Character Modeling with Andrew Klein

Take a look at this site (below). It focuses on simple diagrams showing which muscles effect the face when creating expressions.

Now that we have created our Characters, we will flush out the modeling further by employing Vowel, Consonent (collectively refered to as Phonemes) and Expressive movements to the face. These three motives for facial movement can be summed up in the following definitions:

A vowel is a sound in spoken language that is characterized by an open configuration of the vocal tract so that there is no build-up of air pressure above the glottis.Vowels letters in the English language include A, E, I, O, and U representing the closing of the mouth in that order. By combining vowels or by stretching them, AE, EE, AH, EH, EWW, IH, OOO, OI, and the Schwa are pronouncable as well.

A consonant is a sound in spoken language that is characterized by a closure or stricture of the vocal tract sufficient to cause audible turbulence. Consonant letters in the English alphabet are B, C, D, F, G, H, J, K, L, M, N, P, Q, R, S, T, V, W, X, Z, and usually Y, with the additional CH and SH, LL, and RR sounds beings spoken. Consonants can be nasal (through the nose), stop (complete obstruction of air), or approximant (vowel like).

A Facial expression results from one or more motions or positions of the muscles of the face. These movements convey the emotional state of the individual to observers. Facial expressions are a form of nonverbal communication. They are a primary means of conveying social information among Humans, but also occur in most other mammals and some others animal species. Some examples of feelings that can be expressed are: anger, concentration, contempt, desire, disgust, excitement, fear, happiness, confusion, sadness, surprise, blank, frown, glare, laughter, pout, shock, smile, smirk, sneer, snarl, etc...

Here is a great example of what you should be aiming for:

I create my blenshapes inside of Zbrush using the techniques in the video below (click on image for larger version):

Skeletal Rig:
A large amount of facial movement is created via the rotation of the hinge in the jaw. For the most part, we will see that Vowel sounds (sounds created by passing air through the mouth) must be created by opening the jaw. Conversely, Consonant sounds (sounds which are created by the stopage of air moving through the mouth) will be created by closing the jaw.

Usually a skeltal Rig for the face is composed of the following bones:

-bottom of the neck
-top of the neck
-head
-jaw rotate
-jaw end

As you see above, the jaw rotates downwards pulling the skin. There are also bones for the tounge which are parented to a null group and then parent-constranied to the jaw rotate. This way the tounge moves with the jaw, but isn't part of the actual hierarchy. The eyes, hat, and any hair in the image above would be parented to the head joint at the top in the skull.

When weighting the jaw, I try and account for this amount of movement roughly:

Eyeball Rig with Clusters:

To Drive my Eyelids, I prefer not to use Blendshapes. The reason is that blendshapes create a linear transformation where as clusters can drive rotational information as well. And what do eyelids do? They ROTATE around the eyeball. To set up eye lid clusters firsts I create locators and using V-snap (snap to point) to position them exactly in the center of my eyeballs.

Next, I can select all the vertices on the Left Upper Eyelid (or the Right Upper Eyelid, or the Left Lower Eyelid... etc.. you choose, I don't care, you'll have to do all four anyways), and go to Animation-Create Deformers-Cluster. I make sure to set the cluster as relative in the settings box so that the eyelids don't stay in place when the character moves. You will notice if you try and rotate these lids now though that your deformation will be a full on or off, 0% or 100%. To create a smooth rotation, we need to paint influence:

The Paint Cluster Weights Tool works just like the Paint Smooth Skin Weights tool that you are no doubt already used to, but with one AWESOME exception... you can smooth and erase with this tool since nothing is added beyond the initial membership set. So feel free to smooth away.

When you complete your painting, you should be able to rotate te cluster in a single axis to drive 1/4 of the blink. Repeat this 4 times for all the other clusters and your eyes will have full blink control. From there just use an SDK to a contol setup and you can rig the blink easily. With this method you can also create eyes that track.

Teeth and Tounge
Modeling teeth and a tounge is an extremely important step in achieving accurate facial animation. The Orbiciularis Orbis muscles, which are the muscles that move the most during the pronounciation of sound, are affected by the teeth. If the facial modeling for expression or phoneme is done without teeth in place, later, during the animation, the model will fold into the teeth or the orbicularis oris muscle will not move properly. This, as a result, will lead to awkward and unnatural animation.

In the image above, the teeth and tounge are clearly visible simply by opening the mouth. It should be noted that the tounge in particular should posses a wide range of movement including out (as in "Uh"), curl, down, up (infront ofor between teeth as in "Th"), up (behind teeth as in "L"), etc...

Here is a diagram you can use when planning the modeling for your teeth:

(note.... these are Andrew's Actual Teeth!!!.... awesome)

Blend Shapes
While most of the shapes you will see below are accomplished through blendshapes, their structure is also formed by the movement of the underlying jaw as well. As a result, often times the blendshapes must be modeled with only the movements of the orbicularis orbis, frontalis, and other such muscles in mind, while allowing the Jaw's rotation to be made seperately with a skeletal component. The movements made to create these blendshapes are shown below using arrows to denote a pulled area of verticies. ***Note, arrows on the jaw denote skeletal movement.***

Expressions
Here are the movements for four simple expressions, Blink, Smile, Surprise, and Anger:

NOTE HOW IMPORTANT THE MOVEMENT OF THE MUSCLES SURROUNDING THE EYE ARE IN CREATING AN EXPRESSIVE FACE!!!!!

Phonemes
Lets start our with some vowels. As a general rule, vowels that occur earlier in the alphabet will be created through a wider opening of the jaw. Sometimes as well, additional stress through the facial muscles might produce subtle movements of the Frontalis downwards. For the A and I vowel sounds, the lips are generally pulled a bit wider, teeth open, tongue visible and flat against the floor of the mouth. The E phoneme is similar to the A and I, but the lips are stretched a bit wider, the corners uplifted more, and the mouth and teeth closed a bit more.

For the U (OOO) sound, the lips are pursed outwards, drawn into a pucker but still somewhat open; the teeth open, and the tongue somewhat lifted. For O, Again the mouth is drawn to a pucker, but the lips don't purse outwards, and the mouth is rounder, the tongue flat against the floor of the mouth. For TH, the lips are mostly closed, stretched wide, teeth closed or nearly closed with the tounge pressed foward. This foward tounge is what seperates it from S, Z, C, D and N. For F and V, Mouth at about standard width, but teeth pressed down into the lower lip. At times there can be variations closer to the D/Th configuration.

For M, B and P; These sounds are made with the lips pressed together. It's the duration that matters. "M" is a long hold, "mmm"; "B" is a shorter hold then part, almost a "buh" sound; P is a quick hold, puff of air. For CH, SHH, and GUH, its the assumption of an EEE shape a in combination with a rising of the tounge behind the teeth, a widing of the lips, a Schwa push of air. S, Z, C, D, and N, like Th, require the lips to be mostly closed, with the mouth stretched slightly wider and with the mouth opening wider in the order the letters were presented. N has the largest downward movement of the jaw when compared to the rest of the group.

As a result of the modeling of these multiple facial shapes, we should be able to create a fairly convincing facial animation of any sort.

JUMP TO A SPECIFIC WEEK:

week 1, week 2, week 3, week 4, week 5, week 6, week 7, week 8, week 9, week 10, week 11

BACK TO TEACHING HOMEPAGE BACK TO COURSE HOMEPAGE

GA 3324 : Character Modeling This course covers advanced modeling techniques for building 3 dimensional characters. Students wil explore techniques of character modeling to include various approaches to figure construction.