For a modern Hollywood sci-fi flick, the make-or-break moment often boils down to an onstage audition in San Diego. That’s where Comic-Con International, the yearly gathering of comic-book lovers, gaming enthusiasts and self-professed science-fiction nerds, takes place each July. It’s also where big movie studios are becoming a larger and larger presence, eager to gather highly in uential genre fans’ advance reactions to upcoming films while there is still time for course corrections. But even by those standards, the 2008 screening of footage for a Tron sequel was an extraordinary exercise in early test marketing.
At the time, 34-year-old director Joseph Kosinski had no feature . lms and only a handful of commercials under his belt, and the 2 ½ minutes of footage he was about to show were from a movie that didn’t exist. There was no final screenplay, no studio green light – the movie didn’t even have a name. Inside the San Diego Convention Centre, 6 000 people who had been listening to a panel discussion about the Disney film Race to Witch Mountain were asked to stick around for some “surprise footage”The title of Kosinski’s clip summed up its status – VFX Concept Test – as just a few minutes of visual eff ects with no context. “If it had flopped,” Kosinski says, “it could have killed the project.”
On the screen, two futuristic motorcycles streaked out onto a neon grid, leaving glowing trails of blue and yellow light. Many in the audience hadn’t even been born when the original film was released in 1982, but a cheer of recognition went through the crowd.The light cycles and glowing geometric cityscape were instantly identi able as the environment of Tron – and the audience wanted more.
It was, in a way, a testament to the power of the original, a movie with a surprisingly parallel backstory. Three decades ago, a young animator named Steve Lisberger pitched the movie Tron by creating a 32-second preview with abstract, blocky computer-generated (CG) vehicles and an animated disc-throwing warrior. It demonstrated what he could do with the emerging computer-graphics technology of the time, and Disney greenlighted the movie. The hero of the film, Kevin Flynn, is a video-game developer who becomes digitised and gets sucked into a mainframe computer, then has to compete for survival in cyberspace and nd his way back to the real world. Tron’s animation stretched the limits of the computers of the time, so the visual-effects team came up with a technological kluge: part of the film was rendered on a supercomputer, while the rest was animated by hand.
Tron’s stylised environments and high-tech looks were no match at the box office for 1982’s real sci-fi blockbuster, Steven Spielberg’s ET: the extraterrestrial, but it’s obvious now which of those films had more influence over the actual process of moviemaking. The puppetry and creative lighting of ET are becoming a lost art, while CG effects reign. Tron inspired a generation of animators to embrace computer graphics. Pioneers such as Pixar co-founder John Lasseter credit the movie as a major influence. “Without Tron,” Lasseter says, “there would be no Toy Story.”
The original Tron arrived at a time of massive technological change in cinema. And whether by accident or design, the sequel born of the 2008 Comic-Con footage arrives at a similar inflection point. Tron: Legacy moves the narrative of the original film forward 28 years, with Jeff Bridges, who played brash young programmer Flynn in the first movie, now reprising the role as an older, more meditative version of the character. The new movie pushes the technology forward as well. It is a single film that combines and refines almost every cutting-edge technique in cinema today: digital performance capture, advanced 3D cameras and sophisticated computer rendering of live actors into digital sets.
If any movie could succeed based on visual effects alone, it’s Tron: Legacy. But the film also comes at a time of serious debate about the stresses such technology puts on the movie-making process and its effect on the art of cinema. Shooting in 3D is expensive, the equipment is cumbersome, and audiences seem increasingly picky about what movies they are willing to pay extra to see in 3D. Plus, risky experiments in CG and performance capture can end up unraveling the virtual environments visual-effects artists strive so hard to create. Tron: Legacy is the type of movie that can shape agendas in Hollywood. If audiences love it, Tron will become the new standard for innovation. But it only takes a few missteps to turn a technological tour de force into a cautionary tale.
3D or not 3D
Given the success of lms such as James Cameron’s Avatar (R20 billion worldwide and counting) and Tim Burton’s Alice in Wonderland (R790 million opening weekend), 3D seems a no-brainer business decision for movie studios. Jeffrey Katzenberg of DreamWorks Animation and other studio executives have portrayed modern 3D technology as an evolutionary Nearly every moviegoer who has worn 3D glasses knows how great the format can be when utilised correctly (Avatar, Up) and how terrible it can be when slapdash conversions of 2D films go awry (Clash of the Titans, The Last Airbender). Now box-office trends are showing that audiences might be unwilling to pay more for substandard quality. Stereoscopic trailblazer and Avatar director James Cameron is happy that consumers are lashing out against bad conversions.
