LSD Docs/LSD Docs.iso/FILEZ/lsdspr06.dms/lsdspr06.adf/sculpt4d_1.doc.pp/sculpt4d_1.doc

Text File | 1990-09-07 | 209KB | 4,472 lines

                           Sculpt-Animate 4D- Part 1

                        Professional 3D Animation Software
                         for the Amiga Personal Computer

                                PREFACE

    Several years ago one of my professors gave an
interesting definition of an animator.

    "When you find someone who's good at animation," he
said, "you will also have found an extremely weird bird."

    His point was that animation required infinite patience,
good math skills, artistic talent and technical savvy.  Those
of us who lacked these attributes were put on notice to turn
our efforts to the more generalized techniques of film and
video production.  And leave the animation to the weird birds of
the world.

    I don't know that Dr. Eric Graham ever heard this
definition.  Thanks to his Sculpt-Animate series of
programs, however, anyone can produce an animation these
days.  Weird bird or not.

    The Sculpt-Animate series is one of the most amazing
graphics packages to reach the personal computer.  The
menus and methods are straightforward.  Logical.  Beneath
this interface, then, lurks powerful 3D code--capable of
operations that were once in the exclusive domain of
mainframes.

    Given this easy interface, someone who has some
knowledge of 3D representations and/or animation might
proceed to animating simple objects shortly after boot-up.
For the rest of us, a few observations may be in order.

    As you read this manual, look at the objects around you.
Don't just see what they look like, but how they are shaped.
you can see that almost every object has clear and definite
features, outlines, and dimensions.  Every feature has a
certain position with respect to the rest of the object, and the
world.  These are the elements used to model objects in
Sculpt Animate.

                                                                        i

PREFACE

    After you are comfortable creating object models, sit
back for an hour or so and watch a few animations.  Observe
how the characters really move.  Study how limbs operate,
how objects rotate and change shape.  See how your eye
blends many discrete images together, perceiving smooth
motion.  These are the elements used to create animation in
Sculpt Animate.

    Read this manual once to acquaint yourself with the
general method.  This first reading is a nice time to link the
Sculpt Animate strategy with your own observations
regarding objects and animation.  By the second reading,
you'll be like a kid on Christmas morning.  As you do the
tutorials, you'll begin to get the feeling that Dr. Graham has
given you a weird bird all your very own.  Your objects will
be made to fly, bank, walk, and squash.  In so doing, Sculpt
Animate will open up an exceedingly powerful medium
limited only by the boundaries of your imagination.

    Ten years hence, when rendering is done in real time,
disk storage is tabulated in gigabytes, and the average RAM-
count is the size of a Buick, Sculpt Animate will still be
around, lighting the way.  Given these anticipated hardware
advances, the Sculpt Animate sequences of the future may be
indistinguishable from camera-reality.  And that brings us to
another definition.  My old college professor also reminded
us that it doesn't take genius to make the complex.....
complex.  "It takes genius to me the complex simple."  By
that definition, the Sculpt Animate series is a work of
genius.  Sculpt Animate flies fast and high.  It also flies low
enough so that non-computer/non-math types like me, guys
that can't even spell cowkulus, much less do it, can jump
onboard and do the animations that we've always dreamed of.


Floyd Wray
Earth & Sky Films

                                                                        ii

     =======================================================================
        This manual and the software described in it are copyrighted
      with all rights reserved. Under the copyright laws, neither this
      manual nor the software may be copied, in whole or in part,
      without the written consent of Byte by Byte Corporation.  The
      software program may be copied for backup purposes, or
      installed on a single hard disk, or both.  Multiple installations are
      expressly forbidden.  Site license agreements are available.
     =======================================================================

    In other words, you may (and should) make backup
copies of the Sculpt Animate 4D distribution disks.  You
may also copy the program onto a hard disk for more
efficient use.  Please do not make copies of the program for
others.  You are honoring the copyright if only one copy of
the software you purchased is being used at any one time.

    When we release Sculpt Animate 4D's predecessors,
we took a risk.  We thought low-cost software could be sold
without copy protection.  We were wrong.  It seems that
some people who never see what goes into software
development don't understand that a couple of pounds of
paper and plastic can be a bargain at $500.  Yet this software
compares with packages costing 100 times as much.  Given
the popular (extreme) falsehood that software developers are
all rich, it is entirely too easy to give copies to friends, etc.

    This has made continued development of Sculpt Animate
extremely difficult.  It has literally become a decision of
whether to kill the series or to implement some protection
against illegal copying.  We will not kill the series!  Instead,
we have chosen a form of protection--the simplest and least
hazardous we could find.

    It is this simple: you may run any copy of the program,
but when it starts you will be asked to type in a word from a
specific place in the manual.  If you do this correctly, the
program will proceed normally.  This scheme is painless
(for all concerned) and as effect as any other.

                                                                        iii

     ===============================================================
     Limited Warranty on Media and Manual:

     If you discover physical defects in the media on which this
     software is distributed, or in the manual distributed with the
     software, Byte by Byte Corporation will replace the media or
     manual at no charge to you, provided you return the item to be
     replaced with proof of purchase to Byte by Byte Corporation
     within 90 days of the purchase date.  Pleas include your
     warranty registration card if you have not already sent it in.
     ===============================================================

    Registered Sculpt Animate 4D owners may purchase a
technical support policy at $15 for ten minutes or $75 per
hour.  Technical support is free of charge for 30 days after
date of purchase.  We MUST have your warranty registration
card before any assistance can be given.  Support is ONLY
available at:

        [phone]      (10am - 5pm, CST)

    Please do not call for technical assistance until you are
sure that this manual does not answer your question.  The
manual is very complete, and it can explain the program
much more effectively than we can over the phone.  It will
also be easier for you to describe your difficulty if you are
familiar with the manual's explanation.

        ==============================================
        Copyright (C) 1988 Byte by Byte Corporation
        Arboretum Plaza II
        9442 Capital of Texas Highway North, Suite 150
        Austin, TX 78759
        ==============================================

        ==============================================
         Sculpt Animate 4D written by Dr. Eric Graham
        ==============================================

Manual: Written by Cathryn E. F. Graham
        and Marcus Brooks

Cover:  Art by Rick Unland
        Layout by Tony Tomsu and Steve Bittick

Printing: AlphaGraphics #151 and the Whitly Company

Revision 1a

                                                                        iv

                        CONTENTS

INTRODUCTION
        What is Sculpt Animate?...............................xi
        About this manual.....................................xii
        Notes for upgraded users..............................xv

CHAPTER 1: AN OVERVIEW
        Sculpt Objects........................................1-1
        The Sculpt Universe...................................1-3
        Images................................................1-4
        Animations............................................1-5

CHAPTER 2: GETTING STARTED
        The menu..............................................2-5
        Loading an image......................................2-8
        Loading a scene.......................................2-11
        Making a picture......................................2-18
        Saving an image.......................................2-20
        Saving a scene........................................2-21

CHAPTER 3: THE OBSERVER
        Imaging modes.........................................3-1
        Placing the observer..................................3-8
        Aiming the observer...................................3-8
        Camera lenses.........................................3-9
        Exposure control......................................3-10
        The Jaggies...........................................3-11
        Dithering.............................................3-12
        Image size on the screen..............................3-12
        Tilting the camera....................................3-14
        Starting an image.....................................3-15
        Hiding and restoring an image.........................3-15
        Aborting an image.....................................3-16
        Memory usage..........................................3-17
        Other status display items............................3-19

                                                                        v

CHAPTER 4: BUILDING OBJECTS
        Vertices..............................................4-1
        Edges.................................................4-2
        Faces.................................................4-3
        Selecting vertices....................................4-4
        Erasing objects.......................................4-9
        Color.................................................4-10
        Texture...............................................4-13
        Smoothing.............................................4-15
        The Grabber...........................................4-17

CHAPTER 5: POWER TOOLS
        Spheres...............................................5-1
        Subdividing objects...................................5-3
        Other curved objects..................................5-4
        Cubes.................................................5-4
        Prisms................................................5-6
        Drawing curves........................................5-6
        The Fill command......................................5-7
        The Edge Maker Tool...................................5-8
        Spinning new objects..................................5-9
        Helices...............................................5-10
        Extruding shapes......................................5-14

CHAPTER 6: more tools
        Saving and loading objects............................6-1
        Rotating objects......................................6-3
        Expansion and contraction.............................6-5
        Finding the center of an object.......................6-7
        Manipulating shapes...................................6-7
        The magnet tools......................................6-8
        The Selector tools....................................6-10
        Duplicating objects...................................6-11
        Building mirror images................................6-11
        Multiple use of the extrude tool......................6-12
        The unslice operation.................................6-14
        Hiding objects in the Tri-Viewer......................6-15
        Creating complicated objects..........................6-16

                                                                        vi

CHAPTER 7: THE WORLD
        Creating lamps........................................7-1
        Illumination..........................................7-4
        The sky...............................................7-5
        The ground............................................7-6

CHAPTER 8: SPLINES
        Splines and knots.....................................8-1
        Slope and speed.......................................8-3
        Cusps.................................................8-7
        Using splines.........................................8-8
        An example: A gearwheel...............................8-9

CHAPTER 9: WHEELS WITHIN WHEELS
        Hierarchies of objects................................9-1
        Building a hierarchy..................................9-4
        Using an object hierarchy.............................9-9
        Editing the hierarchy.................................9-10
        Loading and saving named objects......................9-12

CHAPTER 10: MOTION PATHS
        Paths.................................................10-1
        Creating paths........................................10-2
        Tumbling..............................................10-4
        Motion hierarchies....................................10-11
        Local Origins.........................................10-14

CHAPTER 11: CREATING AN ANIMATION
        The Take..............................................11-1
        Two kinds of animation................................11-4
        Global Animation......................................11-6
        Example: global animation.............................11-6
        Previewing an animation...............................11-9
        Looking at individual frames..........................11-11
        Key frame animation...................................11-12
        In-betweening.........................................11-14
        In-betweening distortion..............................11-17
        Example: key frame animation..........................11-18
        Global vs. Key frame..................................11-20

                                                                        vii

        RAM animation.........................................11-22
        Making a RAM animation................................11-24
        Playing a RAM animation...............................11-26

CHAPTER 12: ADVANCED TECHNIQUES
        Loop mode.............................................12-1
        Frame duration........................................12-2
        File name conventions.................................12-4
        Disk space............................................12-5
        Organizing floppy disks...............................12-6
        Saving images.........................................12-8
        Color and exposure locking............................12-8
        Animating the observer................................12-10
        Animating lamps.......................................12-11
        Model Animation.......................................12-12
        Object Cycling........................................12-14
        Animated objects......................................12-15
        Backgrounds and foregrounds...........................12-16
        Motion blurring.......................................12-18
        Single frame controllers..............................12-18
        Frame buffers.........................................12-19
        Connecting up the hardware............................12-20
        Using a GenLock.......................................12-21
        Genlocking the Tri-View...............................12-22

CHAPTER 13: BY THE NUMBERS
        Sculpt Animate coordinates............................13-1
        The Coordinates window................................13-2
        The coordinate origin.................................13-4
        The tape measure......................................13-5
        The protractor........................................13-6
        The grid..............................................13-7

CHAPTER 14: KEYSTROKES, FONTS, SCRIPTS
        Keyboard menu equivalents.............................14-1
        User settable hot keys................................14-2
        Fonts.................................................14-3
        Scripts...............................................14-6
        Script commands.......................................14-8

                                                                        viii

        Batch mode............................................14-20
        Conclusion............................................14-21

APPENDIX A: FUNCTIONAL REFERENCE
        The Tri-View Gadgets..................................A-2
        Menu commands.........................................A-6
        The file requester....................................A-32
        The hierarchy requester...............................A-33
        The expansion requester...............................A-34
        Color requesters......................................A-35
        The modify faces requester............................A-35
        The modify lamps requester............................A-35
        The modify know requester.............................A-36
        The modify tumble requester...........................A-37
        The modify Take requester.............................A-38

APPENDIX B: GLOSSARY

APPENDIX C: FILE FORMATS

APPENDIX D: ERROR MESSAGES

INDEX

                                                                        ix

                                INTRODUCTION

What is Sculpt Animate?

    Sculpt Animate 4D, the latest member of the Sculpt-
Animate series, is an integrated solid modeling, animation,
and image rendering system.  The program enables you to
design three-dimensional (3D) objects and display them as
still or moving images, in full color.

    This means that you can use Sculpt Animate to construct
a computer simulation of virtually any 3D shape that your
mind can conceive, and then see what it would look like by
displaying its image on the computer screen.  You don't
have to be a mathematician or know anything about 3D
geometry to use Sculpt Animate.

    Any shape can be formed and manipulated with the aid
of the mouse and its two buttons.  You are not limited to
simple geometric shapes like spheres and cylinders.  Indeed,
with a little imagination you can build 3D models of almost
any real-world object.  With some artistic skill you can even
create very difficult irregular shapes, such as human and
animal forms.

    Regard Sculpt Animate as the next step beyond a paint
program.  A paint program enables you to daub simulated
pigments on the screen of your display without having to
deal with the physical encumbrances of painting, and
without messing up your best shirt.  Using a paint program
effectively requires artistic skill, however, particularly if you
are trying to produce a naturalistic scene.

    Sculpt Animate takes care of many of the messy technical
details that any 2D artist would have to deal with, such as
perspective, shading, and shadows.  In fact, making an
image with Sculpt Animate is very much like taking a
photograph, just point and shoot and out comes a picture.

                                                                        xi

    When you are using Sculpt Animate to build a solid
object, the construction is much easier than with any real
construction medium.  Changes can be made easily, and you
can save your intermediate stages.  If you make a mistake or
change your mind you can quickly backtrack.

    Sculpt Animate is not limited to making still images.  The
same tools that you use to build three dimensional objects
can also be used to describe motion paths for those objects,
or to specify the way an object is to change its shape or
series of images showing the motion that has been
described.

    The data for some moving images can be compressed
enough so that the animation can be displayed directly by the
Amiga, this technique is called RAM-Animation and was
used for the 'Juggler' and 'Kahnankas' demonstrations, as
well as others that have been shown throughout the Amiga
community.  With appropriate hardware and drivers, an
animation that's too ambitious for RAM playback can be
generated for storage on file or video tape.

    Software packages like Sculpt Animate have been
available for some time.  What makes Sculpt Animate
unique, besides its extra-simple interface, is that the earlier
packages cost about one hundred times as much and required
hardware that cost a hundred times more than an Amiga.

    What is the catch?  Well, Sculpt Animate can only display
an image within the constraints of the Amiga hardware.  It
can never generate a big-screen image as detailed as those
seen in the movie 'Tron'.  Similarly, because the Amiga is
not a supercomputer, some tasks that Sculpt Animate can
perform will take a while.  Does your Amiga really work 24
hours a day?  Now it can be doing useful work while you are
asleep, or away at the office!

    Seriously, while it could take many hours to make an
image of a taxing scene in the best available quality, excellent
images can be created without going to such extremes.

                                                                        xii

Sculpt Animate also provides ways to get a quick preview of
what the final image will look like, in only a minute or two.

    Even within its hardware-imposed constraints, Sculpt
Animate offers you a powerful ability -- the ability to build
your own private universe.  Once you have finished
construction, you can roam your imaginary world and take
pictures from any viewpoint, in any direction.  Look at
ordinary objects in extraordinary, otherwise impossible
ways, or look at objects that have never existed, and see
what they'd look like if they did.

    Indeed, while we have been emphasizing Sculpt Animate
as a recreational and artistic tool, it is by no means limited to
this role.  Professionals involved in construction or
manufacturing--of buildings, vehicles, coffee mugs, or
whatever--can use Sculpt Animate to view designs before
committing to actual production.  Scientists can present or
visualize objects that couldn't be seen otherwise, such as
molecular shapes or mathematical constructs.  Videographers
can create flying logos and moving illustrations to give their
work visual impact.

    No special skills are needed to operate Sculpt Animate.
You can 'take a picture' at any time, to see if your 3-D
objects are shown the way you'd like them to be.  This
manual will lead you through all the stages of solid modeling
and image generation.  Unlike many programs, you can start
getting useful results from Sculpt Animate with a minimum
of training.  In a few pages you will be told how to generate
your first picture, and by the time you have finished the
book, there will be no scene that you cannot create.

About this manual

    The main portion of this manual is organized as an
introductory text, which walks you through Sculpt Animate
feature by feature, explaining what each feature does and
how it may be used in the creation of your objects, images
and animations.

                                                                        xiii

    Chapter 1 provides a grand overview of the program, to
whet your appetite and to give you a feeling for where you
are and where you are going when you move on to the
chapters that follow.

    Chapter 2 introduces you to the use of Sculpt Animate's
solid modeling and static image generating capabilities, and
lets you familiarize yourself with the essential elements of
operating the program.  A section at the end of this chapter
provides troubleshooting hints.

    Chapters 3 through 9 thoroughly explore the static
imaging and modeling aspects of the program, progressing
from the basics through the advanced techniques, and laying a
foundation for animation.

    Chapter 10 introduces Motion Paths, the first element of
the program which is entirely unique to the purpose of
creating animations.  Don't think that you can easily skip to
this chapter, though.  Every aspect of Animate is built firmly
upon Sculpt's foundation.

    Chapters 11 and 12 complete the presentation of
animation techniques.  Chapter 12, in particular, details
special cases and advanced techniques which may be used.

    Chapter 13 explains coordinates, grids, and other
positioning aids which can be used throughout Sculpt
Animate to fulfill specialized purposes.

    Chapter 14 details the use of keyboard equivalents, hot
keys, fonts, scripts, and other 'power user' methods for
accelerating your work.

    A reference section follows the main text.  This contains
appendices which may be consulted for concise definitions
of Sculpt Animate functions and terminology.

