Director Bekmambetov finds the idea of “fix it in post” a horrific concept—to him, visual effects are intended to take the shot further than captured in-camera, not to wholly create something that didn’t start on the set.
Bekmambetov explains, “For me, it is the emotion that is important, not really the effect. It may be a little old school, but this is how I get what I need from my actors and crew. I don’t use effects to make up for what is not there. If it’s in the character, in his or her emotions, it will be on the screen.”
Longtime collaborator/editor and second unit director of Wanted, DMITRI KISELEV, has worked with Bekmambetov in Russia for the last 10 years and is only just now starting to understand his friend’s vision. Kiselev describes, “Timur breaks so many rules, but he is always looking for something natural, something real in a shot before he even contemplates using CG to complete it. He will create his visual effects around an existing shot.”
Producer Lemley picks up: “When we were drafting the screenplay and developing the sequences, Bekmambetov would go away to his ‘science lab’ and come back with pre-visual sequences that illustrated exactly how he would shoot and what the focal point of each scene would be. That focus was always specific to the emotions we wanted the characters to convey and the impact those feelings would bring to the visual effects that completed the scene.”
Visual effects producer Farhat adds, “He uses his pre-vis as a tool—he uses it to educate and discuss with those who are trying to capture the action, but he doesn’t cripple people’s visions with it.
“There is no bad idea with Bekmambetov,” Farhat continues. “He knows there is more than one way to do something and he’s very open to ideas, but as experienced as he is, he understands that ideas always need to be fresh.”
The director has his own visual effects house in Moscow, Bazelevs (a production and postproduction/effects facility), which served as a “clearing house” for the effects created for Wanted. While not all of the effects were completed under its roof, Bazelevs maintained an overall watch on all out-of-house work.
Farhat says, “One of the biggest challenges making a film where you’re using multiple facilities is keeping the continuity, the look and the style consistent all across the board. Bazelevs works not only all over Russia, but the world. The visual effects were split up among various facilities—some worked on modeling, others created texturing, others animating and so forth. So Bazelevs created this digital pipeline, a digital asset management system, where they could actually follow the progression of any shot and compare it to the progression of any other shot in a sequence or anywhere else in the movie. The separate houses really acted as one house—a virtual company.”
One effect that stayed under the Bazelevs roof, however, was the creation of computer-generated stunt doubles. Even with Bekmambetov’s insistence on shooting as much of the story as can be captured in the real world, there were certain things (the height of assassin mode and high-risk action sequences such as running on the roofs of moving trains) that could not be filmed—even with the help of the best stunt performers and the most advanced wire works. To complete those scenes, digital stunt doubles were created through cyber-scanning.
Think of cyber-scanning as an enormous 3-D copy machine, which rotates around the actor (for around 15 seconds) and creates a 3-D model of that person. The model is transferred to the computer in a CG mold, which then has to be “rigged” (inserting skeletal and muscular systems and texturing the exterior) and fitted with wardrobe (which has also been scanned). That rig then has to be “taught” to replicate the way the actor—and the actual stunt double—move. The result is a digital double that doesn’t balk at engaging in the most life-threatening stunts imaginable.
In addition to high-end film cameras, Bekmambetov adapted a still technique for use in several sequences, particularly the “L” train chase. The director and cinematographer Admundsen employed a series of six synched 35mm cameras mounted on a special plate that could rotate 180°. The cameras’ lenses angled to capture the horizontal top of the train as it drives through Chicago and overlapped frames to produce a contiguous all-around view. When seamed all together and matched with green-screen shots of the actors, the result is a scene with a cylindrical or spherical texture and a complete 180° view of the nonstop action atop the speeding “L.”
Farhat concludes, “This technique really freed us up a lot. We’ve all seen tiled stills, where you take a series of stills and you match them end-to-end spherically, basically freezing the action and rotating the point-of-view. But in this case, we’re doing it with moving footage. I’d say it was one of the toughest sequences in the show, because we’re basically taking real actors and putting them in a world that doesn’t really exist.”
No comments:
Post a Comment