The Animated Head For The Common Man

#1

Here is the last link in the series, “The Common Head For The Common Man”, showing the result of modeling one “embryonic” or generic mesh and transforming this head model into 5 very different character types. After this, the mesh is subjected to the extremes of facial animation to test its integrity. The simple mesh was modeled in Silo, a very nice and inexpensive subdivision surface modeling package. The same model constructed in Animation Master has even greater integrity and is faster to construct, also.
Animation of an “embryonic” facial mesh

I hope you find this useful,

Greg Smith

#2

Animation-wise, it looks really rough. It appears that many of the facial targets involve both the mouth and eyes at the same time. This makes for a really cumbersome facial animation rig, and severely limits the level of expression the character can achieve. Generally it works much better to have groups of controls over different parts of the face, allowing for much greater flexibility in facial expression.

#3

err… Looks like shape animation… doesn’t AM do this already? In fact it looks like you imply that AM does it even easier and faster than Silo does… why is this here?

I am teh confused.

#4

Justin:

This example was made very quickly, not with the intent of generating a critique, but merely to show that a rapidly constructed and extremely simple facial mesh can withstand the stretching necessary for extreme facial gestures. It was not intended as any kind of example of competent facial animation technique. I think I spent about 20 minutes on the “poses”. Yes, many, far too many overlapping facial parts, agreed. What I was hoping to demonstrate was that quite ordinary drawing techniques can be used to model the facial contours, and that one can be very frugal as far as adding extra density to such a mesh and still achieve a wide range of expression. I’ll try to assemble a more professional demonstration of actual facial animation at some point. Sorry this one looked rough.

The poses and animation were not done with Silo, I’m sorry if that impression was given. This segment was actually the last in a string of facial modeling sessions done in Silo, however I used Art of Illusion to create the gestures and then string them together for a final, unpolished Phong render. Mesh simplicity and integrity were the focus of the exercise.

Greg Smith

#5

Jesse:

I posted a link to this exercise because some interest has been expressed in importing simple quad meshes from external sources into Animation Master, over on the regular Hash forum. Success by those experimenting with this has been achieved with simple Silo quad models. Silo has a great subdivision preview which closely approximates a spline-based version or conversion of the same thing, so, for those liking to use a subdivision surface modeler to construct characters for final refinement in Animation Master, this “head making” session might be of assistance. It might be useful to note that both the native spline rendering abstraction in Animation Master, and the subdivision surface abstraction in Silo both are just that - an abstraction. An attempt to simulate, as accurately as possible, the rendered polygonal output necessary for a complete animated piece. Everything ultimately boils down to polygons once the render stage is encountered. Some people simply prefer one abstraction to the other. The animated and/or rendered end result is the goal.

Greg Smith

#6

The End .