Trying to Understand Imagine 3.0's Forms Editor, or, Going Mad by Degrees
by Bill Graham

    Imagine's Forms editor is probably the most ignored editor in the
program. I know of users who've used Imagine since it's release,
and the Forms editor does not exist as far as they're concerned.
They have opened it once or twice, played with the very strange default
sphere, made some odd looking objects that they could never duplicate in a
million years, and gave up feeling frustrated and maybe a little mystified.
And with version 3.0, point editing can be done in the Detail editor in the
Perspective view, the New Mode, and this seems to be another good reason to
not worry about the Forms editor.
    But the Forms editor is incredibly powerful. There is nothing like
it in any other program. It is a symmetrical object editor. This means that
you edit an object's topography without affecting it's geometry. In other
words, any object created in the Forms editor can be morphed into any other
object created in the Forms editor, as long as the point and edge counts are
concurrent. Keeping them concurrent is easy, due to the way that this editor
works. By using a simple top down object construction strategy, making the
most complex object first, then the next most complex, etc., you can create
some mind blowing 3D morphs. And considering that you can morph an object's
surface attributes at the same time as it's geometry, truly amazing effects
can be accomplished. Some animations that I've done with Forms objects
include morphing fruit, with a banana turning into an apple that turns into
an orange that turns into a pear which turns into a plum which turns into a
banana again. And with States, you can control the "flow" of the morph, easily
getting around the problem of objects' points trying to move linearly when
doing an object morph. I morphed a meat cleaver into a serving platter with
utter smoothness. This ability to construct controlled morphing complex objects
would persuade me to buy Imagine just for object creation if I was already
fluent in the use of another package that supports object geometry morphing.
Assuming I had object file conversion software, of course.
    However, all this power comes at a price. And that price is a willingness
to completely rethink the way you plan and build objects. The hurdles are in
the conceptualization, not in understanding the various menus and commands. In
fact, the Forms editor has fewer menus and commands than any of Imagine's 
seven editors.
    I am going to attempt to demystify this editor. I teach Imagine
classes and have succesfully gotten through to people who had given up on the
Forms editor. Of course, we were all sitting in front of computers with Imagine
running, and I took them through it step by step both verbally and by showing
them on screen. So this may be a little more difficult, perhaps. All's we
can do is try.......

    I think what bothers most people at first is the very concept of a
symmetrical object editor. Trying to understand symmetry as it applies in the
Forms editor can be daunting. Symmetry, in the Cartesian coordinate
environment, can only be in three directions, those three corresponding to the
Top, Front, and Right orthogonal working views in Imagine. Because Forms is a
symmetrical editor, there can be no working in New Mode here. Working in Forms
is working in the abstract. The three orthogonal views are a representation
of the object, the only "true" view is in the perspective window.