    Finally, an Index is provided after the reference section.

                                                                        xiv

Notes for upgraded users

    This version of Sculpt Animate naturally differs from
earlier versions.  This section is not intended to enumerate all
of the changes and additions that have been made.  Some of
the changes made, however, although always made with the
intent of increasing the program's efficiency, might prove
momentarily confusing to users who are familiar with earlier
versions.

    The most disrupting changes take the form of menu
shuffling.  The greatest of these is that all of the face attribute
functions--color, texture, and smoothing--have been
combined into a single requester, EDIT MODIFY FACES.
This has eliminated several menu items, while preserving all
of their functionality.

    A similar change has been made to the Lamp functions.
EDIT LAMPS COLOR and EDIT LAMPS BRIGHTNESS
have been combined as EDIT MODIFY LAMPS.  LAMPS
CREATE has been changed to EDIT ADD LAMP.

    All of the LOCAL ORIGIN commands have also been
moved to the EDIT MODIFY submenu, combined in the
EDIT MODIFY LOCAL ORIGIN requester.

    The PROJECT ANIMATION submenu has been broken
up, and its functions moved into the PROJECT LOAD,
PROJECT SHOW, and PROJECT UNLOAD submenus.

    Two other moves have been made for purely logical
reasons.  EDIT DO BE SPHERE has become EDIT SNAP
CONNECTED TO SPHERE, and OBSERVE SHOW has
become PROJECT SHOW IMAGE.

    These are by no means the only changes that have been
made to the program.  The many other changes are primarily
additive, however, being features that simply never existed
before.  Other changes should be obvious enough that you
won't have to think twice about them.

                                                                        xv

    This manual and those of earlier Sculpt Animate products
naturally have several similarities.  Be sure to read through
this manual at least once, though.  It thoroughly explains the
program's new features, and great effort has been expended
to enhance and clarify the rest of the text as well.

                                                                        xvi

                                CHAPTER 1

                               AN OVERVIEW

    This chapter contains an overview of Sculpt Animate.
Everything in this chapter is described elsewhere, so you
may skip ahead if you wish.  You may find it useful to read
what follows, though, so that you have a grasp of the big
picture before we delve into all the details of the program in
the other chapters.

    While Sculpt Animate is easy to use, some aspects of its
use are not immediately obvious.  You may want to begin
using the program immediately.  This is a good way to get
your feet wet, so to speak, but you can quickly get in over
your head.  If this happens, don't be too discouraged, just
turn back to this manual and continue reading.

    You will find that there are several different ways to do
certain things.  If what you are trying to do seems too
difficult, you can probably find an easier way if you review
the related portions of this manual.  When a reference is
made to another chapter, be sure to look there as well.

Sculpt Objects

    The objects that we will be building, be they cubes,
aircraft or elephants, will all be made from a single building
block--a triangular face.  Such faces can be joined together to
approximate the surface of any shape.  The smaller and more
numerous the triangles, the better the approximation.  This
needn't be carried very far, however, for although the
triangles would normally create a faceted, angular shape,
they can be shaded as if they were curved.  This makes the
joint between adjacent triangles quite invisible.  This is called
Phong smoothing, and it's particularly easy to employ,
you'll just set smoothing 'on' for certain faces, and the
program will take care of the details.

                                                                        1-1

    Each triangular face can be given other attributes besides
smoothing, such as color and surface finish.  Faces can be
shiny or dull, luminous or metallic, even mirrored or semi-
transparent, so you will be able to accurately model a wide
range of materials.

    Although the triangular face is the basic building block
for objects, each face is described in terms of simpler objects
called vertices and edges.  A vertex describes a position in
space representing the corner of a face, while an edge is a
link between vertices, and is used as the side of a face.
Vertices and edges can be shared between more than one
face, so Sculpt Animate objects are really built up from a
triangular network of edges.  By moving one or more
vertices, you can changed the shape of an object.

    Objects are built in an interactive environment called the
Tri-View that displays the vertices and edges in a so-called
wire frame view.  The Tri-View lets you look at a box-like
region of space from three different angles at once, usually
northward, westward and downward.  You can use the
mouse to manipulate vertices in all three directions, but you
will have to switch from one view to another to be sure that
you have created the intended change.

    You need to use multiple views because a window on a
flat screen can only represent two dimensions at once.
Think of a shoebox: if you look directly at its end, you can
only see two of its dimensions--height and width.  If you
want to look at its length, you have to look at either its top or
side.  The Tri-View allows you to do this.  Although it may
seem difficult at first, especially if you are new to such
things, take heart.  After a little practice you will be surprised
by the intricate structures you can build.

    Sculpt Animate gives you commands for producing
many standard shapes automatically.  These can then be
stretched or distorted to produce still more shapes.  Shapes
that have an element of symmetry can often be generated
from simple curves, for example a wine bottle shape can be
generated by building its profile and using the Spin
command to sweep the profile around, creating a curved

                                                                        1-2

surface.  Similarly, the Extrude Tool will let you stamp out
shapes like a cookie cutter, and the Unslice command will
join cross-section profiles, creating quite irregular shapes.

    Sophisticated users can create spline curves: smooth
curves made of many vertices, of which only a few need be
adjusted to alter the curve's shape.  Objects can also be
decomposed into named hierarchies, making it easier to
manipulate parts of an object while leaving other parts
unaltered.

    Objects can be saved and recalled from disk, so you can
quickly assemble a library of useful objects.  Once built,
these can be reused in any number of scenes.  Because of
the success of programs in the Sculpt Animate series, there
are numerous objects whose builders have placed in the
public domain, or included in commercially available object
libraries.  When you use pre-constructed objects, it is almost
trivial to assemble interesting and complicated scenes.
Dazzling images can be created with an absolute minimum of
effort.

The Sculpt Universe

    Once you have assembled your objects, Sculpt Animate
can be asked to provide certain environmental elements to
round out the scene.  One useful item is a ground plane that
stretches to infinity.  The ground can be given any solid
color, or painted with a checkerboard of two contrasting
colors.  A sky can also be provided which can be shaded
from one color at the zenith (overhead) to another color at the
horizon.  If your demands for a background setting are more
complicated--for example if you need a distant mountain
range, or a lake of rippling water--then these can be created
in the form of objects.

    A major purpose of Sculpt Animate is to create images of
your modeled objects, and the most important element of any
image is its point of view.  In the Sculpt universe, this is
called the 'Observer'.  Think of the Observer as a camera
that can be positioned and pointed anywhere.  The view that

                                                                        1-3

this hypothetical camera would capture is what your screen
displays when an image is rendered.  You are provided with
a number of controls over the Observer, including a lens
with a alterable focal length, so that you can zoom in or out
of the scene and modify the apparent perspective of the
image.

    In addition to objects and the Observer, a scene will
normally contain lamps to illuminate the objects.  As any
painter, theatrical set designer, or photographer knows, the
placement of light sources can be instrumental in creating a
desired effect.  In Sculpt Animate, any number of lamps can
be used, and each can be given a different position, color,
and brightness.

Images

    There are numerous ways to produce a computer
generated image.  As you might expect, the better quality
images usually require more computer time.  While simple
images can be produced in a few seconds, realistic images
with shadows, highlights and reflections can take hours or
even days to generate.  Sculpt Animate allows you to trade
image size and realism for savings in computing time.
Numerous rendering modes and image sizes are provided,
so you can pick the degree of realism you require for a
particular application.  For complex scenes, it's best to try
one of the faster modes or a smaller size to preview the scene
before committing your Amiga to a lengthy computation.

    Images can be stored on disk in the Amiga's standard
IFF format.  Images can also be recalled from disk to be
displayed on the screen.  The images can be modified using
either paint programs or image processing programs,
although not every program can handle the Hold-and-Modify
(HAM) or the overscan images that Sculpt Animate is
capable of generating.

    Sculpt Animate can even generate images that an Amiga
cannot display.  These exist as raw data representing almost
any desired resolution and up to sixteen million colors.  The

                                                                        1-4

data can be sent out to files for programmers and the like to
use, or to a separate frame buffer, which is a device built just
to display high quality images.  Each frame buffer type is
accessed through its own custom device driver, so Sculpt
Animate won't have to be re-written for every new frame
buffer that appears.

Animations

    Sculpt Animate can be used to make animations.  Objects
can be made to move by specifying their positions at
different times.  Each feature of the Sculpt universe can be
made to change, lamps can dim or brighten and change
color, objects can change size and shape, or even vanish, the
observer can fly through the scene and change its viewing
direction, the sky can be made to darken to simulate dusk.
There is practically no limit to the effects that can be
animated.

    An animation comprises many images, each differing
from the previous one by just the right amount to create the
illusion of smooth motion.  Sculpt Animate uses a
mechanism called a 'Take' to keep track of all the details of a
single animated sequence.  The Take knows the names of all
the files that describe the objects as they move around, as
well as the name of each image that is generated.  When a
Take has been properly specified, the process of rendering
all the frames can be accomplished with a single mouse
click.

    Object movement can be specified in several ways.  You
could, for instance, create a series of scenes, setting the
position of each object for every picture in the animation.
This would be extremely tedious.  Since most motion is
smooth, however, it's much easier to just set up a few
images called key frames.  The program can usually figure
out where the objects should be for the frames in between.
This technique is called key frame animation.

    If an object's motion is fairly simple, for example a car
driving along a road, the path that the object follows can be

                                                                        1-5

described, and the object 'attached' to the path.  The path
describing the motion is just another object, so a single scene
can contain all the information needed to animate the object.
This is called global animation.  Very complicated motions
can be described by a hierarchy of motions, with movements
superimposed on each other to produce complex motions.
Global and key frame animation can even be mixed in the
same take, with some objects being described in a global
manner while others are described in terms of key frames.

   The motions can represent actions that are impossible in
the real world, or they can simulate reality with uncanny
accuracy, it is up to you to decide what effect you wish to
create.

    There are two ways for completed animations to be
displayed.  If the animation is not too large, is possible to
store its image in memory in compressed form, then display
them quickly enough to simulate smooth motion.  At video
frame rates, however, even a second's worth of animation
can use a lot of memory, and if motion is extreme, the time
required to construct each image may prevent full speed
playback.  Nevertheless, RAM animations can be useful
demonstrations, or can be edited together on videotape for
longer presentations.

    The other, more versatile way to display animations is to
copy the images directly to motion picture film or to
videotape.  As you might expect, this method requires the
use of special purpose hardware.  Such devices require
custom 'interface' software.  Sculpt Animate can access this
software through a device driver facility, without any need to
customize Sculpt Animate itself.

    As you can see, Sculpt Animate offers a great many
facilities, which you can use to create sophisticated images
and animations.  Sophistication is not a prerequisite for the
user, however.  While it is true that advanced tools and
'power user' techniques exist, the essence of Sculpt Animate
is to open up the world of 3D graphics and animation to
everyone.  Here it is, climb in!

                                                                        1-6

                                CHAPTER 2

                             GETTING STARTED

    This chapter introduces you to the basics of operating
Sculpt Animate.  you will be shown how to load a scene
from a disk, generate a picture of the scene, and finally save
the picture to disk.

    Before you do anything else, read the support policy
statement and warranty information that came with this
package.  Fill out the warranty registration form and send it
in.  This ensures that you will be able to obtain support
when you need it, and registers you for future updates and
announcements.

    If you have already used a few Amiga software packages
you can probably skip to the next section, but be sure to
make a backup copy of the Sculpt Animate distribution disks
first.  If you are new to the Amiga, you should know that
while floppy disks are very secure, accidents can happen, so
it is wise to make a copy and keep the original disks in a safe
place, far from coffee spills, magnets, and small animals.

    The book INTRODUCTION TO THE COMMODORE AMIGA, that
came with your Amiga, tells you how to make backup
copies.  When this important step is complete, you are ready
to start using Sculpt Animate.

    The program can be started from the Workbench Screen
by first double-clicking on the icon for Sculpt Animate
disk.  Then, when the disk's window opens, double-click
on the 'SA' icon that appears.  If you are an advanced user
you may run SA from the Command Line Interface (CLI),
but you must first set the CLI's stack to 50,000 or more.

    After a few moments of disk activity, Sculpt Animate
starts by displaying three windows, arranged like three sides
of a folded-out box on the screen.  These are collectively
known as the Tri-View.

                                                                        2-1

    NOTE: If your system disk (usually 'Workbench') is
not in a drive, it is likely that a requester will appear when
you run Sculpt Animate, asking you to insert the desired
volume.  This means the program needs to read a file on that
disk, and can't proceed without it.  Whenever this happens,
do not click CANCEL.  Instead, insert the requested disk
and click 'Retry'.

              +-----------------------------------+
              | | |NORTH=====|^|==============| | |
              |-+-------------------------------+-|
              | |                               | |
              |-|                               |-|
              | |                               | |
              | |                               | |
              |<|       A Tri-View Window       |>|
              | |                               | |
              |-|                               |-|
              | |                               | |
              |-+-------------------------------+-+
              | | |          |v|              | | |
              +-+-------------------------------+-+

The Tri-View

The Tri-View is a necessary part of three dimensional
editing, because the Amiga can only display flat images on
the screen, while Sculpt Animate works with three
dimensional solid objects.  The three windows show three
different views of a rectangular chunk of space.  Initially
each window is empty except for a little blue cross in the
center, and four letters denoting the cardinal directions that
apply in that window.

                                                                        2-2

    The illustration shows a representative Tri-View
window, there are three windows like this on the Sculpt
Animate screen.  The window at the upper left is a view
looking downward, like a map, with north at its top, west to
the left, etc.  The window at the lower left shows a view
looking northward, so west is to the left and up is towards
the top of the screen.  The third window, at the lower right
of the screen, is a view looking west, with south to the left
and north to the right of the window.

    If you have ever encountered technical drawings, or
blueprints of a house, you will be familiar with this way of
representing solid objects.  As with most Amiga software,
these windows may be moved around on the screen in any
way that you wish.  To move a window, use the mouse to
position the pointer anywhere on the Drag Bar at the top of
the window.  If you press and hold the left mouse button on
this bar, an outline of the window will appear.  This outline
can be moved anywhere on the screen.  When you release
the button, the window will be re-drawn at the new position.

                                                                        2-3

    The size of a window can be changed by using the sizing
gadget that can be found at the bottom right corner of each
window.  To change the size of a window, use the mouse to
point at the sizing gadget, then hold down the left mouse
button while you drag the corner of the window to create a
new shape of size.  When you release the button, the
window is refreshed, but you may notice that something
funny happens.  Whenever you change the size of one
window, one or both of the other windows will change also!

    This is because the windows represent a rectangular
chunk of empty space, they're like three sides of a box that
has been opened out and laid flat.  If only one of the sides is
changed, then it won't fit when the box is folded back up.
One or both of the other sides will have to be changed as
well, so the program does this automatically.  Suppose that
you reduce the size of the box at its south end, then the
change would also be seen in both the view looking down
and the view looking west.

    If your windows overlap one another on the screen, you
can use the back gadget or the front gadget to change their
depth arrangement.  As usual on the Amiga, these gadgets
may be found at the right hand end of each window's top
border.  To use a gadget, point to it with the mouse and click
the left mouse button.

    The initial sizes of the Tri-View windows are small
enough that each window can be displayed without covering
another.  This is recommended when you first use the
program, because it is easy to get confused while working in
three dimensions.

    When you have gained confidence, you may wish to
make the windows as large as possible and only use one
window at a time, using the depth arrangement gadgets to go
from window to window.  The windows can be re-sized
manually, but it's quicker to use the command EDIT DO
MAKE TRI-VIEW BIG.  The complimentary command
EDIT DO MAKE TRI-VIEW SMALL will return the
windows to their default configuration.

                                                                        2-4

    Each Tri-View window also has more than a dozen extra
gadgets tucked away in its borders.  We will be describing
each one in detail later in this book.  To avoid confusion, try
to restrain yourself from using the gadgets, until you know
what they are supposed to do.  Of course, if you fell
adventurous you can try them anyway.  None of them can
cause permanent damage to your Amiga.

    If you use a gadget incorrectly, a message may appear to
warn you of this.  Such messages usually display a 'button'
gadget, marked 'CONTINUE', beneath their text.  To
continue with the program, move the mouse pointer into the
button's outline and click the left mouse button.

    The blue cross shown in each Tri-View window is called
the cursor.  There is only one cursor, but you see it in all
three windows at once.  It represents a unique position in
three dimensional space.  To move the cursor, point with the
mouse at a new position in any window and click the left
mouse button.  The cursor will jump to the pointer's
position.  If you hold down the left mouse button and move
the mouse, the pointer will vanish and the cursor will move
in its place.

    You will notice that you cannot move the cursor outside
any window.  You should also notice that when you move
the cursor in one window, it also moves in the other
windows.  The mouse can only move in two dimensions,
but the Tri-View makes all three dimensions readily
accessible anyway, just by switching to another window.
You will quickly become accustomed to this style of
working in 3D.

The menu

    Most of the commands that you can issue to Sculpt
Animate are made through the Amiga's menu system.

    To make the menu visible, press the right mouse button
and hold it down.  If nothing happens, move the pointer into
one of the Tri-View windows, and click the left mouse

                                                                        2-5

button (this 'selects' the Tri-View).  Then try again with the
right mouse button.

    The menus show up as a bar at the top of the screen,
showing the words 'PROJECT', 'EDIT', 'TOOLS',
'OBSERVER' and 'WORLD'.  These are the major menu
headings.  To see the choices in each menu, move the
pointer up to the desired heading, keeping the right mouse
button down all the while.  When the pointer touches one of
the words, a list of choices will spring down.

