Navigation Bar - Text Links at Bottom of Page


The Oompa-Loompas and the Squirrels
Bringing the Oompas to life involved the full cooperative effort of all the effects artists on the film, but it all began with one man, the Oompa-Loompa prototype: Deep Roy. 

If five, six or 20 Oompa-Loompas appear in a scene, Roy played all of them. In separate takes, and from different starting marks, he would act out each single part on the motion capture stage, whereby his body and facial movements were recorded in the computer. If the scene was one in which the Oompa-Loompas join to dance and sing a number about the fate of each wayward child on the tour, the entire routine would be meticulously choreographed for months to composer Danny Elfman's music. Then Roy would perform the steps from each individual spot on the line, subtly adjusting his gestures and expressions from one to the next so that when the collection of images were later joined together onscreen he would have created an entire troupe. 

"I think of it as doing nineteen second takes,” offers Roy, whose extensive training for the roles included daily pilates sessions and dance classes. "The most challenging part was trying to remember my position from one performance to the next, counting in my head and remembering at what point to turn or where to look. It was a lot of rehearsing.”

"This was extremely tricky, partly just for the volume of shots it created,” says Chas Jarrett, Visual Effects Supervisor of The Moving Picture Company, one of the companies that joined the production team to work on Oompa-Loompa footage and contributed nearly 500 shots. "Although effectively the Oompas look alike, we've slightly altered the facial tones of each. Their hairstyles may be a little different and each performance is slightly varied from one character to another.” 

What the relatively new facial-capture process offers over standard animation, Jarrett believes, "is subtleties around the eyes and mouth shapes, the way the jaw moves and the skin stretches around the nostrils when he speaks. Those are the kinds of details that animators find most difficult to recreate. And here we get it free, with Deep's performance.”

As if that wasn't complicated enough, the Oompa-Loompas are only two-and-a-half feet high, so Deep Roy's virtual image had to be proportionately reduced. This wouldn't be a problem if he played his scenes solo but the Oompa-Loompas are in nearly every frame of the film and interact with all the human characters in various settings. 

To illustrate how complicated it was just keeping track of the scale issue, Alex McDowell offers an impromptu checklist that sounds like one half of an Abbott and Costello routine: "Our environments had to be in two different scales. We had to be constantly aware of Oompa-Loompa scale, which is 30 inches high. Hand tools, controls, pathways and architecture had to conform to Oompa height. A lot of the time that's Deep Roy, who is actually twice that height. So there's Oompa scale and Deep Roy scale. Oompa scale is sometimes the same as human scale, with tiny props that stay tiny in human scale but appear larger in Deep Roy scale. Sometimes you have Deep sitting in a human chair so you build a double-size chair for him so he appears half-size to humans; sometimes Deep is in the Oompa environment in which case you build a set for Deep, at his scale, and a half-size set for Willy Wonka so that he appears large in the Oompa environment. The terminology alone is hard to get a handle on.” 

Partly to provide a scale reference point in some scenes as well as a focal point and something for the actors to react to, the production enlisted animatronics and prosthetic makeup effects specialist Neal Scanlan, of Neal Scanlan Studio, an Oscar winner for his work on Babe. 

"Our goal,” says Scanlan, "was to make a photo-realistic Oompa-Loompa.”

He and his team assembled five completely motorized puppe

Next Production Note Section


Home | Theaters | Video | TV

Your Comments and Suggestions are Always Welcome.

2018 6,  All Rights Reserved.


Find:  HELP!