Creating a detailed face using a polygonal modeling tool like 3D-Studio Max has traditionally been very difficult. Until recently, most people doing this kind of work have needed to spend thousands of dollars on 3D digitizing devices. Other 3D packages that allow B-Spline modelling have faired better at the task, but users do not have NURBS modelling capability.
The tutorials that come with Surface Tools only touch on what is possible with the plugin. This document is designed to supplement the tutorials, not replace them. You should read the tutorials that come with the package before reading this one. You should also be somewhat familiar with MAX in general. This tutorial assumes you have the latest version of Surface Tools, as well as the rewrite of the EditSpline modifier called EditSpline2. EditSpline2 is available on the Spectral Imaging home page in the Surface Tools section.
The first step involved with modelling a face with Surface Tools is gathering reference material. Even if you intend to model a face of your own design, you should use source material. Source material commonly includes hand drawn sketches or photographs. You may also wish to model the object in clay or plaster and then take a picture of it with a digital camera. The reference material will give you a basis for creating splines. Splines make up the structure of the face and provide the framework for which the surface modifier will create the skin. The more views you have for reference, the better off you will be. Mostly you will rely on the front and side view of your face. I used the following reference images when modelling the face above.
These images are photographs of a model created with plaster. The photographs were then scanned into the computer using a mid range quality scanner.
The next step is to create two boxes (12 hits) of suitable height. Make them thin and arrange them perpendicular to each other at their midpoints, as shown in the figure below.
Map your reference images onto the boxes. You may need to resize your mapping coordinates or scale your map to match both of the images. You want the various parts of the two images to correspond. In other words, the eye ball socket should be roughly the same height along the Z axis on both of the boxes.
These images will provide you with enough information to begin creating an outline of the face. Since most faces are symmetrical, you will usually only want to model half of the face and then mirror it. If you have an asymetrical face, you will need to model the whole thing, but working with one half at a time is probably a good idea to help you modularize the process.
At this point you will begin tracing the contours of the face. Through this process, you will create a basic framework of splines that intersect each other at common vertex points. Each area of intersection must create either a three or four sided area. The reason that you must create three or four sided areas is that the Surface modifier functions by placing either Tri or Quad patch objects in the areas outlined by the splines. These Patch grids become the surface or skin of your face. You should pay attention to where you place your vertices on your splines. If vertices are placed on a spline and don’t serve to connect with another spline to create a three or four sided area, the surface modifier won’t function correctly. You will end up with pieces missing from your face where the unattended vertices lie.
You will want to start by creating a line and then changing the vertices to Smooth, Corner, Bezier or Bezier Corner. You will probably want to create a rough outline and then apply EditSpline2 to your line. By modifing individual vertices with EditSpline2, you will fine tune your outlines. One advantage to using EditSpline2 over EditSpline, is that the vertex ticks appear while out of sub-object mode, which helps you visualize where you need to create connecting splines. It is a good idea to start with the major curves which outline the whole face, and fill in detail from there.