    To activate one of these choices, move the pointer to that
item and release the right mouse button.  If, as you look through the
menu, you decide you'd rather not execute any of the
commands, move the pointer away from the menu region
and release the right mouse button.  Be a little cautious about
where and when you release the button, or you may execute
a command unintentionally.

                        Menu and Submenu

    As you move the pointer down the list of menu choices,
sometimes an additional set of choices will pop up to the
right.  This is a 'submenu.'  To execute a submenu item,
keep the right mouse button pressed down and move the pointer to
the right, then down the new list to the item you want.

                                                                        2-6

When you release the button oven an item, that action takes
place.

    At times you may need to issue several commands from
the menu, one after another.  Rather than releasing the right
button, and then having to wend your way throughout the
menu system each time, there is a way to generate multiple
commands.

    When the pointer is over the menu choice that you want,
you can keep the right button depressed and click the left
button instead.  Sill holding down the right button, you can
keep moving through the menus, selecting additional
commands by clicking the left button on their entries.  When
you release the right mouse button, the items you selected
will be executed in the order in which they were selected.

    There is one exception to this, if one of the items selected
brings up a requester (requesters are discussed below),
subsequent selections in a multiple command will be
ignored.  You will quickly become proficient at this speedy
way of using the menu system.

    For those of you how are not averse to dealing with the
Amiga's keyboard, there is an even more fluid and efficient
way to get to the menu's commands.  In the menu, you will
notice odd two-character symbols to the right of some menu
items.  Each of these is a 'keyboard equivalent' for that
menu item.  If you are familiar with keyboard equivalents in
other Amiga programs, you can use them now.  If not, they
are explained fully in chapter 14.

    Keyboard equivalents can also be defined 'on the fly'.
The simplest of these is the ESC key, which will repeat the
last menu command.  You can also hold down the ALT key
and press any alphabetic key after executing a menu
command.  This defines that alphabetic key as a 'hot key'
equivalent for that menu command for the rest of the
session, or until the key is re-defined.  This is also explained
in chapter 14.

                                                                        2-7

Loading and image

    To get an idea of the capabilities of Sculpt Animate, take
a look at some of the images it has created.  Hold down the
right mouse button and point to the word PROJECT at the
top left of the screen.  Move down to the word LOAD, and
then move to the right to the word SCENE.  Finally move
down to the word IMAGE and release the right button.  You
have issued the command PROJECT LOAD IMAGE.

    If you have done this correctly, the menus will vanish
and the pointer will be replaced by an image of grinding gear
wheels.  This indicates that Sculpt Animate is busy and
should not be disturbed.  In this case it is looking through the
disk to see if it can find any files that contain images.  You
will notice that the disk drive light is turned on.

    After a few seconds, the normal pointer will return, and
a new window will be displayed.  This window is a
'requester'--the Amiga's standard method of asking the user
for more input.  There are many kinds of requester, this one
is asking you to choose the name of an existing file that
contains a Sculpt Animate image.  This File Requester is
used by several commands, so you will be seeing it a lot.

    The File Requester contains three large boxes for making
selections.  The contents of the boxes will depend upon what
files you have in your computer.

    The box on the upper right is used for selecting a
volume.  Each floppy disk that the Amiga knows about is
regarded as a separate volume, with its own name.  The
RAM disk is also a volume, as are any hard disks, etc., that
may be mounted (recognized by the system).  A volume is
distinct from a disk drive or unit, in that a drive may exist
that has no volume in it.  Data resides in a specific volume,
but the drive that a volume is mounted in is of little
importance to Sculpt Animate.

    If you wish to inspect to contents of a particular
volume, double click on the name of the volume.  After
some disk activity, the display will change, and some of the

                                                                        2-8

contents of that volume will now be visible in the requester's
other two large boxes.

    The large box at the upper left of the requester contains a
list of drawers, if there are any in the selected volume.  You
can double click on the name of a drawer to select it.  The
uppermost entry in this box is called '/Parent Directory' and
it is used to select the drawer, if any, that contains the
current drawer.

                        The File Requester

    The large box at the lower left contains a list of the
names of all the images in the selected drawer on the selected
volume.  At this time you could double click on the name of
the file and the image would be displayed.

    Above the Drawer and File boxes are slots where you
can type in a name.  At the lower right are four named boxes
called 'RENAME HIGHLIGHTED FILE,' 'DELETE
HIGHLIGHTED FILE', 'CANCEL' and 'OK'.  These
boxes are called buttons--each one can be activated by
clicking on it with the left mouse button.

                                                                        2-9

    The CANCEL button is present on most requesters and
enables you to cancel the current command.  The OK button
informs the system to proceed with the current selection.

    The RENAME and DELETE buttons can be used to
change the name of a file and to delete it from the system,
respectively.  CAUTION: If you delete a file, there is no
reprieve, its contents will be lost forever.

    If there are more than six items to be shown in any of the
three large listing boxes, only six items will be shown, but
the gadget to the right can be used to reveal more.  This
gadget is called a scroll bar, and it is similar to scroll bars in
other Amiga programs you may have used.  The white
rectangle in the tall narrow box is called the slider knob.  The
size of the knob relative to the size of the slider box indicates
how many items are shown relative to how many exist.

    You can grab a slider knob by placing the pointer over it
and holding down the left mouse button.  If you move the
slider to the bottom of its box and release the left mouse
button, the listing box will display the last six items in its
list.  If the knob is moved to the top, the first six items will
be shown.  With the knob in the middle of the slider box, the
middle six names are displayed.

    To move up or down in the displayed list of files, click
on the arrow gadgets above or below the slider box.  If the
listing box contains six or fewer items, the slider will be
inactive, filling its box to show that there are no extra items
to display.

    Advanced users, especially those with hard disks, may
wish to keep different projects in different drawers.  This
makes it easier to find files and keep things organized.

    When you have found a file that looks interesting,
double click on it.  The screen will clear and the image will
appear one line at a time, starting at the top of the screen.
When you wish to stop viewing the picture, click on it first
with the left mouse button, and then click with the right
mouse button.  If you want to remove the image from

                                                                        2-10

memory, hold down the right mouse button, go to the
PROJECT menu, move down to UNLOAD, then across to
IMAGE and release the button.

Loading a scene

    A scene is a collection of objects and lamps.  A scene
may also contain information about the observer and a
description of the ground and sky.  The scene is essentially a
description of what Sculpt Animate is to make an image of,
and how the image is to be made.  Scenes are manipulated in
the Tri-View and may be stored on disk.  Your Sculpt
Animate disk comes with some scenes, so let us try loading
one of them.

    Execute the menu command PROJECT LOAD SCENE.
After a short pause, a requester will pop up that looks very
much like the requester used to load images.  Scroll the list
until you find a scene called 'LITTLEHOUSE'.  Double
click on it to indicate that you have made your choice, and
that you still want to load a scene.

    Objects and lamps that are already in the Tri-View will
not be erased when you load in a scene, but at this point a
second requester will appear, giving you a choice of what
parts of the new scene to load.  You may selectively load
parts of several scenes to combine them.  This is explained
more fully in chapter 6.

    If you want to start with a clean slate, use the command
EDIT ERASE ALL before using PROJECT LOAD SCENE.

    The 'Load What?' requester contains a list of the scene's
various parts.  Beside the name of each item is a box
containing either the word 'YES' or 'NO'.  Initially, each
box contains the word 'YES'.  If you click on a box with the
left mouse button the contents will change from 'YES' to
'NO' or vice-versa.  You can click on the boxes until they
correctly show what you wish to load.  For now, we want to
load everything, so you only have to change the boxes if you
have fiddled about with them earlier.

                                                                        2-11

    Below the other boxes is one called 'Select' that affects
the way that objects are loaded, normally you will wish to
leave it as YES.

                   The 'Load What?' Requester

    When all of the boxes contain the word 'YES', click on
the OK button.  After a few seconds, an image will appear in
each window of the Tri-View.  At first, the display may
seem confusing.

    This kind of picture is called a wire frame, because the
object is displayed as though it were transparent, with only
the edges visible.  In addition to the object's outlines, Sculpt
Animate divides each rectangular side with an extra diagonal
line.  Although this seems like a drawback, we will later see
how it becomes an advantage when we are building more
complicated shapes.

    Look at the window with the name 'North' in its title
bar.  It should be at the screen's bottom left, unless you have
re-arranged the windows.  This is the view that faces
northward.  Notice that there is a small letter just within each

                                                                        2-12

border of the window.  At the left edge is a 'W' for 'west',
to the right, 'E' for 'east'.  'U' for 'up' is at the top, and
'D' for 'down' at the bottom.  These are the directions you
can move things in this window.

    The view in this window should be recognizable as the
end view of a simple house with a ridge roof.  Note that the
wall seems to be braced diagonally in both directions.  In
fact, the north wall (the one furthest away) is braced in one
direction while the south wall (the one nearest to you) is
braced the other way.  In a wire frame view, you cannot
distinguish between near and far objects because far objects
are not hidden by those nearby.  Now we will look at one
way to make sense of these confusing images and at the
same time learn about some of the Tri-View gadgets.

                        The North View

    Move the pointer to the 'West' window (the window that
looks west, with an 'S' for 'south' at the left side, 'N' for
'North' at the right, etc.).  Unless it has been moved, it is the
one at the screen's bottom right.  Now look at the middle of
the left hand border (south, remember?).  You will see a
gadget with a left pointing arrow.  This is called a Move Tri-
View Gadget.  Click on this gadget with the left mouse

                                                                        2-13

button.  After the usual screen flashing, try to see what has
happened.

    At first glance, the little house seems to have moved to
the right (i.e. north).  It's more useful, however, to say that
the Tri-View has moved to the left (i.e. south).  That is why
the gadget's arrow points to the left.  It is better to think of
the objects in Sculpt Animate as fixed in space, and the Tri-
View as moving.  As we said before, the Tri-View windows
display three views of a 'box' of space.  This box itself can
be thought of as a sort of window, into which we can look
to see a piece of the Sculpt Universe.  As you might expect, it
makes more sense to move the window around in the
universe than it does to move the universe around in the
window!

                   The Tri-View "Box" From Inside

    If you want a 'real world' analogy, think of riding a
bicycle.  Most of the time it's more convenient to imagine
oneself rolling down the road, rather than thinking the
bicycle's wheels are moving the earth beneath you.  After a
while, the world of Sculpt Animate may seem just as real to
you.

                                                                        2-14

    By now you should be able to guess what effects the
arrow gadgets in the centers the other borders will have.
Each will move the Tri-View in its respective direction.

    Back to our example.  Look again at the northward
window.  If you are particularly astute you will have noticed
that one of the diagonal braces of the wall has vanished.
This is because the brace was part of the north wall, which
now lies outside the Tri-View.  Click on the gadget in the
middle of the right border of the west Tri-View window to
move the Tri-View back to its original position.  The whole
of the house fits inside the Tri-View, so the second brace
becomes visible again.

    The Tri-View will only display objects, or parts of
objects, that are within the chunk of space that the Tri-View
represents.  When you are dealing with complicated scenes,
such as an entire village of little houses, this is a decided
advantage.  You can move the Tri-View to the area that you
are interested in, and not be bothered by neighboring
objects.  By the same token, you can use the Tri-View to
isolate just a part of an object, as you have already done with
the little house.

    Now try using the Sizing Gadget in the bottom right
corner of a window to make a window bigger.  See how the
other windows adjust themselves in order to be consistent.
Also notice that although you make the window bigger by
stretching out the bottom right corner, as far as Tri-View
is concerned it gets bigger uniformly with respect to its
center.  If the little house was in the middle of the window
before enlarging the window, it wills till be in the middle
after enlargement.  The image of the house does not get
any bigger, there is just more space between the house and
the border of the window.

    There is another way of changing the size of the Tri-
View.  Click on the gadget that is near the bottom of the right
hand border.  This is called the Expand Tri-View Gadget.
Somewhat to your surprise, the Tri-View windows don't
expand, in fact the image of the little house gets smaller.
This is relativity once again.  What you have done is enlarge

                                                                        2-15

the Tri-View's 'box' so that more 'Sculpt space' is
contained within it, even though the Tri-View windows are
the same size on the screen.  Click on the Expand Tri-View
Gadget a few times to make the Tri-View even bigger (which
makes the house's image on the screen still smaller).  Now
can you see why we like to think of the Tri-View's volume
increasing, even though its windows stay the same size.  If
the windows had expanded this much, they would have
overflowed off the screen.

    The Expand Tri-View Gadget increases the volume of
the Tri-View equally in all three directions, its width
increases by about 40% with each click.  If you hold down
the left Amiga key (it is called the Commodore key on newer
Amigas, but will keep on using the old name) while the
gadget is clicked, the amount of expansion is reduced.
Similarly, if the right Amiga key is held down when the
gadget is pressed, each click of the gadget will have twice
the usual effect.  Similarly, the left and right ALT keys can
be used in the same way to decrease or increase the effect of
the gadget still more.  Several other gadgets, such as the
Move Tri-View Gadgets, can also be modified by the Amiga
keys or the ALT keys.

    Let's try moving around in Sculpt space a little.  First,
press one of the Move Tri-View Gadgets a few times until
the house vanishes from view.  At first, the cursor (the blue
cross) will stay with the house, but when the edge of the
windows hits it, the cursor will be swept along in the Tri-
View.

    Next, expand the Tri-View with the Expand Tri-View
Gadget until the house becomes visible.  Point to the middle
of the house and press the left mouse button.  This moves
the cursor to the middle of the house in that window, but
look at the other windows to see if the cursor is really in
the middle of the house in all views.  If not, repeat the operation
in another window.

    Now click on the gadget that lies second from the left of
the bottom border.  This is called the Center Tri-View
Gadget.  It moves the Tri-View so that it is centered on the

                                                                        2-16

cursor's location, putting the cursor in the middle of the Tri-
View windows.

    Finally, click on the gadget that is second from the right
in the bottom border.  This is called the Contract Tri-View
Gadget.  It makes the Tri-View smaller so, the image of the
house seems larger.

    You have just completed a very useful sequence of
commands, which you'll use often: Expand the Tri-View to
find a point of interest, place the cursor on it, then center the
Tri-View and contract it again.  Except for short distances,
this is often easier than using the Move Tri-View gadgets.

    You can use the expand and contract gadgets like the
zoom control on a camera.  Expanding the Tri-View is like
zooming out to a wide angle setting, the field of view gets
bigger so objects seem smaller.  Contracting the Tri-View is
like zooming in for a telephoto effect, the field of view gets
smaller and objects seem magnified.  You aren't changing
the size of things, just changing the way you look at them.

    It is easy to get lost in the Tri-View--if you contract it too
much there may be nothing visible at all.  Similarly, moving
the Tri-View too much will move you away from the
inhabited part of the universe into a region of empty space.
You can always use Expand Tri-View to 'step-back' and
figure out where you are.

    When you run the program, the Tri-View always starts
out at the center of the Sculpt universe, and it always starts
out the same size relative to the universe.  You can always
get back to this setup if you get lost.  To return to the center
of the universe, hold down the right Amiga key and click on
the Center Tri-View Gadget.  To make the Tri-View its
original size, hold down the left Amiga key and click on the
Center Tri-View gadget, the Tri-View windows will be re-
drawn in their original size and the effects of any expansion
or contraction will be removed.

    Let us look at one more Tri-View Gadget before moving
on to more interesting things.  This gadget is called the

                                                                        2-17

Reverse Gadget, and it lies just below the top of the right
border.  It performs a mirror image swap of the particular
window.  Alternatively, you may think of it as reversing the
direction of view, looking at the opposite side of our 3D
box.  A north view is turned into a south view, for instance.
Our little house is so boringly symmetrical that you may not
even notice the reflection, except that the letters indicating the
direction near the left and right borders change places.  With
a more complicated object, this gadget can be more useful.

Making a picture

    If all that Sculpt Animate could do was to manipulate
objects in wireframe, it would be no more than a three
dimensional computer aided design package.  But Sculpt
Animate has extensive image generating capabilities.

    Sculpt Animate makes use of a special object called an
observer.  The observer can be placed at any location in a
scene, and can be instructed to look in a particular direction.
In addition the observer can employ one of several imaging
techniques.  The acts of the observer are akin to
photography, and to follow this analogy the observer can be
equipped with different camera, lens, and film combinations.

    Let us look at how the observer operates by providing
instructions to make a picture of the little house.  First we
must pick a location from which to view the house.
Photographers know that to make a picture of an object with a
regular lens, they must stand far enough back so that the
object's image fits within the film size.

    To make room for our observer to move back, use the
Expand Tri-View Gadget until the house appears to be about
one eighth of the size of the window.  Going to the
downward looking view, position the cursor at a reasonable
distance from the house.  Check in one of the other windows
to see that the cursor is at some sensible distance above
ground level, about as high as the top of the wall will do
fine.  Move the cursor, if necessary.  When the cursor is
where you want it, hold down the right mouse button to

                                                                        2-18

bring up the menus.  Go to the OBSERVER menu and pull
the pointer down to the word LOCATION.  Release the
button.

    If you look at the Tri-View closely you will see that a
tiny circle has been drawn at the cursor location.  That is the
observer!  It will be easier to see if you move the cursor
away.  The observer will stay fixed until you next perform
an OBSERVER LOCATION command from the menus, or
you load another scene from disk.

    The next thing to consider is in what direction the
observer's camera is pointing.  Although the observer is not
equipped with feet, you will be disappointed with the picture
if the camera is pointing downwards.  Rather than have you
specify camera bearings in degrees, Sculpt Animate just lets
you specify a target position.  The observer then points the
camera at the target so that it lies in the center of the image.
A good target would be the center of the house.  Move the
cursor there, checking in two windows to see that the
position is correct.  Hold down the right mouse button and
select TARGET from the OBSERVER menu.  Although you
may not be able to see it for all the clutter of the house, a
small x will be drawn at the target location.

