and of further developing global production capacity, so that companies can work more closely with studios and artists in places where labor is less costly.”
One recent CG project — Through the Moebius Strip — attempted to take advantage of inexpensive labor overseas by building a PC-based digital studio in China. Like many others, however, it came up short. A principal player in that effort, director Frank Foster — now CEO at Tigar Hare Studios, Sherman Oaks, Calif. — has not lost his enthusiasm for the concept, however.
“We ran into problems — some expected, some unexpected — like problems in communication and language barriers,” says Foster. “Obviously, there were financing issues, as well, so I finally left the project. But I haven't given up on the notion of using offshore solutions. Right now, at Tigar Hare, we are working with a studio in India [Maya Entertainment], and as part of our plan, we are talking about a CG film. I think it's inevitable that this model will eventually result in a commercially viable CG film. I think the combination of a boutique-type 3D animation company in the U.S. with an offshore facility to stretch the budget as far as possible is a good formula.”
DNA's John Davis also feels strongly that independent CG movies will become commonplace in the near future. “People running 3D facilities like ours, we are all storytellers in search of some sort of distribution,” Davis says. “When we come up with something good enough, and make smart business and technology choices, we can earn distribution. Some of them will obviously be low-budget, but that can actually be liberating. With a lower budget, you can make niche stories. When the budget is over $100 million, you are limited to something with the widest possible audience appeal. CG animation from small studios is now well entrenched in television and the direct-to-video market, so why not in feature films also?”