“Titans was a pretty good movie, but the 3D sucked,” he says. “It was sort of 2.5D. I want the studios to get spanked for making bad decisions.”
As it turns out, manufacturing 3D from 2D footage is hard. Our eyes, set 5 cm apart, naturally allow us to perceive depth. When we look at a tree, for example, we can distinguish that one branch is in front of another. Shooting that tree with a stereoscopic camera rig would capture those different depths, but in a conversion every object must be placed in a depth plane. There are a number of ways to convert movies to 3D, says Hugh Murray, the vice-president for technical production at IMAX. Some companies rely on automated computer algorithms, while others carve out objects and arbitrarily place them in different depth planes. Both methods can lead to underwhelming 3D.
At IMAX, conversion begins with segmentation artists, who draw precise outlines around the objects in a scene that will be in 3D; simultaneously, animators build wire-frame geometric models of those elements. The models allow converters to accurately set the objects’ depth within the scene.
Next, a camera algorithm computes where the film pixels would be if they had been shot with a second camera, which generates the left-eye view. The areas that second camera would have seen had it been on set are filled in by hand-painting or, in visual-effects shots, with information from the original animators. All that’s left to get the fi nal shot is aesthetic manipulation.
Murray says, “It’s a combination of sophisticated technology, a lot of grunt labour and creative calls that go into making a good conversion.” step in moviemaking, akin to sound and colour. Cinema operators have also bought into the technology in a big way – in the US, some 5 000 digital cinema screens are now equipped with 3D projectors. Yet there are already signs of a growing backlash. Innovative directors such as Christopher Nolan (The Dark Knight) and Zack Snyder (300, Watchmen) have expressed concerns about the limitations of the technology and are making upcoming features in 2D. Studios have also pulled back on plans to release some high-profile movies in 3D. Last October, Warner Bros changed course with Harry Potter and the Deathly Hallows: Part 1 and decided to release it in 2D. Plus, audiences burned by shoddy 3D postproduction conversions are voting with their wallets, opting for less expensive 2D versions or avoiding bad productions altogether.
Directors working in the medium certainly hope that audiences care about good 3D, because making stereoscopic movies can be a technical pain. Kosinski shot the actors in Tron: Legacy with an updated version of the same Pace Fusion 3D rigs James Cameron used for Avatar, out. tted with two Sony F35 cameras. “The Pace-Cameron F35 system is a very cumbersome beast on the set,” Kosinski says. “It’s such a large animal. It really informs the way you shoot.” As a result, Kosinski went for longer, more static shots. Those shots, however, produce better 3D cinematography, giving more depth to each scene.
Much of the elaborate camerawork that is routine with 2D productions is intensely difficult with 3D, where precision is the coin of the realm. “We’re really in the infancy of shooting 3D movies correctly,” says Eric Barba, visual-effects supervisor for Digital Domain, which created the digital imagery for Tron: Legacy. “We’re taking two cameras that aren’t really intended to be together and linking them with a mechanical system. Five years from now, when they’ve designed a camera that just shoots stereo, this is going to seem like rubber bands and glue.”
Capturing 3D footage on camera is only part of the process. Many of Tron: Legacy’s shots are 90 per cent CG. During pivotal light-cycle races, for instance, actors rode bike rigs on a soundstage, then their movements were rotoscoped by computer and animated into a virtual environment that was completely computer-generated.
To directors such as Kosinski and Cameron, this immersive experience is what separates 3D as a cinematographic tool from 3D as a gimmick – a distinction they believe audiences can see and should care about. As Tron: Legacy producer Sean Bailey puts it: “When they come in to see a 3D movie, they should get what they pay for.”
A tale of two bridges
The premise of Tron: Legacy is ironically similar to the process of making it. Kevin Flynn has again been sucked into a virtual world of his own creation, and his longestranged son, Sam (played by Garrett Hedlund), follows him into the software. Shortly after Sam enters the computerised game world of Tron, he comes face-to- face with his father, whom he hasn’t seen in 20 years. Yet something seems off , since apparently his father hasn’t aged at all. In the movie, it’s a trick of the software – the real Kevin Flynn is trapped elsewhere in the program, while a youngerlooking digital doppelgänger named Clu has emerged to become the virtual world’s villain.