    For the observer's camera to record an image, there has
be light.  Some background illumination is provided as a
default, but this gives a flat, uninteresting image by itself.
The primary light source in most Sculpt Animate scenes is
one or more lamps.  To put a lamp in this scene, place the
cursor somewhere near the observer, a little to one side and
above will do, then hold down the right mouse button again
and select LAMP from the EDIT ADD submenu.  A small
white star should appear at the cursor's position.  Lamps
shine in all directions, so you don't have to worry about
pointing them, and the camera's exposure setting will
automatically be adjusted to account for the extra light.

    Now you can instruct the observer to start take a
picture.  Go to the OBSERVER menu and select START.  A
tine picture will appear near the bottom of the screen, and
after a few seconds will start 'developing'.  If you have not

                                                                        2-19

accidentally changed any of the observer's attributes, then
the picture size will be TINY and the image type will be
SNAPSHOT.  These settings will be explained a little more
later, and fully in the next chapter.

    While the image is rendering ('developing'), and after
it's finished, you may want to move it out of the way.  You
can do this as follows.  Point to the picture and first press
the left mouse button.  Release it and press the right mouse
button.  The picture should vanish.

    The picture is still in the computer's memory, however,
and can be recalled by going to the PROJECT menu and
selecting SHOW IMAGE.  This brings the Amiga 'screen'
on which the image was rendered 'to the front'.

    Let us try a different image of the same screen.  Go back
to the OBSERVER menu and select MODE PAINTING.
This mode renders more quickly than SNAPSHOT, so we
can afford to make the image larger.  Go to the OBSERVER
menu again and select IMAGE SIZE FULL.  Finally, select
OBSERVER START.

Saving an image

    The image that you have just rendered can easily be
saved to disk.  Hold down the right mouse button and go to
the PROJECT menu, follow down to SAVE and then across
to IMAGE.  Then release the button.  A file requester will
appear on the screen, asking for the name that is to be used
when the image is saved on the disk.

    If you want the data to go to a particular volume or
drawer, click on the appropriate names in those listing
boxes.  You will probably want to use a new name for this
file.  Next to the heading 'File:' over the files listing is a box
called a 'text entry gadget'.  You may click on this box, then
type in a new name for the image.  When you hit the
RETURN key, the image will be saved.

                                                                        2-20

    If you want to save over an old image, just double-click
that image's name in the file listing box.  Since the file
already exists, the program will ask you to confirm that this
is really what you want to do.  Saving data using an old
name will cause the old data to be destroyed.

    Images are saved in a standard 'IFF-ILBM' file, which
may be loaded into some other Amiga programs, such as
paint programs, etc.  Sculpt Animate supports a wide range
of image types and sizes, though, so not all existing
programs will be able to use all Sculpt Animate images.

    Images are not given icons.  You may copy, list and
delete images using the Amiga's CLI interface.  All image
files are given the name extension '.image' by Sculpt
Animate.  This is how the program keeps track of what files
are images, what files are scenes ('.scene'), etc.

Saving a scene

    The contents of the Tri-View an be saved to disk with
the command PROJECT SAVE SCENE.  This requester
works the same as the SAVE IMAGE requester.  After
choosing a file name with this requester, a second requester
will appear, asking you what parts of the scene you wish to
save.  This is very much like the requester that appeared
when you loaded the little house.  Usually, you will just
leave all the gadgets set to 'YES' and the 'which?'
gadget set to 'ALL'.  The meanings of these gadgets will be
described in chapter 6.

Leaving Sculpt Animate

    If you wish to perform other tasks while Sculpt
Animate is still running, you can push its screen back with
the Back Gadget at the top right of the screen.  You are then
free to use the Amiga's multitasking capability, as long as
your machine has enough memory.

                                                                        2-21

    Be sure not to forget that Sculpt Animate is running,
however!  You will waste memory if you start twice.

    To exit Sculpt Animate completely, execute the menu
command PROJECT QUIT.  Because you may have some
valuable data, scenes or images, that you have neglected to
save to disk, Sculpt Animate will ask for confirmation
before actually stopping execution.

If You have problems

    It would not be too much to hope that everyone who
ever uses Sculpt Animate is able to do so without any
trouble, whatsoever.  It would probably be too much to
expect, however.  This section attempts to provide some
guidance on how to deal with trouble, should it occur.

    If you have been successfully using Sculpt Animate to
follow the text up to this point, then you have already cleared
some major hurdles--especially if you are new to computers.
Your Amiga is plugged in and connected together properly,
and probably free of any grossly debilitating defects.  You
have begun to master the boot-up process, the Workbench
screen, and the vital process of backing up floppy disks.
Furthermore, you have managed to run Sculpt Animate and
operate some of its menus and other controls, and to grow
somewhat accustomed to the style of this manual.  In short,
you have gained a valuable foothold on the rocky slope of
the learning curve, and you can skip right now to the next
chapter.  This is by far the most likely circumstance.

    If, on the other hand, your efforts so far have not been
entirely successful, do not be too discouraged.  Rest assured
that many people have been using this program and its
predecessors for some time, with good results.  Any
problem you may be having is only a setback, not a defeat.
What follows is essentially a short course on troubleshooting
any computer problems, as applied to Sculpt Animate.

    If your Amiga is simply not working, don't fail to
check the obvious--be sure all of its connectors are properly

                                                                        2-22

plugged in.  Even the most experienced computerists have,
at one or time or another, been tripped up by a loose cable.
Other problems might be a bad system disk, or perhaps even
a hardware problem.  Don't hesitate to call your dealer if you
can't find the answer in your Amiga's manual.

    If you have trouble backing up the Sculpt Animate
disks, or if the program won't load, it could be a problem
with either the distribution disk or with your Amiga's disk
drive.  The first thing to try is using the disks in another
drive, either your second drive or a friend's drive.  If the
program works in another drive then the hardware is
probably at fault.  Otherwise, the program disk may be
flawed, perhaps because of heat, cold, X-ray or magnetic
damage incurred during shipping.  If the disk seems to be at
fault, read the warranty statement that was enclosed with the
program, it will explain how to obtain a replacement.

    Finally, if the program runs, but doesn't behave as you
expect, there are a number of things that could be wrong.
We will consider these one by one.

    Of course, there is the possibility of bugs (errors) in
Sculpt Animate's documentation or software.  Few, if any,
programs escape this problem entirely, but it is always the
goal.  Every effort has been made to ensure that Sculpt
Animate is bug-free, and we continually hunt for bugs.

    If you encounter a situation in which the program
behaves in a confusing manner, contrary to the
documentation, please write down step-by-step instructions
telling how to re-create the problem and send them to Byte
By Byte (Attention: SA4D Bug Reports).  Be sure to include
specifics about your Amiga's model, peripherals,
enhancements, and system software.  Also note the Sculpt
Animate version number you are using; this is displayed at
the top of the screen when you run the program.

    We would also like to hear of any changes or additions
you would like to see made to the program (Attention: SA4D
Enhancements).  Updates and new releases will be made
available periodically.

                                                                        2-23

    Before you file a bug report, read the remainder of this
section.  It describes the majority of problems that users
have had in the past, and could provide a quick solution to
your particular problem.

    First, the obvious.  It is very possible that the program
simply doesn't do what you expected it to.  Sculpt Animate
is a remarkable program in many respects, and it does things
that has never been done on personal computers before.  To
do these things, it must behave differently than, for instance,
a paint program.  The nature of solid object modeling on
microcomputers makes it impossible to actually 'draw' in 3-
D, the way you would draw on paper.

    If you encounter confusing behavior, read the related
part of this manual carefully, remembering that every method
was designed for a specific purpose, with the Amiga's
limitations in mind.  Even if what a function does is not quite
what you expected, in most cases it is the best or only
solution to a difficult problem.  Never forget the value of
simply doing something else for a while--the problem will
usually seem simpler the second time you look at it.  If you
simply cannot fathom some aspect of the program's use,
consult the support policy statement that is enclosed with the
program for information on how to obtain special help.

    At times Sculpt Animate may seem to take an extremely
long time to do something.  This is difficult to pin down,
because different scenes in Sculpt Animate can present
entirely different problems to be solved.  Extremely simple
scenes, such as the example in this chapter, should never
take more than a few minutes for any operation.  If you
increase the size and complexity of your scenes gradually,
then you will surely develop an idea of how long things
should take.  The following chapters explain some of the
factors that can make a scene develop more or less quickly.

    There are also some times when the amount of RAM
memory in the system is inadequate for a particular task.
This situation is described more fully in the next chapter
('Memory usage').  Also, users working with floppy disks

                                                                        2-24

will soon wish for more spacious storage media, chapter 12
has some information on how to do without.

    A more insidious problem can occur if the program
'crashes' or 'hangs up'.  Again, this could be a hardware
problem, especially if it seems to happen at random, and it is
good to try the program on one or two other Amigas if
possible.  Sculpt Animate exercises the Amiga hardware as
much as any program possibly can, using more memory and
performing more intensive calculation than had previously
been the norm (it has even been used to 'acid test' memory
board designs).  If any part of a computer or peripheral is
'marginal', then it is more likely to glitch when running
Sculpt Animate than when running most ordinary software.

    If the crash or hang-up always occurs when you
perform a particular task, something might be wrong with
the Sculpt Animate program file itself, even if the program
starts up OK.  It is certainly worthwhile to try making
another working copy of your original disk, although this
will not help if the original itself has been corrupted.  Again,
see the warranty information for replacement instructions.

    Crashes are also certain to occur if you run Sculpt
Animate from the CLI without first setting the CLI's stack
to about 50,000 bytes (or more), or if you use the wrong
version (before 1.2) of the Amiga's Workbench and
Kickstart disks.  Problems can also occur if system files on
the Workbench disk are somehow corrupted, so it is a good
idea to try rebooting with a different copy.

    It might seem alarming that there are so many things
that could go wrong when you are using Sculpt Animate, but
these are only things that could go wrong, and usually don't.
The program and its predecessors have been intensively
tested for literally thousands of hours by Byte By Byte, and
used by thousands of people.  The single most often
reported 'problem' with Sculpt Animate is that users can't
keep their hands off of it, it is so absorbing!

                                                                        2-25

                                CHAPTER 3

                               THE OBSERVER

    In the last chapter you were briefly introduced to the
observer, a feature of Sculpt Animate that controls how
images are produced.  In this chapter we will look at the
observer in more detail and describe the different kinds of
images that can be produced.

Imaging modes

    In order to record images, a real-life observer has to have
a camera, and the camera needs film.  What kind of film
determines the quality and resolution of the resulting image,
as well as how quickly it can be captured.  Although your
computer doesn't use film, there are still a number of
different ways for it to 'take a picture' of the objects in your
scene.  These methods all differ in speed and realism, just as
various types of film do.  The MODES submenu in Sculpt
Animate lets you set the kind of 'film' the Sculpt Animate
observer will use.

    The first mode in the menus is WIRE FRAME.  This is
the highest-speed rendering mode, but the lowest in quality.
Wire Frame yields two-color images that closely resemble
the objects seen in the Tri-View, with one significant
difference.  You see, while Tri-View windows display a flat
'orthographic' projection of their views, Wire Frame shows
you a true 'perspective' view, as seen from the observer's
viewpoint.  Only edges are shown, but each edge is shown
in the same place it will appear in the more sophisticated
modes.  This makes Wire Frame an excellent mode for
taking a look through the observer's 'viewfinder' before
rendering in a slower mode.  The command EDIT MODIFY
WIRE-FRAME COLORS can be used to change the colors
used in wire frame renditions.

                                                                        3-1

    The rest of Sculpt Animate's modes render objects as
solid shapes in multiple colors.  The simples way to do this
is to display each visible portion of the objects, with color
and shading determined by the specified light sources.
Sculpt Animate calls such an image a SKETCH or a
PAINTING.  These modes include the effects of perspective
and the resulting pictures look reasonably solid and realistic.

    As we shall see in chapter 4, the objects in the world of
Sculpt Animate are made up from triangular pieces.  In these
two modes, the image of each piece is 'painted' with a single
color.  While this can yield good results, the image is not
strictly accurate.  For instance, if one end of a triangle lies
closer to the light source than the other, it will not seem more
brightly lit, as it should.  Most objects will also have a
faceted appearance because adjacent faces cannot be
'blended' together.

    SKETCH mode differs from PAINTING in two ways.
Firstly it is faster, but a price is to be paid for the speed, so
SKETCH is also less accurate.  Sometimes one object that
should be hidden by another will be incorrectly drawn in
front.  This is because SKETCH mode uses a naive way of
determining which of two objects is in front of the other.  An
average distance from the object to the observer is calculated;
objects with closer average distances are assumed to lie in
front of other objects.  This is not always correct, so some
images produced in this mode will be flawed.  The speed
advantage makes SKETCH an attractive way or previewing
a scene in any case.  PAINTING mode resolves depth
ambiguities by breaking up some faces into smaller ones.
This does well, even for faces that intersect, but it requires
extra calculations so PAINTING takes longer than SKETCH
to generate an image.

    The SCANLINE PAINTING mode yields result that
look similar to PAINTING, but it employs a different
approach to creating the image.  Briefly, it sorts faces much
as PAINTING mode does, resolving intersections.
SCANLINE PAINTING does this pixel by pixel, however,
which can improve speed as scene complexity grows.
SCANLINE rendering's advantage increases with objects

                                                                        3-2

that cover a relatively small portion of the image, or objects
that have many intersecting faces.  Its output is similar to that
of PAINTING (hence its name).

    These three rendering modes are fairly limited in the
number of colors which they can display.  This is because
they use the Amiga's 'normal' video display capability of 16
or 32 colors in any image (or fewer).  Images rendered in
this mode can be written to disk and subsequently used in
almost any Amiga 'paint' program.

    Sculpt Animate's higher-quality imaging modes all use
the Amiga's Hold and Modify (HAM) facility.  The
mechanics of HAM are rather complex, but its effect is to let
these modes display as many as 4096 different colors on the
screen at once, at the cost of some horizontal resolution.
With all these colors available, it is possible to display
images with subtle shadings and blends of color.

    A HAM image's many colors are derived from a 'map'
of only 16.  This can make color transitions difficult, so
sometimes a false color fringe will be seen between areas of
different color.  As soon as Sculpt Animate finishes
rendering an image, it will pause for a time as it finds the
best possible map of 16 colors to minimize these fringes,
then the image will be re-drawn with the new color map.

    The fastest imaging mode to use HAM is SCANLINE
SNAPSHOT.  This mode uses the additional colors to
enhance realism by shading color smoothly across faces,
and, if desired, smoothing adjacent faces together.  A face
may also be made 'shiny', so that spectacular highlights will
appear on its surface wherever appropriate, as determined by
the scene's lighting.  Color, smoothing, and surface texture
may be set differently for various faces.  This will be
discussed in Chapter 4.  Only the DULL and SHINY
textures are used in SCANLINE SNAPSHOT.

    The next step in increasing realism is to use a technique
called ray tracing, which traces rays of light from the
observer's eye to their source, computing an intensity and
color for each pixel on the screen.  A pixel ('picture

                                                                        3-3

element') is the smallest area of the screen that can be given a
distinct color and brightness.  The number of pixels in an
image will vary according to its size and resolution, as will
be discussed later in this chapter, but a typical image might
have hundreds of thousands of pixels; meaning hundreds of
thousands of rays to be traced to an indefinite length.
Understandably, ray tracing taxes the Amiga's computing
power.

    The rewards are worthwhile, though.  Like SCANLINE
SNAPSHOT, ray tracing can determine shading and color
smoothly across flat surfaces, and smooth rough edges.
Ray tracing can go still further, though.  Since each pixel's
light value is traced all the way back to its source, rays can
be traced through transparent objects, and off of reflective
surfaces, automatically creating accurate and realistic images
of objects that would be very difficult to draw or paint.
Uncanny realism can be achieved.

    Sculpt Animate calls its simplest and fastest kind of ray
traced imaging a SNAPSHOT.  Lacking only shadows, this
mode can still create very realistic images, rivalling
photography in some cases, where it not for display
limitations.  Still more realistic is PHOTO mode, which
shows all the ray tracing prowess of SNAPSHOT, plus
distinct, realistic shadows cast by objects on the ground and
each other.  PHOTO mode's extra calculation incurs an extra
time penalty, though, so if shadows aren't likely to be a big
part of your image, feel free to use SNAPSHOT or
SCANLINE SNAPSHOT.

    Whatever mode you use, the time taken to generate an
image depends upon the complexity of the scene and the size
of the image, as well as the imaging mode.  The time needed
can vary from a second or two for a WIRE FRAME of
something simple to many hours for a difficult PHOTO.

    With SCANLINE, SNAPSHOT and PHOTO modes, it
is a good idea to try things out with a TINY or SMALL
image size, because rendering time varies with the size of
the final image.  Of the three HAM imaging modes,

                                                                        3-4

SCANLINE SNAPSHOT is generally fastest, although
SNAPSHOT may be faster for certain simple scenes.

    Now that you know about the various imaging modes,
you need to know how to select one.  You can specify which
is to be used with the OBSERVER MODE menu command
(From now on in this book, we will assume that you have
mastered the operation of the menu).  The MODE submenu
lists the choices with a checkmark to show which mode is
currently active.  If the marked item is not the one you want,
select the correct menu item in the usual way.  The
checkmark will move to reflect the change.

    The OBSERVER MODE command also lets you choose
the screen resolution in both vertical and horizontal
directions.  Resolution determines the amount of detail
which can be displayed in a given area of the screen.  On the
Amiga, resolution can vary in either direction.  LO-RES and
HI-RES refer to low resolution and high resolution
horizontally; NO-INTERLACE and INTERLACE refer to
low resolution and high resolution vertically.  Thus, a LO-
RES, NO-INTERLACE image would be the Amiga's
grainiest image, HI-RES INTERLACE its sharpest.

