3D Animation Workshop: Lesson 8: Mind Candy
Lesson 8 - Mind Candy - Part 1
Welcome back to our newly revised 3D Animation Workshop. We have changed from weekly to bi-weekly releases, and have broken the each tutorial into three parts for easier loading and navigation. The idea is to provide even meatier tutorials, but divided into "easy to handle" pieces. Notice the new navigation buttons at the top and bottom of each page to move you between the parts of each lesson. We look forward to having you join us every two weeks for something new and exciting.
The last lesson took us through the basic principles of keyframe animation. We took a very crude and tiny little sphere and animated different parameters--transformation, scale and color. This was very instructive, but not very artistic. This week we will work through a more aesthetically pleasing animation concept and, indeed, consider what it even means to speak of an "animation concept."
Of all the parameters that we can animate, the most important and difficult to master is object geometry. By "object geometry," we mean the shape of a 3-D object. We play with the geometry of a model in two overlapping instances. First, we typically play with geometry in modeling a 3-D object in the first place. Unless we really want a perfect sphere or cube, or cone, or whatever PRIMITIVES the application gives us, we will have to modify those objects to obtain the shapes we want. For example, heads with eyes and mouths and noses are typically created by starting with a sphere and tweaking the geometry to push points in and pull them out. An important measure of the power of a 3D application, one that separates the high-end from the low-end tools, is the control it provides a user to modify the surface geometry of an object. The best applications permit the user to edit the shape of an object in the finest and most delicate ways. Even where we do not use primitives, and create 3-D objects from scratch, it will often be necessary to modify the shape.
The second, and as I say, overlapping instance, in which we play with object geometry is in animation. We have not only the power to create 3-D models, but also to modify the geometry of those models over the course of an animated sequence, and one of the most distinctive aspects of 3-D animation is the fluid change in the object geometry over time. This kind of animation is possible because of the nature of a 3-D MODEL. The reader will recall that a model that can be rendered is composed of a mesh of polygons joined by shared points (vertices). In the model file, each of these points is assigned a location in 3-D COORDINATE SPACE, with coordinates (x,y,z). These coordinates (those of the individual points) can be changed, and the changes keyframed. Thus an object changes geometric shape over the course of an animated sequence because the individual points that define the geometry are being translated. Take a moment to make sure you can imagine this process before you go on. An inherent aspect of this process is that only a single model can animated. The animation changes one version of the model into another version. You cannot animate one model to change into a different model, because shape animation depend on there being one, coherent, list of vertices that are translated from keyframe to keyframe. Thus shape animation is achieved by taking a given model, changing its shape, and then keyframing the changes.
|To Continue to Parts 2 and 3, Use Arrow Buttons||
Created: Apr. 22, 1997
Revised: Apr.22, 1997