Lexicor Graphics Class for the Compuserve System

Date: October 5, 1991

By Lee Seiler

All Rights Reserved


               Three Dimensional Animation:  What is it?


Contrary to popular belief, three dimensional animation is not an
invention of the modern computer world.  It all began when motion
pictures were first created using glass plates and round hat boxes.
We will build one of these interesting devices called the "ZOETROPE"
using the computer.  Some of the very first French motion pictures
were animations using fruit and color film.  The process of hand
dyeing the separate pieces of film and sandwiching them together was
so slow that it was abandoned for single pieces of film which were
first arranged in a side by side fashion.  The first motion picture
projectors were large flat boards, carbon lamps and wood hand cranks.
Then the idea of putting the frames  on drums with the frames stacked
vertically changed motion pictures forever.  Eventually they were made
into long strips and became what we know today as motion picture
film.

There have been hand drawn cartoons, science fiction and dramatic
films from the very begining.  It seems that from the first, film
makers began creating moving images that grew from their imaginations.
As time passed, the art of hand drawn animation was honed to a fine
edge by such men as Max Flisher and Walt Disney.

The biggest and longest standing hurdle to hand drawn animation has
been the task of making the subjects look right when they turned to
face a new direction relative to the viewer.  Many a hopeful animator
has failed because they just could not get the newly exposed part of
the subject to look right.

One solution to this problem was found by Willis O'Brian when he
created small movable models of giant animals in his now famous film,
"King Kong."  The technique was called Stop Motion animation.  In this
medium the animator would painstakingly move the miniature character
just a tiny bit, expose just one frame of motion picture film, move
the character a tiny bit, then expose another frame of film, and so
on.  This worked remarkedly well, but like all things done by hand
there was a certain artificial look to the finished film.  This is not
to say that the work was not exciting and entertaining, just that it
was always not quite real.  There were other problems as well, one
being the lack of motion blur and another being the limitations to the
speed at which an object could move before a phenomenon called
"strobing" would set in.  As computer technology developed, many new
methods were developed to speed up the stop motion process using
models and computers to control the model motion. This was taken to
its pinnacle by Industrial Light and Magic, and John Dykstra the
father of "Magic Cam" and CCMAP.

Anyone who has ever seen a Star Wars film, or a daily television
commercial during the mid and late 70's saw this technology in action.


For the home animation enthusiast this was all well and good, but
quite impossible for them to accomplish.  Given that the enthusiast
had the modelling skill and could build perfect miniatures to film,
just the cost of film would make this approach unreachable.  The
cameras, computers and staging for even the simplest project cost many
hundreds of thousands of dollars.

Today in the opening of the 90's that has all changed.  The home
animation enthusiast can now for a very reasonable cost do the most
elaborate animations imaginable.  How is this possible?

It is possible for two reasons.  First the inexpensive desktop
graphics computer (any computer with Graphics hardware), and the magic
of CAD Animation tools.

First a word about computers: it makes no difference what the assorted
computer manufacturers claim about the products they want to sell you,
they are all about the same.  Each has it's strong selling points,
each has it's weaknesses.  The fact is however that with the advent of
the 16 and 32 Mhz computer and massive amounts of RAM, the truly
functional desktop computer animation system is finally here.

These computers make it possible to build and animate virtually
anything that can be imagined.  The animator can now build anything he
can think of and then make a useable animation which can be viewed by
others.

Of course, if you are reading this tutorial then most of this will be
nothing new since you will be using a computer to read the text.  You
may also have all the tools required to do animations using your
computer.  So from here on we will not talk of specific tools or
computer platforms.  We will be exploring what CAD animation is and
how it works.

So just what is CAD animation?  The term CAD stands for Computer
Assisted Design.  It was first used to describe engineering
applications that used the computer to do the complex computations in
designing structural parts of machines.

