CULVER CITY, CA - The third installment of any hit VFX movie franchise faces two key challenges: deliver visuals more spectacular and more innovative than audiences have seen before, and remain true to the characters and core themes that have made the films popular.
Spider-Man 3 is no exception. “In one sense we really had to top ourselves, to have Spider-Man move, look and feel better than ever before and make our buildings and environments look as real as possible,” says Scott Stokdyk. He has served as VFX supervisor on all of the Spider-Man movies, working with senior VFX supervisor John Dykstra on the previous two films.
“But the big challenge was the new villains — the new Goblin, Venom and Sandman — who represent a combination of effects and character animation we really didn’t have on the first two movies,” Stokdyk continues. “This intricate layering of effects with the character part of performances and how intertwined they were raised things to another level of complexity.”
While some 14 visual effects houses contributed to Spider-Man 3, Sony Pictures Imageworks (www.sonypictures.com/imageworks.com) was tasked with most of the complex shots, including the main character work and the villains.
NEW CHARACTERS
Venom, with his alien symbiotic goo, and Sandman, in all his different forms, particularly stood out. “We had to tie a layer of goo moving and transforming or sand forming different body parts into the emotional performance of the characters,” explains Stokdyk. “Both the symbiotic goo and sand required a huge amount of software development. Ryan Laney wrote the tools for animating the goo in Autodesk Maya, and Doug Bloom, Jonathan Cohen and Chris Allen formed the basis of the sand team who, programming-wise, spent over 10 man years of time writing code.”
The sand proved especially challenging because the animators needed “so much control over what individual particles could do in directing the sand behavior and more sophisticated fluid and collision behaviors,” he notes. “The core software had to work independently and as a group; it had to pass particles between different systems.”
In addition, working with sand meant dealing with “hundreds of millions of particles and required clever ways of culling out unneeded particles and being very efficient,” Stokdyk points out.
Spider-Man 3 reveals a new, dark side of Peter Parker when the same symbiotic goo that later affects Venom takes over Spider-Man’s body turning him into a black-suited character with enhanced abilities.
“Character animation director Spencer Cook gave Spider-Man’s body a bit more musculature and showcased him flying at heightened speed with faster falls and rises. Black-suited Spider-Man is stronger and more agile,” says Stokdyk. “Spencer took ideas from director Sam Raimi and Topher Grace, who plays Venom, and referenced footage of different animals and their strengths to push the abilities of the black-suited Spider-Man and then push further for Venom, the goo’s next host.”
This exercise “was more of an artistic development for Spencer and his animation team” than the technical task of crafting tools to bring Venom and Sandman to life, Stokdyk notes. “They drew from the abilities of the classic Spider-Man and pushed him in ways that kept him consistent with the last two movies. Imageworks has built layers and layers of character tools into Maya; it’s a great platform to build upon.”
INTEGRATION
Another goal for Stokdyk was to integrate CG with greater amounts of live-action footage than in the previous movies. “It was very important for the filmmakers and myself to get real footage of actors performing — even in giant stunt effects shots,” he reports. “In the last two movies we might have gone to all CG for action shots that were more complex than just take offs, impacts or landings. Nowadays, all-CG shots are very common, and in many ways they’re easier to do than incorporating live action with CG. For Spider-Man 3, if we had any chances to blend live-action pieces into CG elements, we would do it, even if it was difficult.”
Stokdyk explains the preference for integrating CG and live action today: “There’s only so much you can do once you’ve gone all CG. But if you can get a piece of live-action footage and CG and incorporate them, it grounds the shot and brings a sense of reality the audience can see and feel.”
He credits Go Stunts with being in “the forefront of breaking down the walls of what stunt people can do,” which furthers live-action capture. “People once talked about CG eventually replacing real stunts, but that’s not the case. Go Stunts has stepped up to the challenge and done some fantastic work.”
Stokdyk says Go Stunts (www.gostunts. com) used previsualizations “to almost motion control their stunt people with high-speed winches pulling them on wires against bluescreen or greenscreen. This encouraged Sam to use real stunts as much as he could while definitely keeping safety issues in mind.”
But live action and CG integration “required us to set up a pretty complex set of tools in Maya,” he reveals. “The idea is to take a performance and map it onto a 2D image plane in Maya. That gives the animators control to manipulate the performance, time warping it and moving it in 3D space. John Schmidt was really integral to making a lot of that work.”
Stokdyk muses that “it’s hard to believe that a CG Spider-Man swinging through a CG city would be the easiest thing to do” today. “In the first movie, it would have been the hardest.” Advances in technology and capabilities in the seven years since work began on the original Spider-Man film meant animators on the new movie “almost took for granted CG skin work, replication of digital doubles and virtual backgrounds, which had been an incredible amount of work on Spider-Man 2,” he reports. “On Spider-Man 3 these things were almost a given.
“It’s kind of great knowing that the hardest challenges of the current show will be easier on the next one,” he smiles. “I can only imagine what the next five years will bring!”
THE POST
Post production for Spider-Man 3 was also complex but, like VFX, it built on the experience of the two previous films. Among the efficiencies it achieved was the improved “time management” of Sam Raimi. “We were able to keep [him] on the Sony Pictures lot by utilizing Technicolor Digital Intermediate [TDI] here as well as our in-house sound facility,” points out post production supervisor John Naveira. He served as post supervisor on the first two Spider-Man pictures and reassembled the same editorial crew for the new film.
Links between production editorial on the lot and Sony Pictures Imageworks also made their extensive communications more effective. A fiber optic network, set up via Avid Unity, enabled VFX to be sent directly to the screening room and TDI to speed up the approval process.
When production editorial, under Spider-Man 3 editor Bob Murawski, turned over cut sequences to Imageworks, Allen Cappuccilli, the senior visual effects editor, was able to import the data immediately without having to copy media. This gave Imageworks the ability to instantly see the same material that production editorial was working with on a complement of Macintosh-based Avid Adrenalines.
Imageworks editorial, in turn, could distribute the sequences to their proprietary edbot (remote viewing stations) system. Production editorial also sent Imageworks count sheets in the form of raw data using the network. Imageworks editorial imported this data and posted it in the facility’s proprietary editorial tracking system. It could then be seen company-wide on Imageworks’ internal Spider-Man 3 Web page.
Negative was scanned and an EDL created by the editors was used to conform the cut at TDI on Autodesk’s Discreet Smoke. Color correction was done with the da Vinci Resolve system in 4K resolution and then, to complete the DI process, files were converted to film-out negative for release printing.
Naveira says his biggest challenge was to meet the deadline for “turnover/lock picture and sound to start foreign dubbing and subtitling without sacrificing the integrity of the film. We knew well ahead what reels and VFX would take more time and worked around those sequences. Flexibility was key, [since] some VFX drop-ins were more challenging than others due to the additional sound design and mixing needed.”
A further challenge was “working around marketing and publicity’s needs, which required the use of our color timer, mixers, sound and picture editorial who were simultaneously working on the feature to keep continuity,” he reports.
Still, the web of experience, talent and technology fabricated by Naveira and his colleagues held strong — another lesson learned from Spider-Man.