    As is usually the case with computers, every
improvement carries a cost.  The highest image resolution is
not available in every imaging mode.  Specifically, the
Amiga cannot display a 32-color 'normal' image or any
HAM image in HI-RES.  If HI-RES is selected, all
SKETCH, PAINTING, and SCANLINE PAINTING
images will be automatically limited to 16 colors.  If any of
the HAM modes is used, a HI-RES setting will be ignored,
and a LO-RES image generated instead.

    None of these limitations apply to INTERLACE, but it,
too, carries a cost of sorts.  Some high-contrast images will
flicker noticeably in INTERLACE resolution--this is normal
on the Amiga.  Special 'high-persistence' monitors are
available which can reduce or eliminate this problem.

    Each of the higher-resolution settings doubles the
number of pixels to be rendered.  In SKETCH or

                                                                        3-5

PAINTING mode this will have little effect on speed, since
the number of calculations depends mostly on the number of
faces.  In the SCANLINE and ray trace modes, however,
each pixel is calculated independently, so doubling the
resolution will more or less double the rendering time.  In
any case, the higher the resolution images will require somewhat
more RAM and disk space.

    Sculpt Animate will generally assume that you want to
use as many colors as possible in a given imaging mode.
The command OBSERVER MODE BIT PLANES gives you
some control over the number of colors that will be used, but
you need to know how this works.  Computer screens, and
hence computers, create colors by mixing red, green and
blue light.  The brightness of each color is specified by a
number.  On the Amiga display, this number can range from
zero to fifteen.  Thus you have sixteen levels of light each
for red, green, and blue, making 4096 total combinations.

    In the normal (non-HAM) display modes, it would
require too much memory for the Amiga to directly display
all 4096 colors in an image, so a particular image may only
use a subset of the available colors, 16 of them in HI-RES
and 32 in LO-RES.  Sculpt Animate attempts to pick the best
set of colors for a given scene.

    Some Amiga programs impose further restrictions on the
number of colors that can be used.  If you plan to use your
images in such a program, you must issue the command
OBSERVER MODE BIT PLANES.  A requester will pop up
asking you how many bit planes you desire.  One bit plane
permits the use of two colors.  Each additional bit plane
doubles the number of colors as follows:

                Bit planes      Colors
                    1              2
                    2              4
                    3              8
                    4             16
                    5             32   (LO-RES only)


                                                                        3-6

    If you specify zero or six bit planes, Sculpt Animate will
use the greatest number possible for the selected imaging
mode.

    Bit plane values greater than six would normally have no
meaning on the Amiga, but in Sculpt Animate there is one
special case.  By specifying 24 bit planes (equivalent to
16,777,216 colors), you can have a raw red, green, and blue
(RGB) image data written out to disk files.

    Most users of Sculpt Animate will never have any use
for raw RGB data, so you will probably want to skip to the
next section.  The ability to write RGB images is provided
so that specialized hardware and image processing software
may use Sculpt Animate images.  For the time being, then,
such things are primarily of interest to programmers and
other technically inclined users.  For those users, here are
some specifics about the raw RGB facility.

    When you ask for 24 bit planes, you will be asked for
three file names, the width and height of the desired image in
pixels, and each pixel's ratio of width to height.  These
settings will only take effect if the image is rendered with the
START menu command, and if the imaging mode is
SCANLINE (either kind) SNAPSHOT, or PHOTO.  RGB
data for animation must be handled through a frame buffer
device driver, as explained in chapter 12.

    Three files will be written, each will contain an image of
the scene in either red, green, or blue.  The data in each file
is stored as one eight-bit brightness value for each pixel.  If
fewer bit planes are required, your special-purpose software
will need to mask off the unused low-order bits and pack the
data as needed.  The pixel bytes are stored serially, one scan
at a time, beginning with the pixel at the upper left hand
corner of the image.  There will be no header information,
just data.  Be sure there is enough disk space for each file, a
704 by 480 image will need 337920 bytes for each file.  You
may type different path names into the three filename
requesters.

                                                                        3-7

    The requested width and height values refer to the width
and height of the image, in pixels.  This depends on your
application, which is up to you.  Except for storage, there
are no known limits to the height values which may be
entered.  The maximum width should be 32768 pixels--
values up to 1024 have been tested with no problems.

    The pixel width and height values refer to the aspect ratio
of pixels on the screen which is to display the image.  For
instance, a 100 pixel square on an Amiga screen is usually
about one tenth higher than it is wide, so its width:height
ratio would be 10:11.

Placing the observer

    In the last chapter, we explained how to change the
observer's position, but we will repeat it again (in case you
look here first).  First you must select a position in the Tri-
View with the cursor.  You may need to make the Tri-View
bigger, so you can place the observer far enough back to get
a reasonable view.

    When the cursor is at the right place (look in at least two
windows), go into the menu and select OBSERVER
LOCATION.  When you execute this command, a small
circle will appear in each window, to mark the observer's
location.

Aiming the observer

    The observer must be told which way to look.  Although
you may know what is the most interesting object in the
world of Sculpt Animate, the observer doesn't.  As we
explained in Chapter 2, the observer always points towards a
special spot called the target.  This is specified in a similar
manner to the way that we picked the observer's location.
We move the cursor to the target position and select
OBSERVER TARGET from the menu.  A small x will mark
the spot.  The target is usually placed near the most

                                                                        3-8

interesting part of the scene, so that part will appear in the
center of the image.

    Never be timid about using the observer as an editing
tool.  As you continue with this manual, you will learn how
to create objects, color faces, and arrange scenes.  It will
often be helpful to move the observer and target in order to
view a point of interest, then 'take a look' with either the
WIRE FRAME or SKETCH imaging mode.

Camera lenses

    We have already drawn an analogy between the
observer and a photographer wielding a camera.  If the
camera is equipped with a zoom lens, it can zoom in for a
telephoto shot where a small angle of view fills the whole
picture.  Alternatively the camera can zoom back so that the
field of view is wide.  In a similar way, the observer can be
supplied with a lens of your choosing.  Use the OBSERVER
LENS menu command to examine and change the current
selection.

    The observer has three standard lenses that should
suffice for most scenes.  Alternatively, you may select the
OBSERVER LENS SPECIAL option.  When you select this,
a requester will prompt you to enter the focal length of a lens
of your own design.  The smaller the number that you enter,
the wider the angle of view, and the smaller an object will
appear.  For instance, of the three standard lenses provided,
the NORMAL setting corresponds to a SPECIAL setting of
50, WIDE-ANGLE corresponds to 28, and TELEPHOTO
corresponds to 135.

    In general, these numbers are analogous to millimeter
focal lengths of lenses on a typical 35-mm camera.  If you
enter a number in the telephoto range of 135 or larger,
distant objects will appear larger than normal, with a reduced
sense of perspective.  If you enter a very small number, such
as 10, then you will get a fish-eye effect with grossly
exaggerated perspective.

                                                                        3-9

    Astute photography buffs will notice that the camera
analogy fails slightly, especially when a fish-eye lens is
used.  This is a rather obscure artifact of geometry, so you
can ignore the explanation and skip to the next section if you
like.

    In a 'real' photograph, long parallel lines will tend to
curve towards each other near the edge of the image.  This is
because camera lenses have a finite 'aperture' through which
light is admitted, and the light is refracted unevenly across
this aperture.  In Sculpt Animate, the observer's lens has an
infinitely small aperture, it is 'perfect', so no such distortion
occurs.  This difference also gives the observer's lens an
infinite depth-of-field, so objects can never be out of focus.

    So which is 'right', the real lens or Sculpt Animate's?
Both, and neither.  Both lenses have their advantages.  The
real lens's field distortion and limited depth of focus can be
used to create exciting effects, but the total removal of these
characteristics presents the Sculpt Animate user with a
creative situation that never existed with a 'real' lens, and
some exciting new results are bound to arise.

    To say that either kind of lens is more 'real' is really
incorrect.  When their images are intended to resemble what
the human eye would see, they both do pretty well, but
neither is perfect.  The only time when their results differ
greatly is when they are used to create effects that look
'unreal' anyway.

Exposure control

    Most modern cameras have an automatic exposure
control so that pictures don't come out too dark or too light.
Sculpt Animate simulates this feature.  This means you
usually don't have to worry about exposure when you add
or move lamps in a scene.

    Like the automatic exposure control on a camera, most of
the time Sculpt Animate does a good job of controlling the
brightness of the image.  It works according to a simple rule:

                                                                        3-10

The object in a scene that is closest to a lamp will be brightly
lit, but not so brightly that it washes out.

    Sometimes, perhaps for artistic reasons, you may want
to change the exposure.  Select OBSERVER EXPOSURE
MANUAL and a requester will pop up asking you to
override the automatic setting.  Enter a number as a
percentage of the standard value.  If you want your picture to
be twice as bright as normal, enter 200.  Because the Amiga
screen has a very limited range of brightness, the image that
you get may look flat and washed out.  Similarly you can
enter a number less than 100 if you desire a gloomy and
somber effect.

The Jaggies

    Computer images are built up from discrete pixels.  Even
with 640 by 400 pixels, image is still ten times more
coarse than a 35mm slide.  The regular layout of pixels
makes them evident to the eye as stair-step bumps,
particularly along shallowly slanted borders between two
very different colors.  Engineers and other technically-
minded folk call these obtrusive artifacts 'aliasing', but most
people just call them 'the Jaggies'.

    When images are rendered in HAM mode, it is possible
to use a technique called anti-aliasing to reduce the effects of
the jaggies.  Anti-aliasing hides the jaggies by blurring sharp
color transitions slightly.  It will increase rendering time
somewhat.  The command OBSERVER ANTI-ALIASING
can be used to control the type of anti-aliasing that is
employed.

    Which is better-blurring or jaggies--will usually depend
on the subject matter and your own taste.  You should
probably try an image with and without anti-aliasing to see
which you prefer.  Images with lots of high-contrast color
transitions will often need anti-aliasing, especially when they
are used in animations.

                                                                        3-11

Dithering

    Although the Amiga display hardware can utilize 4096
colors, this is not enough if you want the highest possible
image quality.  Each of the primary colors can only take on
one of sixteen values, so colors jump by almost 8% from
one level to the next.  The human eye can easily detect this
transition as 'color bands' in the finished image.  In the
HAM imaging modes, Sculpt Animate usually employs a
blending technique called 'dithering' to hide these
transitions.  Dithering works well, but it adds graininess to
the image.

    You can control this graininess with the command
OBSERVER ANTI-ALIASING DITHERING.  A requester
will pop up asking what percentage of standard dithering is
needed.  A value less than 100 will reduce the graininess at
the expense of the color jumps being more obvious.  A value
of zero will eliminate dithering entirely, causing distinct
color banding.  Values greater than 100 cause more dithering
to be applied.

Image size on the screen

    You can select the size of image that Sculpt Animate
creates.  Usually you will want the image to occupy the full
size of the screen, but there are a few reasons why you may
want to make a smaller image.

    If you are planning to import your picture into one of the
Amiga paint programs, then a small picture may be all you
need, particularly if you are using the image as a brush.

    Another advantage of a small image is that you can
continue to work with the Tri-View, with the picture
displayed at the bottom of the screen.  If you point to the
screen near the top of the black area to either side of the
picture, and hold down the left mouse button, you can drag
the image up the screen or push it down so that it almost
vanishes.  To do this, you have to hit the black area just
right, between about four and twelve scan lines from the top.

                                                                        3-12

    Small images also spare your time and system resources.
They will always require less memory for storage and
display than larger images, and in the scanline and ray-traced
modes, smaller images render much more quickly.  It's
usually a good idea to render a small image to check
shadows and lighting before starting a full size picture.

    Take note that the initial scene data processing and
optimization will be the same for all images of a certain
scene.  Still, in the rendering phase a TINY image should be
about 64 times as fast as a FULL one.

    You change the picture by selecting a choice from
the OBSERVER IMAGE SIZE menu selections.  The actual
size in pixels of the resulting images will vary from country
to country.  This is because two different video standards are
common:  NTSC, which is used in the USA and Japan, and
PAL, which is used in Europe and elsewhere.  Sculpt
Animate automatically adjusts both its Tri-View and images
to your Amiga's video display.

        Here are the image sizes (in pixels) that may be used:

        Size            NTSC            PAL
                    Width x Height  Width x Height

        TINY          48 x 25         48 x 32
        SMALL         80 x 50         80 x 64
        MEDIUM        160 x 100       160 x 128
        FULL          320 x 200       320 x 256
        JUMBO         352 x 220       352 x 282
        VIDEO         352 x 240       352 x 290


    The above values are for lo-res, non-interlaced images.
If INTERLACE is employed, there will be twice as many
vertical pixels, but the image size will be the same.  If HI-
RES is used there will be twice as many pixels horizontally,
except in the HAM modes, where HI-RES has no effect.

    You can force the program to use one video standard or
another with the Workbench menu 'Info' facility.  Do this by
selecting the SA icon and calling its 'info' window.  Click
ADD ont he 'Tool Types' string gadget on that window,

                                                                        3-13

type in either 'TVTYPE=NTSC', or 'TVTYPE=PAL', hit
RETURN, and click SAVE.  The setting takes effect when
you next run the program by double-clicking its icon.

Tilting the camera

    Many photographers have difficulty holding a camera
level.  You have doubtless seen photos of perfectly upright
buildings that resemble the Leaning Tower of Pisa, or of
people who seem to have lost one shoe.  In Sculpt Animate,
the observer always holds the camera straight unless told to
do otherwise.  To change this, you can use the OBSERVER
TILT menu command to specify an angle of tilt, in degrees.
A positive value will tilt the camera to the left, negative
values tilt it to the right.  You can use a value of 90 to create
a vertical format image, or a value of 180 to make the image
come out upside down.

    Most of the time it is easy enough for the observer to
hold the camera upright, since it's usually pretty obvious
which way is up.  But what if the observer is directly over
the target?  No one camera orientation makes more sense
than any other in this situation, because 'up' isn't in the
picture!  In this case, the program will pick an arbitrary
orientation.

    What is more likely to happen in this situation is that the
camera will be almost, but not quite, directly over the target.
If it is a little bit east of the target, then 'up' will be west. If
it is a little bit northwest, 'up' will be southeast.  In other
words, 'up' is always the direction opposite the observer's
position.  In an animation, this can make the image seem to
wheel about drunkenly.  A similar behavior occurs when the
observer is beneath the target.

    If you must place the observer above the target, you will
need to do so carefully, but it's fairly simple.  As we've
seen, if the observer is to look downwards, the orientation
of the resulting image can be controlled by placing the
observer just a little bit to one side.  In chapter 13, you will
see how the coordinates window can be used to do this with

                                                                        3-14

precision, easily.  Only a tiny offset will be needed, so the
image will be indistinguishable from a true downward view.

    If you're surprised that 'up' and 'down' can cause us so
much trouble, don't be.  We're in good company, some of
the largest telescopes in the world have trouble looking
straight up, for similar reasons!

Starting the image making process

    After you have set up a scene and the observer's
attributes, you can start the picture making process with the
OBSERVER START menu command.  If a hidden image is
already being rendered (see the next section) you will be
informed, and nothing more will happen.  Otherwise the
image creation will commence.

    Image rendering will proceed differently in different
imaging modes.  As you use the program, you will learn
what to expect.  In the HAM modes, the image will usually
not be displayed until it is complete, but a status window
will display the phase and progress of the calculation.  You
can request that, if possible, the image be displayed as it is
rendered.  The command OBSERVER DISPLAY EARLY
does this.  OBSERVER DISPLAY LATE re-instates the
normal procedure.

    Sculpt Animate renders a HAM image initially in 4096
true colors, then analyzes the image to pick up the best possible
HAM color map.  If memory is low and there is enough disk
space, the program will save temporary image data to disk,
either in the current directory or in any mounted device
named 'SATEMP:'.  A message will warn you if there is not
enough RAM or disk space to render the image.

Hiding and restoring the image

    Once a picture has been created, or sometimes even as it
is being produced, you can hide the image from view.  This

                                                                        3-15

will only be possible if the normal system pointer is visible,
not if the grinding gears are going.

    To hide the image, first place the pointer over the image
itself and click the left mouse button.  This will make the
mouse pointer disappear, so don't move the mouse any more
or it might go off the picture.  Now if you click the right
mouse button the image will disappear.  You will usually
just want to click left-right in quick succession.

    When the image disappears, the Tri-View will become
visible.  Before you can do anything else, you must move
the pointer inside one of the Tri-View windows and click the
left mouse button.  This 'selects' the Tri-View, which makes
the menus and keyboard active.

    You can bring the hidden image back into view with the
menu command PROJECT SHOW IMAGE.  If you want to
remove the image entirely, freeing up the memory it uses,
use the menu command PROJECT UNLOAD IMAGE.

    Experienced Amiga users will find that the hidden image
is on its own Amiga graphics 'screen'.  There can be several
such screen present in the system at any one time, and there
are a number of ways to bring a particular screen into view
(see your Amiga documentation for details).  Sculpt
Animate's JUMBO and VIDEO sized images are larger than
the Amiga's standard screen, however, so they will not be
properly displayed unless you use PROJECT SHOW
IMAGE to bring them into view.

Aborting the image making process

    When the observer is instructed to start making a picture,
there is a period when the full capacity of the Amiga is
required.  During this time, the pointer will turn into the
grinding gearwheel image.  As soon as the regular pointer
returns, you are free to continue working with the Tri-View,
even though image generation is still in progress.  Changes
that you make to the scene or its objects will not effect the
image being generated.