   We will try an analogy. Think of an adjustable weight dumbbell.
Imagine it laying in a left to right orientation. A metal bar, with a round
metal plate on either end. This object has symmetry in all three directions.
If you had a metal cutting blade on a bandsaw, you could slice this object
all three ways, and the resulting halves would be equal.( Well, equal enough
for this example).
  You could slice it (in the Front view) left to right, with the blade going
downward in Z, with the cut moving along X. This corresponds to Imagine's X/Z
plane. The two halves would be symmetrical in Y. 
  If we rotated our bandsaw blade 90 degrees away from us around X, we could
still slice from left to right, but the blade would be moving along Y (in and
out of the screen, in the Front view). The symmetry of the resulting halves
would be in Z. The cut is occurring in the X/Y plane.
  Our final dumbbell mutilation is to cut the bar in half. The blade is still
pointing in and out of the screen, but we have to rotate it 90 degrees in Y.
Now the cut is on the Y/Z plane.We cut the bar in the middle,and the two
halves' symmetry is in X. There are only three axes to deal with. Since the
hypothetical cut must involve two of them, the symmetry or non-symmetry must
occur on the other. It has to!
    Now, let us remove one of the dumbbell's plates. We slice it like we did
the first time above, left to right, into two halves, one half closer to
us than the other in Y. Symmetry in Y, yes! We rotate the blade 90 degrees,
as before and once again slice left to right, this time with the blade
pointing away from us as we did above, slicing it into upper and lower
halves. Symmetry in Z,yes! But when we rotate the blade 90 degrees in Y,
and slice the bar in half, no symmetry. One half has a plate and the other
does not. By removing one plate from the dumbbell, we made it symmetrical
in two axes only.
   We are tired of this dumbbell rolling around on the desk. We go to the
sporting goods store, and buy the latest thing in work-out equipment. The
triangular dumbbell plate. No more stubbed toes! But for the purposes of
this "thought" experiment, we will only put one plate on. With this funny
triangular plate attached, we slice like we did the first time above. Two
parts, one closer to us than the other. Symmetry in Y, yes! We rotate the
blade away from us 90 degrees, and slice along the X/Y plane. Oops! No symmetry
in Z! We rotate the blade 90 degrees again, this time in Y. We slice the
bar in half.(Remember, only one plate) No symmetry in X, either! Yikes!
This object is now symmetrical on one axis only.
   For our last "thought" experiment, we remove all plates from the dumbbell
bar. We stick one end of the bar into a fist-sized blob of adhesive, fast
drying clay. We let it dry, and slice it three ways like we did above. Left to
right, on the X/Z plane, no symmetry in Y! Left to right, on the X/Y plane,
no symmetry in Z! We slice the bar up and down, on the Y/Z plane. No symmetry
in X either. Wow! No symmetry at all in this object.
	This is the key to creating an object in Forms. You must plan ahead,
mentally laying out your object or using graph paper to decide how many axes of
symmetry and what orientation your object will have.
   When you enter the Forms editor, it looks very similar to most of the other
Imagine editors. The Project and Display menu bars are pretty much the same as
others throughout the other editors, with the exception of there being no New
Mode option under the Display menu. There are the programmable User gadgets
at the bottom of the screen, and you should set these up after getting
familiar with all the menu options.
   The Object, Mode, Symmetry, Select, and Cross Section menu items work very
differently from anything in the other editors, however. We will cover the
Object menu first.
   In Forms, you either Load a previously saved Forms object, or create a New
one. Be aware that the Forms editor allows only one object at a time to be
worked on. To Load an object into Forms, it must have been created there, with
it's special geometry intact.
   You can Save a Forms object, and load it into Detail, and apply brush maps,
textures, and even create States, and re-load it into Forms. As long as you do
not alter the object's geometry by scaling, point editing, or using the Strip
command, you will be able to further edit a Forms object in that editor.

