VFX Pro Feature

The Post Pipeline on V for Vendetta
By Matt Hurwitz
May 16, 2006, 11:00

The Wachowski Brothers' "V for Vendetta" (Warner Bros./Silver Pictures/Virtual Studios) tells the story of a masked vigilante known as "V" who seeks revenge for both himself and his country against a government that maimed him, and, as he sees it, England as well. When he rescues an ordinary young woman (Evey, played by Natalie Portman), she joins his struggle against the forces of oppression. With a campaign that fights these forces with acts of terrorism, V attempts to convince the English people to rebel. V's acts of terrorism include the spectacular destruction of the Houses of Parliament and The Old Bailey.

Based on the graphic novel of the same name by Alan Moore and David Lloyd, the film was directed not by the Wachowskis, who decided to take a break from helming for this project, but by their 1st AD, first-time director James McTeigue. McTeigue is a veteran of the Matrix series, as is V for Vendetta visual effects supervisor Dan Glass. The move from 1st AD was an easy one, McTeigue says, particularly since the crew members were all familiar faces. "Having people around that you know you can entrust to do certain jobs really frees you up, creatively, on a day-to-day basis. I was lucky to have a repeat experience."

Besides Glass, accompanying the director on his journey were Production Designer Owen Paterson and Director of Photography Adrian Biddle, BSC, who died in December of a heart attack at age 53, after the film was essentially completed.

McTeigue's vision for the film included a style that draws heavily on the contrasty look of the graphic novel, as well as film noir movies of the '40s and '50s and the work of 20th century painter Francis Bacon and 16th century artists Michelangelo Caravaggio and Matthias Grünewald. "Adrian [Biddle] really understood the look of the graphic novel. He did an amazing job in just letting things fall off to black."

Production Designer Owen Paterson notes, "We tried to create a London that is recognizable, yet frozen by having become this totalitarian state." He and the production team settled on using a palette of grey tones to evoke the bleak, regimented pall that envelops the city and its citizens.

Expanding the DI Process

All of the looks in the film made use not only of Paterson's creative art direction but also of Biddle's insightful application of the digital intermediate process. "With the timeline we were shooting in, it proved very hard to get the feel of the graphic novel exactly right," says Visual Effects Supervisor Glass. (The team had nine months to complete the film and just seven weeks to complete the effects once the edit was locked.) "We decided early on that we should go to a DI to do that kind of fine-tuning."

Colorist Adam Glasman of Framestore CFC in London was brought on board, then, fairly early, to make continual adjustments and feed information back to McTeigue and Editor Martin Walsh. "Normally, it's a process that would come at the end, once you've done your film and your edits are locked," Glass notes. "You'd conform and put it together once. We conformed the reels as they were cut. Although that involves quite a bit more work, I think it's the way that more films will be done."

Knowing the DI would be part of the process, Biddle purposely shot a number of scenes brightly lit, to capture as much as possible on the film negative. "Adrian really understood what can be done digitally in the DI process," McTeigue recalls. "He always had an eye on that when lighting a scene. If you've got the information on the negative, you can really go places with it."

Flashback Sequences

The team conducted quite a bit of testing to achieve the look McTeigue was after for the film's flashback sequences. The first flashback depicts Guy Fawkes' original "gunpowder treason" attempt at blowing up Parliament from a tunnel burrowed beneath it. Where Fawkes failed in 1605, V succeeds sometime in the near future.

For the flashbacks, "I wanted a zoetrope-y, old look," McTeigue says. "[2nd Unit Director] Harvey Harrison shot a lot of the flashbacks with an orange gel and diffusion. Then we pushed it even further in the DI." The flashbacks move toward a more realistic look, he says, as the film progresses to the film's present time.

"We applied quite a strong look to the first one, the Guy Fawkes segment," says Framestore colorist Adam Glasman. "We drained a lot of color out and applied a golden, high-contrast look. We then applied diffusion to both the highlights and lowlights. That process was done, to a greater or lesser extent, to all of the flashbacks."

V's Mask

While McTeigue and Biddle made full use of Framestore CFC's color grading abilities, the same systems were also put to work handling some matters traditionally done by a visual effects department. "You can actually use this system for some low-level compositing," explains Glasman.

The company's color grading system, called Baselight 8, was developed by Filmlight, originally part of CFC, for DreamWorks' Chicken Run in 2000. Baselight provides real-time processing on films scanned at 2K. "It's basically a software color corrector running on its own dedicated hardware. Because it's software-based, we can apply an unlimited number of grading changes to a single shot."

For V for Vendetta, the Framestore DI team began by making a digital version of the feature, which was shot on Kodak 35mm film stock with cameras including ARRICAM ST, ARRICAM LT, ARRI 435 ES Xtrem and ARRI 235. Over several weeks in the fall of 2005, Glasman corrected the color with Cinematographer Adrian Biddle and Director James McTeigue, with occasional visits from the Wachowskis.