                                                                        3-16

    If, after your normal pointer returns, you decide that you
do not want the image, you can use the OBSERVER
ABORT menu command to stop the processing.  After an
image generation has been aborted, the PROJECT SHOW
IMAGE command may be used to recall the partially
completed image, and PROJECT SAVE IMAGE can be used
to save it.  Use PROJECT UNLOAD IMAGE to free up any
memory that is involved in maintaining the fragment of
image.

Memory usage

    If you are curious about what resources are being used,
you can execute the command OBSERVER STATUS.  This
turns on (or off) a display at the top of the screen.  Most of
the items in this display are explained in the next section.  On
the left side of this display, the amounts of unused chip and
fast memory are shown.  Chip memory is required to display
screens, while both chip and fast memory can be used for
storing code or object data.  This distinction will be
explained more in a moment.

    When memory gets low, the program may not be able to
proceed.  Sometimes Sculpt Animate will recognize the
condition and simply refuse to perform a command, but at
other times the memory shortage occurs at a critical point and
there is no alternative but to halt the execution of the program
without warning.

    It is a good idea to keep track of memory usage, and
save any important scenes or images at frequent intervals.
Sculpt Animate is quite a large program, but all the code
need not be in memory at once.  When memory starts to run
low, unused code is jettisoned to allow space for more
important items.  When memory becomes available, the code
will be reloaded in anticipation of being used.  You can
control this automatic operation with the PROJECT LOAD
CODE and PROJECT UNLOAD CODE commands.  The
only drawback of having code unloaded is that a delay will
occur before some commands can be executed, as the code
must be recovered from disk.  If you are working with

                                                                        3-17

floppy disks, you will be asked to insert the program disk
when this happens.

    This is not the only time when Sculpt Animate will load
code.  The program makes use of a facility of the Amiga
called 'disk loaded libraries'.  Most of the time you won't
need to worry about this, but if you are ever asked to
insert the Workbench disk (or whatever disk you used to
'boot up' they system), this is probably why.  Of course,
other programs running concurrently, including the
workbench screen, may also ask for a particular disk.  This
is normal on the Amiga.

    Just as the code can be loaded and unloaded, so can the
workbench screen.  The commands PROJECT UNLOAD
WORKBENCH and PROJECT LOAD WORKBENCH do
this.  Note that the workbench can only be unloaded if it is
not being used, and it can never be unloaded if a CLI
window is present.  PROJECT UNLOAD WORKBENCH
will do nothing in these cases.

    The terms 'chip memory' and 'fast memory', used
above, refer to Random Access Memory (RAM), this is
volatile (non-permanent) memory within the computer which
can be accessed directly by the microprocessor.

    Fast memory is the Amiga's general purpose memory.
All expansion memory on current Amiga models is classified
as fast because, theoretically, the microprocessor has free
access to the memory at all times.  Fast memory is where
programs and object data are preferentially stored.

    Chip memory may also hold these things, if there is no
more fast memory available.  More importantly, though,
chip memory is the only place where graphics display data
can be used, since it is the only memory which the Amiga's
graphics chips may 'borrow' from the microprocessor.  As
you can imagine, this makes it very precious on a graphics-
intensive machine like the Amiga.

    Current models of the Amiga can have no more than
512k of chip RAM.  This means that, no matter how much

                                                                        3-18

expansion RAM you have, there will still be times when an
operation cannot be performed, especially when you
generate images with HI-RES, INTERLACE, and JUMBO
or VIDEO set.  You may need to make it a habit to do an
UNLOAD WORKBENCH and make the Tri-View windows
as small as possible when working with such large images.

    Users in Europe, Australia, and other countries that use
PAL video may not be able to use PROJECT SHOW
ANIMATION (chapter 12) on animations rendered in these
large resolutions.  These animations require the stand-
alone playback program Movie, included on one of the
Sculpt Animate distribution disks.  See the Movie
documentation file for more information.

    It is reported that conversion kits will eventually be
available to upgrade the Amiga 500 and 2000 to one
Megabyte or more of chip RAM.

Other status display items

    As the last section explained, the command OBSERVER
STATUS calls up a program status display.  In addition to
the free memory readout already discussed, the status
display also lists how many vertices (selected and
unselected), edges, faces and lamps are in use.  If an image
is being rendered, a rough estimate of the remaining time is
also displayed.

    The vertex readout can be extremely useful.  It is easy to
count vertices just by selecting an object, or part of an
object, and looking at the status display.  This is especially
useful in creating motion paths for use in animations (see
chapter 10).  The edge and face readouts are primarily useful
as an indicator of how complex the scene is.  The more faces
there are, the longer the scene will take to render.  It is good
to be careful not to use any more faces than necessary.

                                                                        3-19

                                CHAPTER 4

                             BUILDING OBJECTS

    In this chapter we will describe the way that Sculpt
Animate represents solid objects, and how you can design
any shape with your mouse.

    Objects in Sculpt Animate are built up from individual
triangles.  In the Tri-View, each triangle is visible as three
edges joined at three vertices.  In this chapter we will
describe the lowest level of object building, dealing with the
triangles on an individual basis.

    It would be extremely tedious to build a complicated
object by constructing all of its triangles one-by-one.  In
chapter 5 we will be using some power tools that enable you
to conjure up complicated shapes with hundreds of triangles,
using only a few strokes and clicks of the mouse.

    You need to know how the low-level construction is
performed, though, to be in a better position to use these
power tools.  The low-level operations can also be used to
expand upon objects that are hewn by more powerful
techniques, increasing the total number of objects may
be created with ease.

Vertices

    The simplest entity that Sculpt Animate can handle is
called a vertex (the plural of vertex is vertices).  Vertex is
just a fancy name for a point at the corner of a triangle.  We
had better not call it a point, though, because that word is
already overused, we might otherwise have to say 'point the
pointer at the point!'

    If you remember how we located the observer within the
Tri-View, you will not be surprised at how we create a
vertex.  First we position the cursor at the exact spot where

                                                                        4-1

we would like a vertex (by pointing with the mouse and
clicking the left mouse button).  Then we issue the command
EDIT ADD VERTEX.

    There is an easier way to do this, just move the cursor to
the correct place and, while still holding down the left mouse
button, click the right mouse button once.  Instead of
bringing up the menu, clicking the right mouse button with the left
button down will create a new vertex.  It is portrayed in the
Tri-View by a bright yellow dot.

    Try making some vertices, but at this stage do not make
more than three of them, we will see why in a moment.
Create the three vertices in a triangular pattern about half the
size of the Tri-View.

Edges

    The next step in low level construction is to build edges.
Edges always end at a vertex.  Indeed, the main purpose for
having vertices is to define the position of edges.

    If you succeeded in creating three vertices, there is a Tri-
View gadget that you can use to join the vertices.  It is
located near the bottom of the left hand border of each
window.  Its icon is a triangle.  Click on it and see that dark
lines are drawn between the vertices.  This is how the Tri-
View portrays edges.

    If, for some reason, edges were not created, and the
screen simply flashed at you when you clicked on the Edge
Building Gadget, it was probably because you had created
more than three vertices.  This can happen if you accidentally
double-click the right button, creating two vertices in a spot
instead of one, or if you have already been playing with the
mouse.  You can clear away the vertices by using the EDIT
ERASE ALL menu command.  Create three new vertices and
try again to construct the edges.

    Why is the Edge Building Gadget so persnickety?  If you
had only created two vertices, it would have happily joined

                                                                        4-2

them with a single edge.  If you had a dozen vertices, the
Edge Builder would have had to make 66 edges to join all
combinations of vertices.  This might make a credible image
of a bird's nest, but it would have little other use.  This is
why the Edge Builder will only work with either two or
three vertices, building one or three edges, respectively.

    The command EDIT ADD EDGES has the same effect as
clicking the Edge Building Gadget.

Faces

    The next, and fortunately last, element that we must
consider is called a face.  In Sculpt Animate all faces are
triangular.  There is no need to provide a gadget to create
faces, for Sculpt Animate has a rule: whenever three vertices
are joined by three edges, a face is automatically created.
So, unbeknownst to yourself, you have already built a face.
It is hardly surprising that you did not notice, because faces
are not depicted in the Tri-View.  They don't have to be
shown, because whenever edges make a triangle, a face will
be there.

    Although faces are not rendered in the Tri-View, they are
the only parts of an object that are seen by the observer when
making an image in any mode except WIRE FRAME.
Vertices and edges are never seen in the final picture.

    To prove this, go to another window.  Your triangle will
probably look like a straight line in this window, because it
is being viewed edge on.  Move the cursor away at right
angles from the line-expanding the Tri-View, if necessary--
and locate the observer at this point using the OBSERVER
LOCATION menu command.  Next, move the cursor back
to the triangle, checking in two windows to be sure of its
position in three dimensions.  Then use the OBSERVER
TARGET menu command to point the observer at the
triangle.  Start the image making process with OBSERVER
START.

                                                                        4-3

    The picture that will be created will hardly be a work of
art.  Indeed, even if you have done everything exactly right,
you will only see the dark silhouette of a triangle.  This is
because we forgot to turn on a lamp.  Move the cursor back
to the vicinity of the observer and a small distance to one
side.  To create a lamp, execute the EDIT ADD LAMP menu
command.  This time, if you make a picture with the
OBSERVER START command, the triangle should be
flooded with light.  If the lamp was sufficiently to one side
and you are in SNAPSHOT or PHOTO mode, you may
even see a hint of shading across the otherwise bland
triangle.

    This really does seem like a lot of bother just to draw a
triangle.  Fortunately, we will soon be using tools that will
mass produce faces so that you can sculpt shapes of any
kind.

Selecting vertices

    Sculpt Animate uses a variety of methods for choosing
vertices with which to work.  It distinguishes between
selected and unselected vertices.  As a general rule, a vertex
must be selected before you can manipulate it any way
(the exceptions: 'connected' commands such as SNAP
CONNECTED in chapter 4; and spline manipulation in
chapter 8).  A selected vertex is portrayed as a bright yellow
dot.  The vertices that you created by holding down the left
mouse button and clicking the right button were shown in
yellow, so they were selected vertices.

    To change the state of selection of a vertex, you must
point at the vertex with the pointer and double click the left
mouse button.  Try it with one of the vertices you've made
in the Tri-View.  The color of the vertex will change from
yellow to a dull purple.  If you double click again, it will
change back to yellow.  Selected vertices are yellow and
unselected vertices are purple.

    In order to change a vertex, the pointer must be within a
few pixels of the correct spot.  If Sculpt Animate cannot find

                                                                        4-4

a vertex in the immediate vicinity, no vertex will be changed,
but the screen will flash briefly to warn you that you missed.
If this happens, move the pointer closer to the vertex and try
again.

    We say that double click selection works on the
'indicated vertex'.  Several of the commands you will learn
operate on the indicated vertex.  If you want to indicate
something, you do not usually have to position the cursor in
all three dimensions, two dimensions suffice to make your
wish know.  In other words, you only need to indicate a
thing in one window.

    The exception to this is when two or more vertices are
close to the cursor in the current window.  If this happens,
the vertex among them which is closest to the cursor in the
other windows will be chosen.  Double click selection will
cause the cursor to jump, in the other windows, to the
location of the desired vertex.  Make it a habit to look at
another window anyway, so that what you wanted is
what you got.

    Now that you know about selection and deselection, we
can tell you a little more about how the Edge Building
Gadget works.  It will construct between selected
vertices, provided that either two or three vertices are
selected.  If some other number of vertices is selected, the
screen will flash a warning and nothing else will happen.

    You now have enough information to create a triangular
net of vertices and edges.  Turn off the selection of all but
two vertices and use the Edge Builder to create a new edge.
Remember that each time you complete a triangle with three
edges, a face is constructed, and it is these faces that are
visible when you use the observer to generate an image.

    If you move from window to window, your vertices will
not all lie in one plane an you can construct solid shapes.
Try to construct a pyramid, and then see what it looks like to
the observer.  Don't worry if it isn't perfect, the exercise is
what's important right now.

                                                                        4-5

    Remember that you have to have at least one lamp
suitably placed or your constructions will just look like a
dark hulk.  To create a lamp, position the cursor at a suitable
spot, preferably behind, above and to the side of the
observer, and execute the command EDIT ADD LAMP.  If
you don't like the lamp where you've put it, for now just
erase it with EDIT ERASE ALL LAMPS, and add a new
one.  You'll learn how to move laps at the end of this
chapter.

    If you succeeded in fabricating a pyramid, try to build a
cube.  Here's a tip that can help keep it square: if you want
two vertices to 'line up' squarely in the Tri-View, place your
first vertex, then move the cursor to about the right place for
the second vertex.  In one of the other windows, the cursor
will seem almost centered on the first vertex.  In fact, if you
go to the other window and center the cursor directly on the
first vertex, you can make your second vertex and it will be
right in line with the first!  In this way, you can use the Tri-
View as a carpenter's square.  The precision is limited by the
screen, of course--chapter 13 explains some much more
powerful tools for precise placement of vertices, etc.

    Back to the example at hand.  Once you've laid out the
cube's basic shape and connected the corners with edges,
you still haven't made it a visible object.  In order to give it
visible faces, you will have to divide each of the square sides
with an extra diagonal edge, making two triangles.  If you
wanted to make something that looked like an open box, you
could leave out the diagonal on the top side.  Try doing this,
and look at the resulting image.

    More than likely, what you will see is an open box, but
there is a chance that you'll see something else, even if
you've done everything correctly.  Enough suspense--it just
so happens that if a box is constructed in just the right (or
wrong) manner, it will have one or more 'hidden faces'
inside it.  You see, it is possible to create your diagonal
edges (the ones that make the sides visible), so that three
diagonals join at three vertices to make an extra face inside
the box.  This usually won't make much difference, but
remember that it can happen--just in case.  The next chapter

                                                                        4-6

will tell you how to get Sculpt Animate to build a basic cube
for you.  It'll usually be best just to start with one of those
and alter it to suit your purpose.

    Now that you have a couple of more examples under
your belt, it's time to learn some more features.

    Another way to select vertices is with the Select Gadget.
It can be found near the top of the left border of a Tri-View
window, and its icon is three dots.  This gadget changes the
selection state of vertices that are inside the Tri-View--only
the ones visible in the Tri-View, no others.  We will make
use of this property in a moment.

    Here's how the gadget works: if one or more vertices in
the Tri-View are selected, then the gadget deselects the
vertices.  If no vertices in the Tri-View are selected, then the
gadget causes all of them to be selected.  You can use the
gadget to toggle all the visible vertices between selected and
deselected.

    A useful trick, when you wish to select only some
vertices of an object, is to move, expand, or contract the Tri-
View until only those vertices that you wish to change
remain inside.  The Select Gadget will then change only
these vertices.  This is an excellent way to select one side of
a cube, the end of a wing, the front or back of some letters,
etc.

    There are still more ways to change the selection state.
The menu command EDIT SELECT ALL and EDIT
DESELECT ALL can be used to select or deselect every
vertex in the scene, not just those in the Tri-View.  The
commands EDIT SELECT CONNECTED and EDIT
DESELECT CONNECTED operate on all vertices that are
connected to the indicated vertex.  Any other vertex that is
linked by an edge, either directly or indirectly, to the
indicated vertex is considered to be connected to it.

    As an example, if you were to reload the little house and
try out the SELECT and DESELECT CONNECTED
commands, you would find that the roof was connected

                                                                        4-7

together, but was not connected to the walls.  All four walls
are connected together.  This makes it easy to select the roof,
without changing the walls.  Later in this chapter we will see
how to literally raise the roof.

    Another useful command for selecting vertices is EDIT
SELECT SWAP.  As its name suggests, all selected vertices
become unselected and all unselected vertices become
selected.  This command works on all vertices, not just those
that lie inside the Tri-View.

    A single vertex can be selected or deselected with the
commands EDIT SELECT INDICATED VERTEX and
EDIT DESELECT INDICATED VERTEX.  In either case
you must indicate the vertex that you intend to change by
placing the cursor beside it.  Since you can do this even
more easily by double clicking on the vertex, these
commands are only really useful when you use the keyboard
equivalents of the commands.  Keyboard equivalents, where
available, are listed in the menus next to the item they
emulate.  Additional equivalents may be created 'on the fly'.
See chapter 14 for more details.

    The final selection commands that we will consider at
this time are EDIT SELECT INDICATED EDGE and EDIT
DESELECT INDICATED EDGE.  If you place the cursor
on an edge, these commands will select or deselect the two
vertices at the ends of the edge.  This can be useful when
dealing with very complicated objects when several vertices
are clustered together and you do no know which one is
connected to a particular edge.

    There are so many ways of selecting vertices (a few
more will be added in Chapters 6 and 10) because the key to
manipulating complicated objects is to be able to select just
those vertices that you need for a particular operation.

                                                                        4-8

Erasing objects

    Sculpt Animate is an editor for solid objects.  So far, we
have seen how to create vertices and edges, and--by the
rules--make faces.  What happens if a mistake is made?

    The EDIT ERASE portion of the menu provides a
number of options.  These commands remove items from the
current scene in RAM memory, but they have no effect on
scenes that have been saved, unless you save over an old
scene with an altered one.  Be a little cautious with the erase
commands, anything you erase is gone forever--unless you
have saved the scene beforehand.  It's a good idea to save
your intermediate work at intervals.

    The first selection in the ERASE submenu, EDIT
ERASE SELECTED VERTICES does what you might
expect.  Only selected (i.e., yellow) vertices are removed.
Since an edge is defined in terms of vertices, any edge that is
connected to a selected vertex is also deleted, and since a
face requires three edges, the removal of edges may result in
faces being erased also.  When you remove an object's
vertices, you remove its foundation.

    Note that EDIT ERASE SELECTED VERTICES works
on all selected vertices, so you should be careful that you do
not have any selected vertex outside the Tri-View, otherwise
it will vanish also.  It's a good idea to use EDIT DESELECT
ALL before selecting vertices to erase, or at least check the
status display (chapter 3) to make sure none are already
selected.