  When you select New from the object menu, you are presented with a requester
(Figure 1) with several fields for input.  When you are planning to construct
an object in the Forms editor,you must, in your mind's eye, determine how many
axes of symmetry there are in your object. This is essential. Three axes of
symmetry for a barbell, two for a candlestick holder, one for a single edged
knife, none for a human arm, etc. Forms is based on setting up and editing
cross sections of an object, one at a time, and editing "silhouette" views of
your object. Number of Points is the number of editing points in the cross
section of your object. You'll change the number of points in the course of
editing, so while you are learning your way around, you can accept the
default number of 16. Number of Slices is the number of cross sections.If you
were going to make a candlestick holder, Points would be the number of
editable points around the circumference, and Slices would be the number of
editable slices along the length of the holder.Twice this number is how many
points you'll see in the Silhouette views. These can be changed also, so
accepting the defaults is fine here also. 
  Contrary to what has been written elsewhere, you should never use anything
but the default Two Former views. In fact, prior to the release of Imagine
2.0, you had no choice. I say this because now that you understand symmetry,
there is no need for you to work in the other two modes. They are actually
"restricted" modes, for remedial Imagineers. They actually restrict editing
axes of symmetry (there's only three, for God's sake), and you will be
seriously restricting yourself if you use them. So I will not be covering
those other two modes. Try them if you like. But it's sorta like learning
to drive on an automatic, and then trying a stickshift. It can be done, but
it's a lot harder in the long run.
  X-Y Cross section would be selected if you were going to make your holder
right side up. Y-Z Cross section would be selected if you were going to make
it laying on it's side. When you click one or the other, the corresponding
Seal Top End and Seal Bottom End buttons become Seal Right End and Seal Left
End. The default is fine when you're learning. You see, the sphere you get
when you create a new Forms object is not really a sphere. It is a spherical
cylinder (!!??). It is actually a tube, with the number of slices and points
specified in the cross section fields, and Seal XXX End simply adds a central
point with faces that close off the tube.These ends will not be "flush"
with the normal end of the objects, and you cannot edit this single point
directly from Forms. It is a job for the Detail editor.
   In terms of learning to create objects, it doesn't really matter which
cross section selection you choose, object editing is the same either way.
Later on, when you are comfortable with Forms, it will matter, in terms of
making morphable objects and such.
   The spherical tube concept is important, because it is crucial to
understanding how to work in Forms. Basically, in the three non-perspective
views, you are working with one cross section view of the spherical tube, and
two silhouette views. Which of the Top,Front, and Right views is what
depends on which Cross section orientation you select in the New Object
requester. Your choices are two, X-Y and Y-Z. Which axis is common to those
two choices? Y, of course (no pun intended). Looking down Y is always in the
Front view, of course. And the Front view never changes, in that it always is
represented by a broken circle. The orientation of the break(s) always
corresponds with the Cross section axis. If the Cross section selected is
X-Y (tube running top to bottom), then the breaks in the Front view circle
are the top and bottom. If the Cross section selected is Y-Z (left to right),
then the breaks are on the left and right.
   The Cross section view, however, can be either the Top or the Right view.
And the Cross section view is always easy to identify, because it is always
the unbroken circle. So if you pick the up/down orientation, the Top view
will be the Cross section view. And if you pick the right/left orientation,
the Right view will be the Cross section view. Hopefully, this Forms editor
seems a little less formidable by now.
   When working in Forms, you are working on a series of Cross sections,
of which only one is visible at a time. The Cross section view never affects
the size, or diameter of the object, only it's shape. For a Cross section to
be editable,it has to be a Key Cross section. The default object always has
only one of these Key Slices. Later, we will learn how to add more. But for
now, you should be able to understand that when you create an object in
Forms, you are working on a series of cross sections, with corresponding
silhouette points in the other two views. Like the sections of an extruded
object, for instance.
   The only other selection to make in this now-not-so-nasty requester is
Fixed Cross Section Symmetry. These four buttons are mutually exclusive. From
earlier in this text you should remember that there are only three possible
axes of symmetry. Since we are dealing with a cross section of a three
dimensional object, we only have to worry about two axes of symmetry. Which
two, of course, depends on which of the X-Y or Y-Z choices you choose. These
buttons, like the Seal XXX End buttons, will change depending on which
orientation you click. (Remember, this requester is for the Cross sections
ONLY!) Cross section symmetry is always (almost) fixed. Should you be in the
midst of editing, and decide you don't want that fixed symmetry, you (a) save
the object, (b) start a new one with the same cross section points and same
form, only with the new one selecting None in this requester. (You'll be
prompted for that information) If we were to accept the default, our cross
section would be in the Top view. X axis would mean that moving a point on
the left would move the corresponding point on the right. ( You must mentally
divide the cross section into halves, it isn't marked for you.) Y axis would
do the same for the top and bottom. Both Axes move the three other
corresponding points. Try this once or twice and it'll make perfect sense.
Remember, with None selected, just the cross section point you actually work
with will move (Figure2N), with X axis or Y axis, you get two for the price
of one (Figures2X and 2Y). With Both axes, you get four for the price of one
(Figure2B). So for a candlestick holder cross section, you'd choose Both,
for a boat hull or a carving knife, you'd choose X or Y, and for something
organic like a leg or an arm you'd choose None.
    Snap to Grid, the last item in the Object menus, allows you to select one
or more points in the various views and have them all align to the nearest
grid intersection. When you become familiar with the Forms editor, this
command will be very helpful for assuring selective editing symmetry.
    And this wraps it up for the dreaded New object requester. Almost, anyway.
I will leave it as an excercise for the reader to figure out the Ok and
Cancel buttons.