One of the most significant applications of the system to V for Vendetta was in the shadows on V's Guy Fawkes mask, which he wears throughout the entire picture.

Production Designer Paterson collaborated with McTeigue and Art Director Stephan Gessler on the creation of V's mask. Paterson's design was modeled on V's iconic visage from the graphic novel, which Illustrator David Lloyd based on the eponymous masks worn in tribute to Guy Fawkes. But as drawn by Lloyd, V's mask takes on different moods and expressions from frame to frame. McTeigue opted to create a "fixed" façade rather than using CGI or a flexible mask that could be manipulated to form expressions.

"I wanted the face, even though it's very distinct, to have a 'universality' to it," he says. "I knew that if we achieved the right look for the mask, we would be able to tonally and atmospherically change the way it appears on camera through the lighting design and Hugo [Weaving]'s performance."

"It's a director of photography's worst nightmare," notes Glass. "You've got a character that's dressed completely in black, but with this white mask or face. You want to retain some information in his black clothing, which means pumping lots of light at it, but then the mask tends to get washed out," causing naturally appearing shadows to disappear.

While Biddle was quite adept at handling the problem in close-ups (which featured little of the black clothing), the problem in wider shots was unavoidable. Using the close-ups as reference, Glass took advantage of Framestore's DI capabilities and had Glasman build new shadows for the mask. "We did a 3D cyberscan of the mask, created a 3D model in Autodesk Maya, tracked the model to his head motion [using 2d3's boujou and RealViz' MatchMover] and created new shadows for it," Glass explains. "By tracking it, you can position lights and create these shapes. The alpha or matte channels [created in Pixar's RenderMan] were then sent to the DI, where [Colorist Adam Inglis and Senior Compositing Artist Jonathan Fawkner] were able to control the level and intensity of those shadows. Normally, this would have been composited in the visual effects world, but their system handled it quite nicely."

Fingerman Alley

Another long sequence, which takes place in a smoky alley (London's Fingerman Alley), caused the Framestore DI team some difficulties, which were once again solved by integrating the grading process with the visual effects work. "Usually, you grade everything, send it to visual effects, and they'd add atmosphere," says Inglis. "We tried that at first, but we felt it wasn't working. It was quite a long sequence with a lot of cuts, and we needed to see the atmosphere in context."

The visual effects crew rotoscoped the foreground characters and rough shapes of the alleyway to create mattes and created visual effects plates of smoke. The smoke plates were a combination of practical library elements and smoky noise generated in Shake. Because the compositing was dependent on the grade, changing as the smoky atmosphere was increased or decreased, doing the work at Framestore's Digital Lab made a lot of sense. "On the Inferno, we could have made the changes on a background plate, but then the grade would have been made bluer, so we would have had to redo the smoke," explains Fawkner. "On the [Baselight] grading system, we just dialed down the smoke. So, it was slightly more interactive."

The Shadow Gallery

The DI system was also put to work to selectively bring down portions of the background set while leaving the actors more prominent-in V's labyrinthine Shadow Gallery, for example. "This was one of the instances in which Adrian knew the DI process would be applied, so he shot it brighter than he normally would have simply to capture all of the information that was there," says Glass.

"The trick was to get that kind of gloomy look without losing this fantastic set they had spent so much time building," notes Glasman. Using the system's tracking software, he drew various shapes, keying areas to lock those shapes to the camera moves as it would track through the shot. "In a few of the shots, I actually drew the shadow on his face. We would then lock the shadow mattes that I created to the face, and move with face. We were trying to create pools of light, so they had to follow the set, not the actors."

Throwing Knives

While Framestore CFC was responsible for creating the digital intermediate, atmosphere effects and mask shadow effects for V for Vendetta, the film's producers enlisted engineers and artists from Cinesite, Double Negative and Baseblack to help deliver the film's bleak and repressive visual tone within the tight seven-week timeframe. Baseblack performed wire and rig removal, while Double Negative created most of the monitor and television screen inserts.

As the lead visual effects studio on V for Vendetta, Cinesite delivered more than 150 of the film's nearly 500 effects shots. Cinesite artists slightly retouched the London skyline, removing advertising signage and any splashes of color or brightness, and created a massive and imposing video wall in Picadilly Circus. Cinesite's biggest artistic challenge was in devising the miniatures and effects that convey the film's ending sequences, including a massive crowd of freedom fighters, a showdown with knife throwing and the explosive demolition of several London landmarks.

In one of the key effects sequences, which takes place near the climax of the film, V takes on a group of attackers with his shiny knife collection. As the weapons twirl across the screen, they're followed by interesting knife trails, an idea conceived during the filming of the scene.

