AnimfxNZ 2006: Animation & VFX Galore in Middle Earth
King Kong -- A Monkey in the Making In the pursuit of realism, Kong was based on an anatomically correct skeleton and muscles. Aitken, Weta Digital's pre-production/R&D supervisor on King Kong, demonstrated the amount of detail that went into the geometry right down to Kong's teeth.
Aitken also discussed the development of Kong's fur. Because of the sheer volume of fur required, proprietary software needed to be created. Also, in order for the fur to look realistic, it too had to be animated as well as the body -- moving and reacting to its environment. The key, said Aitken, lies in treating the fur as volume, using layers of mesh and volumetric simulations.
The building of Kong lasted 2/3 of digital production time, while the remaining 1/3 was used for putting him in the shots. Aitken defended this ratio, maintaining it is important not to rush modelling and risk cutting the building time too short. It is better to be certain the model can account for any and every need required in the plot -- otherwise, once "shooting" starts, it becomes very costly and time-consuming to go back and make any changes to the model.
King Kong, featuring a huge gorilla as the lead, set new benchmarks in movie animation. The decision to make Kong entirely CG was based on the need to ensure continuity, flexibility and control of performance during the production process. However, Weta Workshop was still involved in the process, building reference models for computer modelers to base their computer-generated models on.
Motion Capture for King Kong However, the complexity of facial animation is huge; so many things happen simultaneously, triggered by the autonomous nervous system: eye motions, our response to our environment, all resulting in multiple, fleeting micro expressions that communicate to an audience what a character is experiencing, feeling or thinking.
For this reason and to ensure consistent performance of the animation, Weta looked at facial motion capture, the technique of using a live actor to drive the performance of a computer-generated character, as opposed to using an animator to create movement and expressions. Weta had already experimented with this process to drive the facial expressions of the Gollum character in The Lord of the Rings movies.
However, as Letteri, Academy Award-winning visual effects supervisor for Weta Digital, explained in the session on "MoCap and the Future of Performance Capture," back then, there was not enough time to complete the system and while MoCap was employed for Gollum's body movements, his facial expressions had yet to be keyframed by animators.
By the time Jackson started to make King Kong, the MoCap system was ready and actor Andy Serkis' facial expressions could now be translated into an animated Kong. The facial performance of Serkis, the actor whose performance was also the basis for Gollum, was captured and transferred through purpose built software to translate the facial structure of Kong into the corresponding expressions of a gorilla.
Sagar, who developed the facial animation system for Kong, gave an in-depth description of the process and FACs, the Facial Animation Code System, initially defined by Ekmann and Friesen, which the system was based on. This model defines groupings of facial muscles, and how they work together to create emotional expressions. A facial motion editing team then worked with the computer generated face, which had 200 controls for adjusting the performance: slowing or toning down expressions, and relaxing muscles where necessary for final touches.
In addition to assuring every effort was made to make Kong look real physically, Jackson also wanted to be able to build a huge amount of emotions into his lead character's face and eyes. It was imperative for Kong's character to be based on emotion -- this called for an actor-driven performance.
























Post new comment