Previously we focused on the integration of CG generated furry creatures in the real world; now we are going to move from fur to hair, discussing the importance of it, starting from japanese and traditional animation, passing through CG milestones of early 2000 and ending with latest movies like Rapunzel, Brave and Frozen that represented great challanges not only for animators, but even (and maybe primarily) for scientists and engineers.
In our documentary we didn’t focus excessively on technical aspects, trying to avoid to make it boring (we hope we managed to do it) but, even for a matter of respect to experts that are behind these productions, we want to underline some interesting features.
In the table (click on it to magnify) we tried to quantify the difference between each product and point out physical aspects like the role of keyhairs and pyisics laws that were exploited.
FizT was the first attempt to move hair according to real world physical laws. The implementation of volumetric method in FizT was a completely new scientific experiment, since it was made an outstanding study by Lena Petrovic, Mark Henne and John Anderson, in which the volumetric representation of hair mass is obtained by calculating the keyhair density at each vertex of the voxel grid. Each keyhair was originally represented as a control hull of a B-spline. To create the volume, they summed up spatial influences of keyhair control vertices. The influence of a keyhair vertex was consider as a 3D tent function, with value one at the keyhair vertex location and linearly decreasing along each of the coordinate axes, becoming zero at a distance equal to the grid unit. We invite you to read the full article here.
Kelly Ward, using the traditional knowledge of the animator Glen Keane, was the senior software engineer that lead the team in the creation of Rapunzel’s hair. Besides the use of Hooke’s Law in the creation of hair they had to reach a compromise between the physical accuracy and the visual effect that 70 feet of hair, weighing 60 to 80 pounds, had.
Dr. Ward explained her work in an incredibly humble way in this video.
Claudia Chung was the simulation supervisor for Brave (she worked with Lena Petrovic here, and they worked together for The Incredibles too) you can read a complete interview made by Art of Vfx here, she explains that the development of the sowtare Taz took three years of work on R&D, she also states that they made sure the simulator could actually finish the film without artists having to constantly clean up and hand-tweak results. Other interesting information about physics studies made for the movie (not only for Merida’s hair) are listed here.
The last work in this world is Frozen. Apparently the software Dynamic Wires realized for Rapunzel wasn’t enough, so Andrew Selle and his team developed another software for Walt Disney Studios called Tonic. Tonic enabled artists to sculpt the character’s hair.
In this interview Gregory Smith, the rigging supervisor, explains a lot of interesting features in simple terms talking about rigging in general and going into detail of the software.
By representing the evolution of the software’s performance with a graph we noticed that it follows an exponential growth, but in this representation we didn’t include the length of hair (that makes a big difference in case of Dynamic Wires).
So this field is made up of a lot of interdisciplinary collaborations, as we’ve already seen with Francesco Giordana, and it is very far to be considered dead.
Chiara Sapio, Edgar Pironti