"James had this idea, from reading the graphic novels, to have [an effect resembling] motion blur trails," Glass explains. Concept Artist George Hull, who had worked on The Matrix, created concept drawings for the effects. These drawings were given to Cinesite UK's digital division as reference in creating the trails. "As the effect evolved, we introduced the idea of the trails being like smoke in the atmosphere in the way they responded to light. When things and people pass through them, they disperse and wisp away."

The Cinesite team built a 3D replica of the knife and animated it to rotate. Seven to eight effects passes were necessary to create a trail with the correct amount of wispy translucency.

Cinesite Delivers a Cast of "Thousands"

Another CG task involved the duplication of hundreds of extras dressed in identical white masks, cloaks and hats. The film's climactic sequence takes place on Whitehall, the iconic thoroughfare running from Nelson's Column at Trafalgar Square to the Parliament Buildings and Big Ben. In the sequence, thousands of V look-alikes confront the government police force.

The production team wanted to avoid using an entirely computer-generated crowd, so Location Supervisor Nick Daubeney secured unprecedented permission to close the street for filming between the hours of midnight and 5 a.m. for three consecutive nights. This timeline allowed the production just four hours of shooting per night, given the setup and removal of equipment, personnel and the production's vehicles, including two army tanks.

With this four-hour work window, things had to move fast, even with five cameras capturing the event. "There was, obviously, a lot of planning for a long time," explains McTeigue. "But, with that short window, it just becomes about grabbing everything you can and fashioning it together," which, it turns out, is done quite effectively.

While the production team shot 500 extras dressed as V, it was the Cinesite effects team that increased the apparent number of freedom fighters into the thousands by replicating those real performers in compositing and also adding fully CG characters. The Cinesite compositing team cut and pasted people from different plates and takes to build the crowd and make the scene believable.

"The faces you see in the foreground, obviously, are real, but those in the far background are doubles," Glass notes. "It's not often you get asked to do hundreds of digital doubles that are actually meant to look the same."

More complicated, however, were aerial shots of the crowd, which Cinesite created in 3D. Character animation was created in React, Cinesite's in-house crowd simulation software. Motion-capture data for realistic walk, run and climb cycles was acquired by Artem Digital with an 18-camera Vicon motion capture system. React applied animation created from motion-capture data to the 15 different crowd models (both male and female of different heights and builds). The animation was rendered in RenderMan.

Exploding Landmarks

Perhaps the most impressive of all the effects in the film involves the explosion of London landmarks The Old Bailey (the criminal court building), Big Ben and the Houses of Parliament. Cinesite model unit supervisor José Granell supervised the construction of the 1/7th-scale Old Bailey and 1/10th-scale Parliament and Big Ben cast plaster models.

While some CG effects were later composited with footage of the models exploding, it was important to the filmmakers that the explosions, which carry great symbolic value, be as realistic as possible, so they opted for the practical effect of detonating physical replicas of the buildings over CGI.

"The models provide a real, tangible environment," Granell explains, "and when you're dealing with physical elements such as water and fire, and especially pyrotechnics, you get a better look when you have real, physical events taking place." The miniatures enabled a better agreement of visual proportions between the size of the buildings and the pyrotechnic events being filmed.

The models were assembled in 11 weeks at Shepperton Studios after a great amount of testing to determine the proper composition to deliver a realistic onscreen detonation. "They're built in the same way that the real structure is built, with large masonry stones at the base, which were added one by one," Glass says.

Granell also supervised unit photography of the miniatures. Visual Effects Supervisor Glass and the V for Vendetta miniatures unit, led by Granell, spent ten days detonating these large-scale models using a high-speed motion control camera in front of a greenscreen at Shepperton. "It's important, obviously, to increase your frame rate when you shoot miniatures in order to compensate for gravity and to give a sense of scale and weight. But sometimes-and José agreed with me-you can overcompensate, which can make miniature work read like miniature work because it's too slow."

To create the background plate, the Cinesite team shot high resolution, high dynamic range stills of the actual London location at varying times of day and in varying light conditions under the supervision of Cinesite visual effects supervisor Matt Johnson. With the guidance of Matte Painting Supervisor Dave Early, the HDR imagery was composited into a final matte and used as the background for the miniature explosions.

The explosions themselves were enhanced with both live-action and CG elements. Cinesite artists shot live-action debris, pyrotechnic elements and water splashes, while the CG department created additional debris, sparks and splashes in Maya. Most of the fireworks in the sequence originated as stock shots from the Warner Bros. library that were scanned by Pacific Title. These reworked live-action elements were combined with CG fireworks created in Maya's particle system.

All of the effects help McTeigue build a story much like a politicized Phantom of the Opera, or, as McTeigue says, like Cocteau's Beauty and the Beast. "Like those stories, it's about two individuals who meet through fate or circumstance, and how that can possibly change your life in a way that you never expected."

© Copyright 2003 by United Entertainment Media, Inc.