The computer would be programmed to solve complex stress problems,
then numbers representing a particular aspect of the part to be
analysed were loaded into the program and the computer would solve for
whatever parameter was chosen in the program.  This process evolved
into what we now think of as CAD applications.  One of the side
benefits of this process was that the computer could also generate an
image of the object being analysed or designed.  Soon programs were
being written that would show the object from any direction in real
time.  At the moment this became possible, CAD animations were born.
It is only a short step from viewing an object to saving a series of
screen images in sequence showing what appears to be movement which
can be viewed at a later time.

The actual process of generating the image of an object is really
quite simple.  A series of files are established in RAM to represent a
given space.  This space does not actually exist, but it can be
represented on the computer screen.  This representation, while
nonexistant in terms of anything real can nevertheless seem quite real
indeed.

The space is divided up into a three dimensional volume.  Any given
point in this volume can be found and described mathematically as a
series of numbers and angles.  To show a point in this imaginary space
is even simpler. The desired point is simply found by reading its
location from a set of numbers and then finding that point in the
imaginary space and telling the computer to display (draw) a pixel at
that location on the computer monitor.

But wait a moment, the screen is flat!  It has no depth, so how can
the imaginary space have depth?  The way that happens is that the
mathematical model itself does have depth, and with the use of a bit
of perspective and angular calculation things "appear" to have depth
on the screen.  When you think about it, your TV set is really flat,
and so is a movie screen, yet things look quite three dimensional when
you are viewing either one.

The best example of this is the demonstration of depth using a cube.
We all know that all the sides of a cube are equal in length and
angle.  Our native intuitive senses tell us that things get smaller as
they get farther away.  That being so, if we deliberately draw our
cube on the monitor screen with some of the sides of the cube smaller
than others, then our impulse is to accept the cube as being solid and
having part of it further away than other parts.  The rest is all just
calculating where an object is in our imaginary space and how it is
oriented.

Although most objects we might want to use in an animation might be
more complex than our cube example, they would all be represented in
the same way.  This means that when you use a program to build an
object you are really just setting numerical values in a series of
lists (files).  As you move points or faces on your object all that
really happens is that the values of the numbers change, based on what
it is you are doing with the points and faces you see on the monitor
screen.  This explains why large RAMs and fast CPU's are needed when
doing Cad animations.  The animator needs a computer that can change
the vast lists of numbers and show the changes in these values as fast
as possible to display them as points and faces on the monitor.

Hummm well, what about color and all those wonderful textures I see in
some computer animations using CAD objects?  Same old story, just more
lists of numbers!  When the computer makes a "White" pixel this can be
only one of two things.

First, in a monochrome system, it is just the CRT gun in the monitor
firing electrons at a  given set of phosphors on the screen causing
them to give off light. In a color system, it is three guns firing
three separate beams of electrons at  three different sets of
phosphers.  The phosphors give off three different colors of light
which are RED, GREEN and BLUE (RGB).  These three colors of light when
mixed produce all visible light you and I can see.

Well this is all interesting but why have to bother with this to do
animations?  This all seems to fall into the "so what" department,
doesn't it? Why not just cut to making objects and whizzing them
around the screen?

The answer might surprise you!  Lets see if it does.

In order for anyone to create an animation using CAD objects, he/she
must first know just what they want to do and how it will look and
work.  It is a fact that the CAD animator not only creates the objects
and the motion but must also create the physical laws that govern the
objects and the object motion.

Let's take a seemingly simple example of breaking glass.

Question: How do you represent a sheet of glass as a CAD object?

Let's see, it could be a rectangular solid that is fairly thin from
front to back.  OK seems easy enough, and now we want to have
something to break it with.  What about a ball?  OK again not too
tough.  The ball can just be plain and solid.

Moving the ball to hit the glass is seemingly just as simple, so
what's the big deal?  Just this: If you did the animation at this
point without further thought, it would not work!

Try it and see.  Just build or up load the objects described and fly
or drop the ball so that it will strike the glass and keep going. Go
ahead and see what happens.

Well what happened?  It just kept going right through the glass, no
breakage at all!  Well now, let's see what we have to do to get a
realistic shattering of glass.

First let's think about what a piece of "broken" glass is.  It is a
smaller version of a whole pane of glass but often having very
irregular and sharp angular edges. OK, so that means that a piece is
the same as a whole, but there is just one problem!

