Spring 1999 CS302 Assignment #6

by Dan Garcia (ddgarcia@cs.berkeley.edu)

Design the first course exam. Try to include problems from each of Bloom's categories (recall; comprehension, translation interpretation; application; analysis; synthesis; and evaluation). Also list the goals for each problem, and provide intended solutions.

Question 1 (6 points)

The first question is a warm-up recall question, intended for them to relax a bit (it's made relatively easy). It tests whether they remember the buzzwords ray-tracing, expensive, and what ray-tracing is.

Your animation partner Buzz says you should try ray-tracing your entire movie. Your other partner Woody disagrees because he says it would be too expensive. What does Woody mean?

Woody means that it would be too computationally expensive, i.e. that it would take too much time given the computational resources at your disposal.

Question 2 (10 points)

This question is also a recall question, but here it's intended to see if they were paying attention to the films we showed early on.

We saw many seminal computer animation shorts in the first few weeks. Briefly, list one of the films and describe why it was ground-breaking.

There are several valid answers. Here are a few:

Question 3 (20 points)

This question tests whether they understand the rationale for performing animation subtasks in a particular order. Specifically, it's remembering that the sounds are always done first so that the timing of the animation can be synchronized. It's also because animation is fully under our control so it's easy to shift things around, whereas audio is less under our control.

Your friend Z suggests re-ordering the computer animation design process. He wants to add the sound last, after the animation is finished. Briefly, what are two disadvantages of this method?

There are several disadvantages to adding sound last:

Question 4 (24 points)

This question tests whether they the recognize the relationship between computer and tradiational animation. Here they're asked the bigger picture why we go through such effort to teach them these tools rather than just hand-draw things. They also need to understand why folks still do draw things.

List three that are easy to do in computer animation but difficult in traditional animation, and three things easy to do in traditional animation but hard in computer animation.

Easy w/computer, hard w/traditional  Easy w/traditional, hard w/computer 
  • Changing viewpoints, moving the virtual camera around.
  • Changing the textures / colors / lights of the scene.
  • Making any changes at all through several iterations. Very little of traditional animation can be re-used between iterations.
  • Getting characters to look 3-D.
  • Real physics to the motion.
  • There are tools to facilitate so many tasks, like libraries of motions, textures, etc. Most everything with traditional animation has to be created from scratch every time.
  • You can import scientific data (say of a cloud formation) and use it in your animation.
  • There are lots of techniques to get realistic motion - data capture for one.
  • Backgrounds. They're usually just matte paintings.
  • Faces. It's much easier to draw an expression than to model a character so that its surface patches can create that expression.
  • Small simple animations. No modeling, no worries about lights, textures, camera positions. That's one reason comic strips are still drawn.
  • Making sure a character's feet don't slide when they walk.

Question 5 (20 points)

This synthesis question brings together some understandings of hierarchies with a completely new idea: infinite recursive geometries, which they've not really seen.

All of the hierarchies you've seen to date have been acyclic trees. Your friend Flik suggests the following for a hierarchy:

Two sub-questions.

  1. Why would you ever want such a crazy hierarchy? (i.e. what objects would be represented nicely here?)
  2. What simple variable would you want to be able to add your model and program to handle these objects?

  1. Fractal or recursive objects, such as a tree, fractal landscape, or the model above.
  2. You'd want to be able to specify the depth of recursion for objects (a number between 1 and infinity), as well as the depth of recursion your system calculates for objects whose requested depth is infinite.

Question 6 (20 points)

This question tests whether they have acquired any evaluation skills in the weekly crit sessions. What is shown will be a very rough animation shown with lots of opportunity for constructive criticism.

In the corner video projector you'll see an animation looping repeatedly. What are two comments you would suggest to the animator to have them improve it? (i.e. what's wrong with it?)

To improve the animation, the animator could:

WWW Maven: Dan Garcia (ddgarcia@cs.berkeley.edu) Send me feedback

Teaching cs302