    The next erase command, EDIT ERASE SELECTED
EDGES, is very similar.  It operates on selected edges,
without removing any vertices.  A selected edge is one that
has selected vertices at both ends.  For example, if you had
made a cube with diagonal braces on all six sides, you could
open up one side by removing a brace.  Simply select the
vertices at either end of the edge to be removed.  The EDIT
ERASE SELECTED EDGES command will remove the
edge.  Again, if any edge of a face is removed, the face
disappears.

                                                                        4-9

    In addition to being able to erase selected items, you can
also erase them individually.  The command EDIT ERASE
INDICATED VERTEX and EDIT ERASE INDICATED
EDGE work on a single edge that is indicated by the cursor.
Again, if more than one edge is near the cursor, look in
another window to make sure the correct one is indicated.

    The next two commands can be used to remove lamps
from the scene.  EDIT ERASE INDICATED LAMPS will
erase a single lamp.  Indicate the lamp as you would
anything else, by placing the cursor within a pixel or two of
it in the currently selected window.  The command EDIT
ERASE ALL LAMPS is much less subtle, it just removes all
the lamps.

    The last erase command that we will consider at this time
is EDIT ERASE ALL.  It removes all vertices, edges and
faces as well as all lamps.  Because this could be
catastrophic--for example if you have just built an elaborate
scene and neglected to save any part of it to disk (which is
not wise in any case)--Sculpt Animate will ask if you really
want everything erased.  This gives you a chance to retract
the command.

Color

    By now, you must have become aware that the objects
you have made have been white, and that effects of
shading have only created different shades of gray.  Every
time Sculpt Animate creates a face, the face is given the
current default color, whatever that happens to be.  White is
the default color when you run Sculpt Animate.  If, at any
time, you want to see or change the current default color,
you need to execute the menu command EDIT MODIFY
FACES.

    The requester that pops up can be used for several
purposes, as we shall see.  This single requester lets you
change every available surface attribute of the faces of your
objects.  It is divided into three regions, one each for these
three attributes.  The regions are accordingly marked

                                                                        4-10

'SMOOTHING', 'TEXTURE', and 'FACE COLOR'.  We
will be concentrating on the 'FACE COLOR' region of this
requester for no.  The next two sections discuss the others.

        +------------------------------------------+
        |                               |          |
        |          FACE COLOR           | SMOOTHING|
        |   +------------------------+  |  +---+   |
        | R ||>                      |  |  |OFF|   |
        |   +------------------------+  |  +---+   |
        |   +------------------------+  | +-----+  |
        | G |                      |>|  | |FETCH|  |
        |   +------------------------+  | +-----+  |
        |   +------------------------+  |  +---+   |
        | B |                  |>    |  |  |SET|   |
        |   +------------------------+  |  +---+   |
        |   +------------------------+  |          |
        | H |          |>            |  +----------+
        |   +------------------------+  |          |
        |   +------------------------+  |  TEXTURE |
        | S |                      |>|  |          |
        |   +------------------------+  |  +----+  |
        |   +------------------------+  |  |DULL|  |
        | V |                      |>|  |  +----+  |
        |   +------------------------+  |  +-----+ |
        |  +-----+                      |  |FETCH| |
        |  |FETCH|  +-----+   +---+     |  +-----+ |
        |  +-----+  |     |   |SET|     |   +---+  |
        |  +-----+  |     |   +---+     |   |SET|  |
        |  |BLEND|  |     |             |   +---+  |
        |  +-----+  |     |             |          |
        |           +-----+             |          |
        +-------------------------------+----------+
        |    +------+                  +--+        |
        |    |CANCEL|                  |OK|        |
        |    +------+                  +--+        |
        +------------------------------------------+

                The Modify Faces Requester

    The current color is displayed in the box near the bottom
of the requestor.  The set of six slides can be used to control
the color--ignore everything else for now.  The first three
sliders control the amount of red, green and blue.  To
operate a slider, point at the triangular knob and hold down
the left mouse button.  If you move the mouse from side to
side, the knob will move to follow it, and the current color
will change.  With a little practice, you can mix the red,
green and blue to achieve almost any hue.

    If you find it difficult to get just the right shade with the
RGB sliders, try using the lower three that are labelled H, S
and V for hue, saturation and value.  It is a good technique
to start by moving the S slider and the V slider all the way to

                                                                        4-11

the right.  If you then move the H slider very slowly it will
change the color smoothly from red to green to blue and
back to red again, with many shades in between.

    Having established the correct hue, you can move the
saturation or S slider to the left in order to dilute the color
with white.  Finally you can use the V slider to make the
color darker or lighter.

    Once you have found the correct color, click the gadget
marked 'OK' at the bottom of the requester.  From now on,
any new faces that are generated will have that color.

    The color of a face can also be changed at any time.
First, you must select the vertices that surround each
triangular face whose color you wish to change.  Then
execute the command EDIT MODIFY FACES and use the
sliders to select the right color.  To set the color of the
selected faces, click the button marked SET that's in the
FACE COLOR region of the requestor.  The change will take
place when you click OK.

    You will have to exercise some care in selecting faces to
be re-colored.  Consider the case of a red triangular face in
the middle of a blue surface.  The red face's vertices are all
shared with blue faces.  If you select all the blue faces, you
will also select the red face!  This may not be what you
wanted.  If this sort of situation occurs, leave one blue face
adjacent to the red unselected, and re-color it later.  It is often
best to make differently colored faces separate, unconnected,
objects.

    The FACE COLOR region of the MODIFY FACES
requester can also be used to examine the color of one or
more faces.  Select the faces that you are interested in by
selecting the vertices that surround them, then invoke the
modify faces requester.

    If all the selected faces have the same color, then the
button marked FETCH will be solid and the button marked
BLEND will be shown in 'ghosted' text.  If the selected

                                                                        4-12

faces do not all have the same color, then the FETCH button
will be ghosted and BLEND will be solid.

    When the FETCH button is active, you can click on it to
make the displayed color the same as that of the selected
faces.  The BLEND button is similar, except that it will
display the average of the colors of the selected faces.  Be
warned that the average of several colors is not always what
you might expect, it is defined by averaging the red, green
and blue components separately.

    Again, the displayed color will become the current color
when you click OK.  If you click SET in the FACE COLOR
region and then click OK, the color of any selected faces is
also changed.

    If you call up MODIFY FACES and both the FETCH
and BLEND buttons are ghosted in the FACE COLOR
region, it means that no faces are selected.  For a face to be
selected, the vertices at all three corners of the triangle must
be selected.  It is easy to have many vertices selected without
having any faces selected.  One particularly confusing
situation can arise if two vertices have been created in the
same place.  This can cause something that looks like a
triangle to exist, and still not be a face.

Texture

    The second attribute that can be set for a face is its
texture.  It can either be dull, shiny, mirror, luminous, glass
or metal.  A face's texture has no useful effect in any of the
non-HAM rendering modes.

    The current default texture value is shown in a box
gadget in the modify faces requester, under the title
TEXTURE.  You can change the current default texture by
clicking on the box.  If you click repeatedly, the texture
values cycle through their sequence.  You can reverse the
direction of the cycle by holding down the ALT key while
clicking on the button.

                                                                        4-13

    A DULL texture scatters light uniformly in all directions
so no highlights will be shown, although the surface will be
shaded from bright to dark portions.  A SHINY finish
throws most of the light uniformly, but reflects some in a
particular direction, which can cause an object to glint, or
have 'specular' highlights.  A surface with a MIRROR
texture reflects all the light in a particular direction, so that
reflected images may be seen in the mirrored surface.  If you
also set a color other than white, the reflection will also be
colored.

    If you color a mirrored surface, reflected colors are
altered as though they had been through a color filter.  This
can have surprising results.  For instance, there is no red
light in a blue sky, so the sky will look black if it is reflected
in a red mirror.

    A LUMINOUS surface emits light regardless of what
light shines on it.  This can be useful if you are constructing
laser beams, bolts of lightning or lighted buttons on a control
panel.  Note that luminous surfaces cannot illuminate other
surfaces, but they shine into the camera so that they have the
same unshaded color in the resulting image that you set in
the COLOR FACES region of MODIFY FACES.

    A GLASS surface reflects about 20% of the incoming
light just like a mirror, while the rest passes through.
Because a face represents a surface and not a solid, Sculpt
Animate does not regard an object built entirely of glass as
being solid, so no refraction takes place.  Glass objects are
useful as canopies of aircraft, or as windows in buildings.
A METAL surface combines the properties of a shiny and a
mirrored surface.  Some light is reflected and some is
scattered, giving a metallic effect.

    The texture of selected faces can be changed to the
current default texture by the SET button in the TEXTURE
part of the modify faces requester.  If all the textures of
selected faces are the same, then the FETCH button will be
unghosted and you can use it to change the current default
texture to that of the selected faces.

                                                                        4-14

    MIRROR, GLASS, METAL and especially
LUMINOUS textures can create strange effects when used
in the SKETCH, PAINTING, and SCANLINE PAINTING
modes.  Avoid using these textures if you intend to use one
of these modes for your final image.

Smoothing

    Smoothing is a special attribute of faces which causes
one face to blend smoothly into the next, so that there is no
visible edge between them.  Smoothing will be especially
useful when you begin building more complicated objects in
the next chapter.  Just as it does with color and texture, a
newly created face inherits the current smoothness setting.

    The current smoothness value is indicated in the modify
faces requester, it is either ON or OFF.  If a face is
smoothed, then it will reflect light as though the surface was
smoothly curved, instead of faceted.  This attribute is only
effective when you make images in the imaging modes that
use HAM.  The particular method that is used to calculate the
curvature is called Phong smoothing.

    If you wish to change the smoothing parameter for a face
or faces that are selected, first change the current smoothing
value by clicking on it to toggle it between ON and OFF,
the click on the appropriate SET button.

    As with texture, if all the selected faces have the same
smoothness, you can use the FETCH button to change the
current default value to that of the selected faces.

    If you are constructing an object that is uniformly
curved, like an egg, all you have to do is to turn on
smoothing for all its faces.  If your object combines several
smooth portions that meet at an angle, like a crease, you will
need some way to inform the program of your wish, or it
will try to smooth out the crease.

    Because the smoothing algorithm works to smooth
adjacent connected faces, you can disconnect the two

                                                                        4-15

surfaces that meet to form the crease.  A small gap does not
even matter, because you can make it small enough that it
will not show in the final image.

    Suppose you were constructing a solid version of the
letter 'S'.  you would build the front flat portion as a single
connect shape, and the curved edge as a second object
with smoothing turned on.  If you need a rear surface, it
could be built as a third object.

    The only drawback of this method is that the SELECT
CONNECT command would have to be used three times
to select the entire letter.  If this is a problem, connect the
rear to the side with one single edge, then connect the side to
the front with another.  These edges will not create faces, but
they will connect the three parts together.

    Flat reflective surfaces can also confuse smoothing.  If a
complex flat surface has been created with a SHINY,
MIRROR, or METAL, texture, then Phong smoothing may
distort reflections in this surface.  Flat surfaces in these
textures may need to have smoothing set to OFF.
Smoothing works very well on curved surfaces in these
textures, though.

    Smoothing can be a great timesaver.  You will learn that
objects can be created with almost any degree of detail in
Sculpt Animate.  By making a rather coarse object and
giving it the attribute of smoothness, the rough edges will be
evened out, so that it will be almost indistinguishable from a
more finely built object.  This can be a great savings in
effort, memory use, and rendering time.

    Keep in mind, however, that Phong smoothing can only
smooth surfaces--it will not smooth outlines.  This is not so
bad, though, since a small amount of segmentation in an
object's outline is much less noticeable than the same amount
of segmentation on its surface.  Sometimes it's good to use
more vertices on an object's profile, leaving the rest fairly
coarse.  This puts the detail where it will do the most good.

                                                                        4-16

The Grabber

    Up to this point, we have assumed that the objects that
you have built are perfect, that you have never made a
mistake or changed your mind.  Except for erasing all or part
of an object, there has been no way to alter the objects you
have created.

    The Grabber is a low level tool that you can use to
change the position of one or more vertices.  Like most
Sculpt Animate facilities, the Grabber manipulates selected
vertices, and leaves unselected vertices alone.

    You invoke the Grabber tool by clicking on the Grabber
Gadget, which resides in the lower left corner of each Tri-
View window.  Its icon is supposed to look like the crane
operated gripping tool that is used in automobile wrecking
yards.  If you click on the Grabber icon with the left mouse
button, you will see its jaws close.  It is then active.  While
the Grabber is active, the normal crosshair cursor will be
replaced by a copy of the Grabber Gadget icon.

    The Grabber may be dropped (turned off) by clicking on
its gadget a second time,or by clicking the right mouse
button.  As an alternative to the gadget, the Grabber tool may
be activated by the menu command TOOLS GRABBER.

    While the Grabber is active, you may move the new
cursor anywhere in the scene, just like the normal cursor.
The difference is that any vertices that are selected will
follow along, maintaining their position relative to the
cursor.

    When you move a vertex, you also move or distort any
edges and faces that are built upon it.  For instance, if you
selected and grabbed just one vertex of the cube you built a
short while ago, you could move it without moving the rest
of the cube.  The edges and faces built upon the moved
vertex would stretch to accommodate its new position.  You
could also select and grab the entire cube, then move it
without any distortion.

                                                                        4-17

    Now give the grabber a try.  First clear the Tri-View
with an EDIT ERASE ALL command.  Next make three
vertices, and join them with a click on the Edge Builder
Gadget.  Deselect one of the vertices and turn on the
Grabber.  All the time that the Grabber is on, the selected
vertices will move whenever the cursor is moved.  The
edges will stretch to follow.  Notice that the cursor doesn't
have to be in any particular place when you invoke the
Grabber.  The selected vertices will always maintain the
same position relative to it when the Grabber is active.  In
this way large selections of vertices may be moved without
altering their structure.

    To try a more elaborate example, EDIT ERASE ALL to
clear the Tri-View.  Then PROJECT LOAD SCENE and
request littlehouse.  When it has loaded, deselect all the
vertices with the EDIT DESELECT ALL command and
move the cursor to one of the vertices at the top of the roof.
Then execute the EDIT SELECT CONNECTED command
and the whole roof structure will be selected.  Turn on the
Grabber and move the mouse pointer to the North window
(the one with the W,U,E and D at its edges).  Then move the
cursor up and raise the roof!

    The grabber can also be used to move a lamp.  To do
this, you must first DESELECT ALL if there are any
selected vertices, otherwise they would be grabbed.  Next,
indicate a lamp with the cursor.  When the grabber is turned
on, it can then be used to drag the lamp to a new location.
Chapter 7 also describes this.

    The Expand, Contract, Center, and Move Tri-View
gadgets all work correctly when the Grabber is active, so
you can move objects and lamps anywhere in the scene.
You must press the right mouse button to use menu
commands, so these are not available when the Grabber is
active.

    Avoid the Rotate (chapter 6) and Select/Deselect gadgets
when the Grabber is active.  Their behavior in this situation
is explicitly undefined, and they can have confusing effects.
At the time of this writing, the rotate gadgets will rotate the

                                                                        4-18

selected objects as though the cursor had not been moved.
The Select/Deselect gadget will only deselect vertices whose
original (pre-grab) position is visible in the Tri-View.
Deselected vertices will go back to their original position.

                                                                        4-19

                                CHAPTER 5

                               POWER  TOOLS

    In this chapter we will present some of the 'power tools'
that enable you to easily build elaborately shaped objects.
The objects you create using these tools are made up of the
same elements that were introduced in the last chapter--
vertices, edges, and faces--so you will be able to adjust and
manipulate these new structures in the same way, if
necessary.  In chapter 6 even more tools will be introduced
that can be used to transform the basic shapes created in this
chapter into almost anything.

Spheres

    Because Sculpt Animate builds everything from
triangular faces, it cannot represent a sphere exactly.  It so
happens, however, that a geodesic structure--constructed of
triangular elements--can represent a good approximation of a
sphere.  Though it would be possible to build such a
structure using the tools discussed so far, it would take
extreme patience and an unusual fondness for geometry.

    Fortunately, Sculpt Animate knows how to construct
spheres, and other objects, for you.  To have it build a
sphere, execute the menu command EDIT ADD SPHERE.
A requester will pop up and ask you how many times you
wish to subdivide the structure.  For now, simply press the
OK button and Sculpt Animate will construct a 20-faced
geometrical figure called an icosahedron.

    To take a look at this new object, set up a lamp and make
a picture of this object with the SNAPSHOT mode.  It is a
pretty poor approximation to a sphere, but if you modify all
its faces, setting the smoothing option, a more credible
image of a sphere will result.

                                                                        5-1

    You have seen (in the last chapter) how to apply the
smoothing attribute to an object, but we'll review the
procedure.  First, you must first select every vertex of the
object.  Since the icosahedron lies entirely within the Tri-
View, you can use the Select Gadget to do this.  Next
execute the EDIT MODIFY FACES menu command, ensure
that the current smoothing state is ON, and then click the
SET button.

    Now make a new image of the object.  Although the
outline still seems angular, the shading will make the
object's surface look quite smooth.  If the final size of the
object is to be fairly small in the picture, this may adequately
represent a sphere.

    To make a better approximation to a sphere, we may
subdivide each face.  The EDIT DO SUBDIVIDE menu
command will divide each selected edge in two, creating a
new vertex at the midpoint of each edge.  Then it links these
midpoints with new edges.  Each selected face will be
divided to make four.  Try this operation on the icosahedron,
being sure that each vertex is selected.

    The resulting object will have 80 faces, but will still look
like an icosahedron, since the new faces lie in the same
planes as the old ones.  What we need to do is to 'inflate' the
object by pushing the new vertices outwards in a small amount.