   The Mode menu is simple to understand. There are three basic things you
do when editing, and these are Edit, Add, and Delete. You cannot Delete
an object to nothing but a simple axis here as you can in Detail, nor can
you Delete an objects defined Key slices. When Adding points, be
careful not to get lost. It is very easy to be in Add mode and forget to
switch to Edit mode, thereby Adding several unwanted points/slices to your
object. I recommend Adding then Editing one point at a time when you are first
learning your way around Forms. This is because what happens to the Added
editing point is not always apparent. You will always see the adjoining
connected lines "jump" when you Add a point. But sometimes you don't see the
point itself. This has made many users crazy. The fact is the new point is
always there, but what happens to it when it's first added is determined by
many factors. These factors include whether you are working in a Cross
section view or a silhouette view, what fixed symmetry is in place (determined
by the New requester), whether you have Locked selected in the Select menu,
whether the adjacent point is a Key slice or not, whether it's raining in
London or not, whether you prefer dogs over cats, whether you live in a
community property state or not, and things like that. What you need to remember
is this: The point is ALWAYS present. If you can't see it, it is either
"underneath" the nearest point you clicked close to, or "underneath" one of
the two adjacent points. It will ALWAYS be under one of those three points.
Sometimes when you Add a point, you will actually be adding two, or sometimes
four points, and instead of losing just one point, you've lost two or four.
But this is no reason to freak, because when you find one, you'll also find
the corresponding second and third/fourth point. Just remember what I said
about adjacent points above. Much more of the reasoning behind this wierdness
will become apparent when we discuss the Symmetry menu, coming up next.

   The Symmetry menu is in many ways the meat and potatoes of the Forms editor.
It has also caused a lot of confusion, because when you look at the various
Symmetry selections, you realize you've seen much of this before in the New
object requester. But the New object requester only addresses symmetry in the
Cross section view. The Symmetry menu addresses symmetry in the two "silhouette"
views. These two views are determined by the New object settings, as you
should remember. With the default settings, the silhouette views are the
Front and Right views. The first Symmetry setting is Off. This means that
when you edit a point in either of the silhouette views, only that point is
affected. The corresponding point remains unchanged. If you were modeling a
candlestick holder with the Off setting selected, it would look very
strange because it would not look "spun" as it should. But if you were model-
ing a carving knife (oriented up and down with the default settings), this
is what you'd want, assuming you were using the non symmetric silhouette
view for the knife's flat outline. Front view and Right view (or Top view, as
this changes according to the New requester settings) means that when you
edit a point(s) in that selected view, the opposing point(s) move also. These
two settings establish symmetry in their respective views. So with just one
or none of these selected, it is possible to move editing points so that they
are not "lined up". For many "organic" objects, this is precisely what you
want. This also allows you to put detail where it's needed, while keeping an
eye on object size. The Both setting simply sets up non-concurrent symmetry
in both silhouette views. So when you edit points in the Front view, only
Front view symmetry is in force, and when you edit in the Right view, only
Right view symmetry is in force. But the other silhouette view is not directly
affected. The 90 Degree setting, however, does establish dual silhouette
view symmetry. So when you edit a point in the Front view, the Right view
points move also, and vice versa. I know this sounds confusing, but when you
try it (on SIMPLE objects when learning) it really will come together.
  The Select menu has four settings that look very similar to those in the
Detail editor, but actually they behave very differently. Click mode functions
like Drag Points mode in the Detail editor. You click on a point with the
left mouse button, and while holding down that button move the point to where
you need. All Select settings work in the Cross section view as well as
the silhouette views. Click mode also supports the "multi" option by holding
down the shift key while clicking multiple points. Drag Box and Lasso are
both multi-select options, you drag-box or lasso points while keeping the
shift key pressed, then release the shift key while holding down the left
mouse button to move the points. It will become obvious to you that by doing
this you can make truly radical changes in your object very quickly. Lock
means that when you move a point or points, they will automatically snap to
the nearest grid intersection. This is a more controlled way of doing the
Snap to Grid function of the Object menu, as you can toggle this on or off as
needed.

