Sponsored By

Postcard from GDC 2004: Building Gollum

GDC attendees gathered during the lunch hour today to listen to Bay Raitt, Creature Facial Lead at Weta Digital, speak about his 4+ years as mastermind of Gollum's facial system in the Lord of the Rings trilogy.

Using dozens of proprietary images and video clips, Raitt explained to the audience how he was able to develop the extraordinarily complex facial system behind the show-stealing CG character.

Brad Kane, Blogger

March 24, 2004

3 Min Read
Game Developer logo in a gray background | Game Developer

 

GDC attendees gathered during the lunch hour today to listen to Bay Raitt, Creature Facial Lead at Weta Digital, speak about his 4+ years as mastermind of Gollum's facial system in the Lord of the Rings trilogy.

Using dozens of proprietary images and video clips from the Lord of the Rings movies and from his own laptop, Raitt explained to the audience how he was able to develop the extraordinarily complex facial system behind the show-stealing CG character.

In order to quickly achieve the vast array of facial expressions of which Gollum is capable -- and to allow animators the ability to transition seamlessly those expressions, without working around expressive "dead zones" -- Raitt came up with a technique he calls "Combination Modeling."

Raitt's vision for Combination Modeling involved the creation of a slider-based GUI that lets animators use 64 independent facial controls to transition between the 935 morph targets on Gollum's face. In what Raitt refers to as "Combination Logic," each morph target can be represented as a set of coordinates in multi-dimensional space, and the transition from one target to the next involves translating those coordinates appropriately. Although very complex at a technical level, Raitt's GUI allowed animators to take Gollum through one extreme facial expression after another, all while maintaining his full acting range from pose to pose.

Combination Logic in itself created enormous complexity issues for Raitt and his team, and in order to track the large number of morph targets available for any given string of facial beats, Raitt created an iconographic tracking system that matched morph targets with simple hand-drawn "glyphs." The glyphs were in turn each assigned a number, and the end result was a system by which Raitt could use simple integer combinations -- say, 272 x 301 x 262 -- to refer to very specific facial expressions.

Raitt emphasized that Combination Logic is so efficient for the detailed sculpting of CG faces that he fully expects "Combination Modeler" to be its own job title in the coming years.

Raitt's expertise in facial sculpting stunned even the most seasoned of facial riggers, and it's no surprise: in his 4.5 years at Weta, Raitt has sculpted nearly 10,000 morph targets for Gollum's face, and singlehandedly developed an entirely new paradigm for the manipulation of CG faces.

Read more about:

Features2004event-gdc

About the Author

Brad Kane

Blogger

Brad Kane is a freelance writer focusing on the film and videogame industries. He has worked on several of the top-grossing animated movies of all time, and on a number of upcoming film and interactive projects. He can be reached at [email protected].

Daily news, dev blogs, and stories from Game Developer straight to your inbox

You May Also Like