According to Digital Domain’s Barba, turning the 61-year-old Bridges into Clu (modelled after Bridges at age 35) was “the most difficult thing that’s ever been tried in visual effects”. But it isn’t entirely without precedent: for director David Fincher’s 2008 movie The Curious Case of Benjamin Button, Barba prematurely aged actor Brad Pitt. Fincher used a similar effect in The Social Network, creating a twin of actor Armie Hammer by mapping his face on to a body double who acted alongside him.
As remarkable as those breakthroughs were, creating Clu was an order of magnitude more difficult: “Everyone has their own image of what they think Bridges should look like,” says Tron: Legacy animation supervisor Steve Preeg. “It’s very difficult to fool the human eye when it comes to seeing another human. We’re such social animals.”
To turn back the clock on Bridges, Digital Domain relied on a similar but more advanced version of the headmounted performance-capture camera arrays used in Avatar. Instead of Cameron’s single-camera set-up, Barba and his crew used four microcameras with infrared sensors to shoot the 143 spots dotting Bridges’s face.
The actor performed each scene first, followed quickly by a younger body double who synchronised his movements. Then Digital Domain used a “facial solve” program to marry the facial expressions of today’s Bridges (and his voice) with a digital actor modelled after the 35-yearold Bridges in 1984’s Against all odds.
The technology Digital Domain pioneered for Clu is a sign of things to come. Expensive cosmetic digital touchups of wrinkles and eye bags are already commonplace on big-budget movies. In fact, Preeg encourages actors to get themselves digitally scanned when they’re young, so that effects people like him don’t have to compensate for an actor’s age to recreate him in a computer. Given the culture of everlasting youth that pervades Hollywood, it’s easy to see how quickly the idea of ageless actor avatars could spread.
Is it worth the herculean effort involved to make such an intensive CG 3D epic? As a movie about the digital world, Tron: Legacy is a natural showcase for advanced technology. But it is possible to push the tech too far. Anything short of perfection will make the Clu character more distracting than compelling – indeed, negative reactions to early footage of Clu sent Barba and his team back to the shop, working down to the wire to refine the character.
Despite all the anticipation surrounding the movie, Tron’s stereoscopically filmed, motion-captured and digitally morphed star, Jeff Bridges, takes the long view of cinema technology. “Filmmakers are always looking for something to bring audiences deeper into the reality of the story,” he says. “Isn’t 3D without glasses coming in the next few years? Maybe that’s where it’s going. Or maybe you’ll just take a pill . . . Tron, the pill.”
How to spot bad 3D conversion
There’s no real depth.
When you look at a scene in 2D, your eyes use perspective and relative size to determine depth relationships. But if the cues from the left- and right-eye images are different, there will be little to no depth in the scene.
Objects wander around.
In real life, the objects closest to you move more quickly than distant scenery. If this motion parallax doesn’t match in both eye views, objects will appear to change position in the frame.
The depth planes are off .
Converters often misjudge where objects should be, which leads to one character looking past another or objects appearing as though they’re part of the background when they should be in the foreground.
The edges are incorrect.
Imprecise outlines leave parts of objects in the background – frizzy hair in particular – and make scenes look like early visual-effects shots, where lines appeared around things in composites.
3D home gear
Big stereoscopic epics are Hollywood’s way to get audiences ack into cinemas, but with the
right technology, you can bring a 3D cinema to your couch.
Manufacturers such as Panasonic, LG, Sony and Vizio make 3D TVs – the Samsung UN46C8000 LCD has performed well in PM’s testing – and the 3D effect really does work. But expect to pay extra for early adoption: 3D sets tend to cost anywhere from 10 to 30 per cent more than non-3D sets.
Shutter glasses are the biggest frustration of 3D home theatre. They’re expensive, not always included with the set and incompatible from brand to brand. Monster Cable’s Monster Vision Max 3D is one of the first attempts at universal glasses.
Most 3DTV manufacturers make 3D Blu-ray players, but a dedicated player isn’t really necessary. Thanks to a firmware update, the existing Sony PlayStation 3 can play both 3D movies and games, and it costs less than most dedicated 3D Blu-ray players.
Article: To read more about the vehicles and weapons in Tron. [click here]