   The Cross Section menu is the last of the Forms editor menus ( as far as
Imagine 3.0 is concerned). When you first create a new Forms object, the
spherical tube only has one defined cross section, so when you select the
Select command the first time, you'll get a box that tells you "The current
cross section is the only one defined". Once you define more than one Key
cross section, the Select command will highlite in orange all defined Key
cross sections. This ability to define key cross sections is how you create
smooth transitions from one cross section shape to another. Imagine smoothly
interpolates between cross sections, making a natural, realistic transformation
from one cross section shape to another. The cross sections have to be selected
from  one of the silhouette views, because only one cross section is viewable at
a time. The program automatically numbers these cross sections, and it is
displayed at the top of the screen. The way to make a new cross section is 
with the Make Key command. In a new object, the one pre-defined cross
section is always at the end of the "tube". After selecting Make Key, you
simply click on a point in one of the silhouette views and that cross section
becomes a Key section. You can verify this with the Select command. Once you
have made a new cross section, that is the one represented in the cross
section view.
    
   When working in the various views, once again depending on the current
settings, you will notice "rubber band" lines going from the points you're
editing towards the center of the representative view. These are guides to
tell you where the corresponding points are in the other silhouette view.
Using Forms depends very much on your ability to visualize, and relies more
on intuition (the human kind) more than any other editor. Although it seems
odd to work with at first, after a while you will find yourself starting
most of your projects here. There are some odd characteristics concerning
Forms objects, and I also have some tips on object creation.
   Forms objects are of a different structure than other Imagine objects. 
In order to make these objects into full fledged Imagine objects, you must
"break" the object geometry symmetry. You do this by loading the object into
the Detail Editor and using the Strip command and checking the Forms Data
button. This is necessary if you plan to export and convert the object for use
in other software.
  When creating two or more objects to be morphed, make the most complex
object first. So, for instance, if you were going to make a candlestick
holder morph into a pencil, you'd want to make the holder first, then the
pencil.(See Figures 3 and 4) Be sure not to delete or change the number of
slices or cross section points, if you do the objects will not morph. The
pencil will seem to have too many cross sections, but the point/edge count
must be the same in both objects. Do not use the Merge function in Detail in
order to simplify your object, it'll blow the concurrent point/edge requirement.
  For the same reason, don't take the object into another program for editing
as the object's structure would be destroyed, and even if you end up with the
correct number of points and edges, the point order will be off, and it'll
look like spaghetti when you try to morph it.
 If you're willing to put in some time, incredible transformations are
possible. Text is one good example. Flying logos are commonplace, but logos
that morph into shape from something else are not.
  Take the time to break things down, and use a little imagination. Break a
butterfly down into head, big wings, little wings, thorax, abdomen, and
with Forms make them along with the propeller, big wings, little wings,
fuselage, and tail fin of an airplane. Morph them while flying, maybe along with
concurrent morphing of cycles also. Your imagination is the only limiting
factor.
   
                                            Bill Graham