There are three parts or phases to our animation: The glass before the
ball strikes it, the glass as cracks propagate from the point of
impact to the edges of the glass, and the separation of the individual
pieces and their individual paths of flight as they fall to the
ground.

Well we took care of the first and third parts, they are the same with
just one minor difference.  The pane is just one object and the broken
pieces are many small pieces with different shapes.

But what about the cracking glass?  What is a crack anyway?  It is an
empty space coming into existance, separating something that was one
solid piece into two separate pieces.  In other words it is nothing,
an empty space of separation of the glass.  Now, how do you animate an
empty space?  How do you make nothing move in exactly the right
direction and speed through the glass?

Remember that the glass is only some lines drawn on the screen between
points you defined as a set of numbers.  All the area between the
points is filled with nothing at all.  There are no numbers in the
object list for the empty space, just zeros.  We are now going to
animate nothing in a list of numbers that are all set to "zero"! So,
how do we do this?

In the real world we can't do it at all, but in the CAD world we can
do it very easly.  First think: If a crack is nothing at all but a
space between two things that are there, then we can't do anything
with nothing.  Let's make the space between the glass pieces an object
just like the glass pane and the broken pieces are objects.  We can
then represent the ever expanding cracks as numbers.  If we can do
that then we can change the numbers.  Now we can make the glass break
any way we want.

We will do this animation as an example in an upcoming class.

See, I told you that the animator not only controls the object, but
the physical laws that govern the movement of the objects.  In the
example above, the animator has changed the physical properties in the
imaginary universe from those that matched the real world (default
thinking), to new ones that allow the animation of something that
would not ordinarily be possible.

Of course, the animator may want to mimic the real world to trick the
viewer into accepting what was created in the computers RAM as
something that looks and feels real.  To do this the animator will
spend many hours exploring the real world and how it works.

By now you have probably realized that this class is not a technically
oriented approach to CAD animation.  I just wanted to give you an idea
of the nature of CAD and a bit of history to sort of get things
rolling.

Now I want to touch on the actual process of building CAD objects.
Here again I won't be dealing with any particular software program.
You can use any one that you may have and follow along with the
construction.  For Atari users there are three programs of choice,
Cyber Sculpt, Dyna Cad and CAD3D2.  If you are using the Amiga you can
use Turbo Sliver, Sculpt 3 or 4D, Imagine.  If you are on an IBM
compatible, any of the AutoDesk modeling products will do just fine.

So lets begin with something really easy, like a piece of glass we can
break, or a simple creature.  Let's do a fish that we can make swim
around.

First things first: We need to get a picture of these two things, so
go get a picture of some glass and a fish.  The fish picture will be
of more use if it shows the fins spread out.

Next let's all agree that these two objects are not complex if we plan
a bit before we create them.  As you will soon see, both objects are
really very simple.  Let's do the fish first.

Look at your fish picture and notice that the fish has several basic
parts, the body and head, eyes, fins and a tail.  Begin by creating
the body.  It will have just two shapes, the profile and the section.

The section (a slice accross the width) is oval in shape with the
bottom tapering to a point.  Create this shape first.  Make a template
of the desired shape and using an extrusion tool, make the body.  Use
12 or so sections for the body.

Now begin at the tail end and reduce the last section to a smaller but
proportional size that will fit into the profile of the fish shape you
have chosen.  Move to the next section and reduce it until it to fits
the profile of your fish.  Continue this process until your fish is
complete.  You can download the file FISH.OBJ to see how your fish
should look.  This is a .3D2 type file.

Next, again using an extrude tool make the tail of the fish.  You
should use at least 8 sections.  You should decide if you will be
animating the fish to determine how the tail should be extruded.  If
you want the tail to flex and swish when animated then use more
segments.  Do the same for the fins of the fish.

The eye is all that is left. The eye is just two simple disks.
The final step is to color your fish.  Color the various parts
differently so that when you do animate it you will be able to see the
body movements more clearly.

END of Part I

CAD Tutorial
