This post is a followup to my earlier piece explaining how 3D modeling can be likened to sculpting in virtual clay. Last time I used a basic torso build for my example; today I’m going to up the ante: I’ll not only go into the beginnings of modeling a “toon” style head, I’m going to do it live, with video. W00t!
Before we run the film I’ll give you a little background to help put what you’ll see into context. First, the style I favor for this sort of modeling is known as “box modeling.” The name comes from beginning with a very basic shape, such as a simple box primitive, then chopping that starting shape up, adding and deleting stuff, coupled with pulling and pushing the remaining points around until the shape you want emerges. There are many primitive shapes available, depending on your modeling software; you needn’t begin with a box. As you saw in my previous post for torso builds I like to start with a faceted tube. For character heads my primitive of choice is the ball, or sphere. The number of polygons (faces) in the starting shape is entirely up to you, and depends on your style and the final form you’re trying to achieve. Sometimes you just wing it and then add/delete stuff ad hoc if it turns out you didn’t start with the perfect shape. It’s all good.
Another important consideration: in most cases a head model will need to be animated. That means that, next to sculpting the face to match the desired model, the geometry (the sum of the points and polygons making up that model) must be conducive to forming the various facial expressions that will be needed. The best way to achieve that is to understand how an actual human face is shaped, how the muscles flow and move. The two most important areas for facial expression are the eyes (taken as a pair) and the mouth. In an actual face those areas are encircled by “loops” of muscles which control their movement. In keeping with art following nature, it turns out the best 3D geometry for those areas follows suit. Here is an example of how you create “edge loops” around those areas of the face to facilitate moving them realistically later:
There are numerous techniques for ending up with those loops where you want them; my favorite will become clear when you watch the video.
A couple more notes: you will see me start with the “polygon cage” and then soon into the build I’ll switch the model to “subdivision surfaces” mode. This is a smoothing technique that mathematically subdivides the base mesh (the cage) and smooths it on the fly. Most character models will employ some variation of this, depending on the modeling application, but in the end they are all pretty much the same thing. When I drag stuff around I’m actually manipulating the now-invisible cage rather than the high-polygon geometry you’re actually seeing. This operation isn’t as difficult as it may sound and you very quickly get the hang of where to grab and drag to make the subD shapes you want. Trust me.
You will also see me working in “symmetry” mode off and on. In that mode when I grab and move stuff its mirror image (when the full model is viewed as left/right halves) will also move. Very useful for avoiding duplicated effort. Now and then I want to use a tool that doesn’t support symmetry; in those cases my approach is to work on one half of the model, then delete the other half and mirror the new stuff over so my new build is again symmetric. When you see me grab just the middle points before I do the mirror I’m making sure all of the ones along the left/right seam are exactly on zero in the X axis in case I accidentally moved any. The seam must be at exactly zero or the mirroring will get borked.
A quick disclaimer: this working model is at a lower polygon count than I usually work with and, in the interest of not unduly boring you, I stop when the model is still in need of massive shaping and tweaking. It’s actually going to be quite ugly but it’s meant to show you the basic geometry and workflow that goes into this sort of modeling, not to win any beauty contests. If you manage to sit through the whole 5.5 minutes at the end you’ll see a finished model going through its paces as well as the ugly mess the hidden polygon cage becomes once a subD model has been sculpted.
Without further ado, here goes:
And there you have it! Hope you’ve enjoyed the peek into what I do; please feel free to hit me up with any questions you might have.
Spinland Studios, LLC is a high-tech branding and marketing studio in the Mohawk Valley of Upstate New York. We leverage the power and magic of 3D modeling and animation to take your company’s image places you can only imagine. Defy conventional marketing and bring your brand to life! Visit www.spinland.studio for more information and examples—then hire us to boost your company’s marketing image into the 21st century!