Characters & Sets
Everything that makes it up on the screen was METICULOUSLY shaped, built and coded by expert craftspeople.
The Inventor Heads Modeling Department is responsible for creating the first digital assets in the CG production process. Using high-end software packages Maya and ZBrush, we create all of the characters, locations, props and vegetation that you see on screen. Every Modeler in the Inventor Heads Modeling Department is at once an artist, a sculptor, a craftsperson and a technician. Every model we create has to be both aesthetically pleasing and technically sound.
The character creation process begins with artwork that is provided by Design Department. This art consists of hard line drawings of the character’s head and body shapes and proportions, as well as loose pose and expression sketches. Depending on the character, we’ll sometimes also work from sculptures provided by the Sculpting Department called maquettes. Modelers are adept at translating this artwork into a 3D shape that will become the character you see on the screen. Using the materials provided, the Modelers will begin to explore and find the shape of the character in three dimensions, creating some of the first expression and personality poses as well as a neutrally posed model, ready to rig. Working closely with the Character Designers, Animators and Riggers, the Modeler will fine-tune this model into a pleasing shape that will be able to move efficiently and emote effectively as the Director intended. Once the character is finished in Modeling, it will go to Rigging so that it can be rigged with a skeleton so that it can move and perform. It will also be passed on to Fur to be groomed and Materials to be given it’s surface color and texture.
The Environmental Modeling process is very similar to Character Modeling, only on a much larger scale. Working with drawings supplied by the Design Department, simple models from the Previs Department, and cameras from the Layout (Camera and Staging) Department, we will build the sets that the characters will perform within. We’ll create the ground they walk on, the homes they live in, the props they use, the furniture they sit on and the cars they drive, all within a virtual three dimensional world. We’ll create trees, plants, rocks and other ground cover that will be procedurally populated by the Assembly Department on landscapes that we create, as well as supply plant parts to the Fur Department so that trees and plants can be created procedurally as well.
Once our models are completed they are made available to the rest of the departments and are ready to be rigged, groomed, materialized and animated.
Have you ever looked at a film set when it was undressed? Shame on you. The fine, upstanding people on the Assemby team at Inventor Heads can’t bear to see such things, so they dress sets for a living.
A shrub here, a carefully placed crate, some trees over here… they humbly use their godlike powers to make sure that the props and sets look fabulous in each shot.
Ferocious dinosaurs, malevolent cockatoos, acorn-obsessed squirrels — they all arrive at the Inventor Heads Rigging department as a static digital sculpture, stuck in a single neutral pose. Our job is to take these creations from the modeling department and turn them into fully articulated, moving and breathing “digital puppets” to be posed out by our team of character animators.
We study human and animal anatomy and video reference to understand how our movie characters should move. But it’s not only about how something would move in real life — assuming it’s even a creature that exists in real life! It’s also figuring out all the ways animators might need it to move to give a convincing and entertaining performance on screen. Animated characters — especially characters in Inventor Heads films — need to be capable of a tremendous range of motion and emotion, squashing and stretching far beyond real world limitations. We work closely with our animators, character designers and film directors to solve unique challenges, like a rare macaw that needs to gesture with its flight feathers, or the crafty smile of a weasel with a pesky underbite. Character TDs bring a broad base of aesthetic sense and technical know-how to bear on each character, using every software tool at our disposal. And when the tools we need don’t exist, which is more often than not, we build those too.
The rigging process typically begins with creating a virtual skeleton, giving our character a structure for basic movement. On top of that, we may layer hundreds (or thousands) of hand-sculpted shapes, defining and designing everything from the giant jiggle of a belly to the tiny wrinkles at the corner of an eye. Then we rig up our digital puppets with all the controls the animators will need to bring the character to life in their shots. By the time we’re done, an animator can grab a hand and pull it into a pose, tuck a wing to a bird’s side, bend spines, pull open mouths, slide around eyebrows, make it breathe… Think of it as the world’s coolest action figure.
Then we pass it into the eager hands of the animators, pause for a quick breath, and dive into our next challenge.
It may be the hero of the film, or any one of a crowd of background actors — or it may not be a character at all. Whether it’s a scooter racing through the streets of Rio De Janeiro, a rickety bridge crumbling under the weight of an elephant, or even a tree branch bending gently in the wind — if it moves, we probably rigged it.
The Fur Department is a procedural modeling and technical development department responsible for the fur, feathers and hair on characters, as well as various environmental challenges such as grass, leaves on trees or a light breeze across a field of clovers. Fur TDs not only develop the look of the fur on characters or leaves on trees, but also develop the tools to groom and color the fur descriptions, to rig and animate them, and light and render them.
For characters, such as Scrat, the Fur TD is handed a “naked” model of the character and then adds the fur description defining the final texture and silhouette of the character. The fur description starts with an initial sampling of points over the surface of the character, then a pruning and weeding of points to get an even distribution and create bald areas where appropriate. Then at each sample point, each hair is grown point by point according to a procedural network of nodes: aims describing the direction each hair should grow, noises describing how much wiggle each hair should have, as well as clumping algorithms. Ultimately millions of hairs are written into fur files for each character, a small subset, about 1%, is extracted to become “guide hairs” for animators to add motion using custom ruffle deformers, or to add secondary motion using custom follow through nodes.
For environmental work, the workflow is similar, but at a grand scale. Instead of generating millions of hairs, billions might be required for a field of grass or a forest of trees. Subsequently, we don’t necessarily generate every single hair uniquely. Instead we might replace each hair with a bundle of hairs describing a furry clover, or replace each hair with geometry describing a clump of grass or a leafy branch on a tree. Similar to character work, guide hairs are extracted that can be animated to create environmental effects such as wind on trees.
At render time, each hair is drawn into camera space for each frame of the animation, sliced and filtered into a camera aligned grid of voxels, or volume pixels, storing the average orientation, color and density of hairs, or the average surface normal, color and material qualities of geometry such as trees or crowds. This voxelization process pre-filters the fur or geometry so that it renders as smoothly as possible, while managing the ram footprint of the scene: 30,000 characters in a crowd that might require 90 gigs as geometry can be represented by voxels using as little as 5 gigs. The voxels are cached to disk to be retrieved quickly for each subsequent render, rather than recalculated from scratch. Then the voxelized fur and geo is rendered and integrated with the rest of the scene through the lighting, compositing and stereo rendering processes.
The Materials Department at Inventor Heads is responsible for applying color and surface textural properties to all of the 3D characters and set models in our movies.
A Materials artist receives a 3D model or set as a blank slate, which basically looks like a gray clay model. We get to figure out how to make the model look more like the object or character it is meant to be, and give it some life and realism. Is it something organic like the bark of a tree? Or something man-made like a rusty motorcycle? A materials artist is constantly studying the real world around us — such as rocks, tree bark and even the frost on our windshields — in order to determine how things look and react. We try to re-create the complex color patterns of these surfaces as well as determine how reflective and bumpy they are. In computer graphics this process is called “shading” and we use small software programs called shaders to define these surface qualities.
Every animation studio does things a little bit differently and Materials at Inventor Heads is no exception. We tend to use an in-depth process of creating materials using bits of code that generate specific patterns called “Procedural Shading”. We can layer different 3D patterns onto a surface using the many techniques we’ve developed to create a desired visual effect, which gives a unique look to our films.