    There is a command EDIT SNAP CONNECTED
VERTICES TO SPHERE that does just what we want.
Unlike most commands, this command does not operate on
selected vertices.  Rather, it works on all vertices that are
connected to the indicated vertex.  Remember that if you
want to indicate a vertex, move the cursor to it in one
window.  The command figures out where the center of the
connected object is, and then moves all its vertices so they
are the same distance from that location.

    To see this happen, move the cursor to a vertex of the
subdivided icosahedron, then execute EDIT SNAP
CONNECTED VERTICES TO SPHERE.  The resulting

                                                                        5-2

object will look quite spherical, even in the Tri-View.  Make
a picture and see.

    You could repeat this sequence of SUBDIVIDE and
SNAP CONNECTED VERTICES TO SPHERE to build
better and better approximations to a sphere, but such
extravagant detail would be lost to the relatively low
resolution of the Amiga's screen.  Each subdivision will
quadruple the number of faces, and before long Sculpt
Animate will start to run very slowly because of the sheer
complexity of the object.  If you persist, you will run out of
memory.  A smoothed one- or two-subdivision sphere is
usually more than adequate.

    You might have already guessed that you can subdivide
the sphere automatically by specifying the number of
subdivisions when the requester EDIT ADD SPHERE
pops up.  Instead of just pressing the OK button, click on
the box containing the default subdivision number, erase the
default with the DEL or BACKSPACE keys, then type in the
desired value and hit RETURN or click OK.

Subdividing objects

    The preceding section described how the command EDIT
DO SUBDIVIDE may be used to increase the detail of a
sphere by dividing each selected edge into two and each
selected face into four.

    The command may be used on other objects as well, but
it's important to remember that, by itself, EDIT DO
SUBDIVIDE only adds vertices and edges--not detail.
Never subdivide an object unless you intend to move or alter
the added vertices in some way.

    For instance, if you wanted to add a nose to a potato,
there might not be enough vertices in the area of interest to
construct this detail.  You could select just this part of the
surface and subdivide it until there were enough vertices to
work with.  Of course, you could also construct the nose

                                                                        5-3

separately and simply move it into place--there are often a
number of ways to achieve a desired effect.

    There is one special function that the SUBDIVIDE
command will perform.  If only two vertices are selected,
and they are not joined by an edge, then the SUBDIVIDE
command works quite differently.  A requester pops up
asking you how many vertices you want inserted.  New
vertices are placed evenly between the two selected ones and
they are joined by edges that lie along a straight line.  No
faces are created, so this new object would not be visible in
an image.  This feature will prove helpful when you begin to
create splines and motion paths in chapters 8 and 10.

Other curved objects

    Besides spheres, the following curved objects can be
created with the appropriate EDIT ADD command:
hemisphere, disk, circle, cylinder, tube and cone.  In each
case you will have to enter a number to specify the degree of
detail that is required.

    In each case except circle and tube, a complete closed
object will be built.  For hemispheres you must specify a
number corresponding to a number of subdivision
operations, while for other commands, the number that
you give represents the number of rotation steps for a full
circle.

Cubes

    Angular shapes like a cube can be represented precisely
by Sculpt Animate, although each rectangular face must be
divided into triangles by a diagonal edge.  The EDIT ADD
CUBE command will construct a perfect cube in the Tri-
View--complete, with a diagonal division on each side.

    The cube that is constructed will be automatically sized to
fit comfortably within the current size of the Tri-View.  All

                                                                        5-4

of the EDIT ADD commands behave this way.  In the next
chapter we will introduce a mechanism called EXPAND that
can change the size of objects, as well as their dimensions.

    With this mechanism, a simple cube may be transformed
into any size of proportion of box.  For now, though, we
will use the Grabber Gadget to do a similar transformation.
Although they are somewhat more awkward than EXPAND,
the techniques discussed here can be used to create a number
of more subtle effects.

    Suppose we wanted to build a block that is twice as large
in one direction as it is in the other directions.  First, clear the
Tri-View with an EDIT ERASE ALL.  Turn off smoothing
in the EDIT MODIFY FACES requester.  Then make a basic
cube with EDIT ADD CUBE.

    Since the cube is sized to fit the Tri-View, we need to
expand the Tri-View to make room for our transformation.
Click on the Tri-View Expand Gadget twice.  Now we want
to select the four vertices that make up one end of the cube.
One way is to move the Tri-View so that part of the cube
sticks outside the Tri-View.  Then use the Select Gadget to
select the four vertices that remain.  When this is done, move
the Tri-View back so you can see the whole cube.

    Now look at the three windows and find one that will let
you move the cube's selected side away from the unselected
side.  The window whose Move Tri-View gadgets you were
using should do.  Finally, turn on the Grabber and--in the
window you just chose--pull out the end of the cube until it
reaches the desired length.

    If you do not have a steady hand, you may find that you
have pulled the cube out to a lozenge shape.  (Chapter 13
will tell you how to avoid this.)  Try to remember this shape-
-and any other shape you create by accident.  You never
know when you might find a use for it!

                                                                        5-5

Prisms

    The command EDIT ADD PRISM works just like the
command for cubes.  The prism will be oriented as though it
were the roof of a building, but in the next chapter you will
see how to tilt and rotate it into any position.

    Just as with the cube, you can easily alter this basic
shape to achieve any number of effects.  For instance, if you
want to create a steeply-pitched roof, you can select the two
upper vertices by double-clicking them, then use the Grabber
to raise the ridge.  You can also move the ridge side to side,
creating interesting asymmetrical effects.  Similarly, if you
want anything other than a square plan, turn off the Grabber
and select just one end of the roof, then grab and move that.

    Again, subsequent chapters will introduce even quicker
ways to alter basic shapes like the prism.

Drawing curves

    So far, the shapes we have constructed have had simple,
regular profiles; either circular, square, or triangular.  The
majority of objects in the real world are not so regular,
however.  Most are clearly distinguishable by a unique
profile.

    Sculpt Animate has several features which allow you to
take one or more profile curves and create solid objects from
them.  In Sculpt Animate, these curves are approximated by
a series of vertices joined by edges.  Although we could
make them using the methods of chapter 4, there is an easier
way.

    To see, execute command TOOLS CURVE.  Notice
that the form of the cursor has changed:  you have picked up
the curve tool.  This tool works very much like the regular
cursor.  You can click or hold down the left mouse button to
move the curve tool around.  If you press the right button
while the left button is held down, a new vertex is created--
just as before.

                                                                        5-6

    Unlike the regular cursor, though, subsequent vertices
will be joined to their predecessor with an edge.  This lets
you easily create a linked set of vertices.  As you move the
curve tool, the last vertex will be linked to the cursor by a
dotted blue 'rubber band line'.  This reminds you that you
are about to create an edge, and shows you the orientation of
the edge.

    While the curve tool is active, the gadgets for moving
and sizing the Tri-View may still be used to move around the
scene.

    You may release the curve tool at any time by clicking
the right mouse button.  If you position the tool very close to
an existing vertex, the curve will be closed by creating an
edge to the vertex, and the curve tool will be dropped.  In
this way you can create closed loops.

The Fill command

    A closed loop would be just what you needed if you had
to create a flat object with an arbitrary shape, like a letter on a
billboard.  If you tried to make an image of such an object,
though, it would be invisible, because although you would
have created vertices and edges, there would be no triangular
faces.  You could construct additional edges to fill the
interior of the loop, but Sculpt Animate provides a command
to do this for you--EDIT DO FILL.

    To try out this command, construct a closed loop in
some interesting shape and ensure that all the vertices are
selected.  Then execute EDIT DO FILL.  Vertices will be
constructed within the loop to fill its outline with faces.

    Erase the object and construct another closed loop.  Then
choose the curve tool again and make another loop entirely
inside the first.  If you use the EDIT DO FILL command,
this time the second loop will serve to define a hole in the
otherwise filled object.

                                                                        5-7

    Of course, if the inner loop is triangular, it will be a face
itself.  If you wish to create a triangular loop that is not a
face, just subdivide one edge of a triangle, then erase the
interior edge that will be created.

    The rules for using the EDIT DO FILL command are as
follows.  Only selected vertices are considered.  They must
all lie (or nearly in) a single plane.  The plane can be
oblique--that is, it does not have to be parallel to one of the
sides of the Tri-View.  The selected vertices and edges must
describe one or more closed loops.  Finally, no two selected
edges may cross.

The Edge Maker Tool

    If you have experimented with the Fill command, you
may have noticed two things.  For complicated shapes,the
command seems to execute slowly.  You might even be able
to do the job more quickly yourself.  You may also be
surprised at how the Fill command chooses to add new
edges.

    In general, rendering images will be faster if you avoid
long skinny triangles.  The fill command sometimes makes
such skinny triangular faces, even though you could
probably figure out how to fill the loop more intelligently.
The Fill operation is one of those things that humans are
good at, but that computer programs are apt to perform less
well.

    If you choose to do the Fill operation manually, there is a
tool that makes the job much easier.  To try it, first make a
closed loop--it doesn't matter if the vertices are selected or
not.  Next, call the command TOOLS EDGE MAKER.
Notice that the cursor changes to a shape that may remind
you of the Edge Builder Gadget.

    To use this new tool, move the pointer to a vertex, then
hold down the left mouse button and click the right button
once.  This starts the edge construction.  Move the cursor
and notice that a rubber band line goes from the vertex to the

                                                                        5-8

cursor.  Move the cursor to a second vertex and do the same
thing again--hold the left, click right.  An edge will be made
between the two vertices.

    You can now point to a new vertex and click the right
button to define the start of another edge, or click on the
right button without holding down the left button in order to
drop the tool.

Spinning new objects

    Many objects have symmetry about an axis.  Examples
include a clarinet, a doughnut, a spindle, a chess pawn, a
candlestick, and an hourglass.  All of these objects can be
described as a curve which has been swept about an axis.
The EDIT DO SPIN command is the mechanism that Sculpt
Animate provides for generating these shapes.

    To make the command work, you must first create a
curve that describes the outline of your object.  This is
simple enough if you use the tools that have already been
described.

    Once you have designed your pattern, make sure no
other vertices are selected in the scene (with EDIT
DESELECT ALL), then select the pattern curve with EDIT
SELECT CONNECTED.

    Next you must indicate the axis about which the pattern
is to be spun.  Do this by placing the cursor some distance
above or below the pattern curve in one of the Tri-View
windows.  It is assumed that the axis runs horizontally
across the screen, passing through the cursor in the selected
window.

    When you are satisfied with the curve and the location of
the axis, execute the EDIT DO SPIN menu command to
generate a new shape.  The command will display a
requester asking how many steps are to be taken as the curve
is swept around the axis.  You must use at least three.  The
more you choose, the smoother will be the resulting object.

                                                                        5-9

But more steps will also consume more memory, having a
similar effect to using many subdivisions on a sphere.  The
default value of 12 is usually adequate.

    After you have selected the number of steps, you will be
asked to provide a spin angle.  Normally you should accept
the default angle of 360 degrees, in order to produce a full
surface of revolution.  If you decide to use an angle less than
360 degrees, then a partial spin will be performed--just the
thing to model a slice of cake.

    Experiment with the spin tool to see how many different
shapes you can build.  Try using a semicircular pattern to
make a sphere.  How does it differ from a sphere made with
the EDIT ADD SPHERE command?

    Note that when you spin a curve less than 360 degrees,
you only construct the curved part of the object.  In the case
of a slice of cake, the flat parts that would be cut with a knife
are not filled.  you would have to select the outline of these
parts and use the Fill command to complete the object.

    You will probably want to use SMOOTHING on objects
created with the SPIN command, just as fore spheres.  This
allows you to specify fewer steps of revolution for better
efficiency, while still creating a realistically curved surface.

Helices

    A variation of the spin command is provided that lets you
build objects with helical or spiral symmetry, for example a
screw thread, a catherine wheel or a coil spring.

    As we've seen, the spin command lets you choose a
profile curve and spin it about an axis by a specified angle.
The command EDIT DO MAKE HELIX lets you do this
while at the same time letting the profile traverse along the
axis, move away from the axis, or both.

    Somehow you have to tell the program how you want
the profile to be moved as the rotation is performed.  This is

                                                                        5-10

done by means of a control curve.  The control curve is
made of a series of linked edges and vertices, in the same
way as a profile curve.  Each edge of the control curve
corresponds to one complete revolution, or twist, of the
helix.  The length of each edge of the control curve will
determine the length of the corresponding twist.

    This is much easier to demonstrate than to explain, so go
to your machine and clear the Tri-View with an EDIT
ERASE ALL.  Create a single vertex near the top left corner
of the 'down' window.  The new vertex is selected--leave it
that way, since it is to be the profile curve and so must be
selected.  Do not worry that the profile is a single vertex and
not really a curve at all, it will do fine for now.

    Now select the CURVE tool from the TOOLS menu and
go to the bottom of the same window.  Create a two-vertex
curve so that both vertices are on the bottom of the window,
spaced apart by about one quarter of the window's width.
These two vertices and their edge will be out control curve.

    To indicate that this is the control curve, place the cursor
on the vertex at one end of the edge.  Either vertex could be
used, but choose the leftmost one for now.  It does not
matter if the vertices that make up the control curve are
selected or not, they will only be used as a measure for
moving the profile.  Any other selected edges and vertices
will be used to construct the object.

    As with the SPIN command, the axis of revolution for
the helix is taken to be a horizontal line in the current
window, passing through the cursor.  Since the cursor is
busy indicating the first vertex of the control curve, you
must naturally construct your control curve so that its first
vertex lies on the desired axis.

    With the cursor positioned at one end of the control
curve, execute the menu command EDIT DO MAKE
HELIX.  You will be asked how many steps you wish to be
used for each turn.  For now, accept the default value of 12
by clicking on OK.  Next you will be asked the total number
of steps that you need.  Again, use the default value of 48 by

                                                                        5-11

clicking on OK.  This means that you will produce a four
turn helix, with 12 steps per turn.

    If you have done everything correctly you will see a few
inverted V-shaped objects in the down window, some
diagonal lines in the north window, and part of a circle in the
west window.  To see what has really happened you will
have to expand the Tri-View a few times.  Then all should
become clear.  The object that you have created is a simple
helix, the same shape as the wire springs you can find inside
retractable ball point pens.

    To get a better look at your creation, you can render an
image of it.  Set the rendering mode to OBSERVER WIRE
FRAME (you have not created any faces, so the other modes
will not reveal anything) and select a FULL image size.
Position the target in the middle of the helix, and move the
observer off to one side.  Use OBSERVER START to make
an image.  Try this again a few times, moving the observer
around until you are sure of the shape that you have created.

    Clear the Tri-View again and repeat the example, except
this time make the control curve a vertical rather than a
horizontal edge.  Remember to place the cursor on the end of
the control curve.  Instead of a helix, this time the command
EDIT DO MAKE HELIX will build a spiral curve.

    The control curve can be of any length, but it does not
make any sense to let it have more edges than the total
number of turns of the helix.  Clear the Tri-View and repeat
the first example, but this time use five vertices, creating a
four edge control curve.  Place each vertex of the control
curve on the bottom of the window, but space them out so
that each edge is longer than the previous one.  Construct a
helix and note that the helix becomes more stretched with
each turn.  See how the pitch of the helix matches the
spacing of vertices along the control curve.  Experiment with
control curves that are neither horizontal nor vertical and see
how the helix develops spiral tendencies.

    Now you are ready to build real helical objects.  Clear
the Tri-View and use the curve tool to draw a small freehand

                                                                        5-12

circle with about half a dozen vertices near the top left corner
of the down window.  Drop the curve tool, then call it up a
second time to create a horizontal control curve with two
vertices, like the first one you did.  Move the cursor to the
leftmost control vertex.  Ensure that the length of the control
is larger than the diameter of the freehand circle, and make
sure the circle is all selected.  Now select EDIT DO MAKE
HELIX again.

    You have created quite a complicated object, much like
the coil spring in the front suspension of some automobiles.
The helix is made up of faces this time, so you can place a
lamp and use some imaging mode besides WIRE FRAME to
look at it.

    Although you can build a multitude of open helical
objects with this command, there is one special way that it
can be used in order to create helical surfaces.  A helical
surface is one where one turn is connected to the next, with
no space between.  In order for this to work, there are two
additional rules that must be satisfied before executing the
helix command.

    The first rule is that there must be no control curve, so
the cursor must not be near any vertex.  The cursor is still
used to indicate the axis of rotation, though.

    The second rule is that the profile curve must be a simple
curve.  That is, it may have no branches, and it cannot close
up on itself in a loop.

    As an example clear the Tri-View and construct a small
V-shaped object at the top of the left of the down window,
comprising three vertices and two edges.  Move the cursor to
the bottom of the window and execute the command to make
a helix.  The role of the control curve is now played by the
profile curve, except that the overall size of the profile sets
the pitch and change in radius.  You will see that the V-shape
creates a surface like the threaded portion of a bolt.

    If one leg of the V were higher than the other, then the
threaded surface would taper.  As you might expect, if the

                                                                        5-13

previous example were repeated with a V-shaped object that
was drawn on its side, then a surface with a helical groove
would be created, like the shape of the groove in a
phonograph record.

Extruding shapes

    The last outline-transforming tool that we'll be
examining in this chapter is the extrude tool.  This tool lets
us use a curve pattern like a cookie cutter, to make solid
shapes from a two-dimensional profile.

    Once again the CURVE tool is used to draw one or more
curves or closed loops with selected vertices.  When you are
satisfied that you have created the cross section of the shape
that you want to extrude, execute the TOOLS EXTRUDE
command.  There will be a short pause, during which the
grinding gear image will replace your mouse pointer.

    When your normal pointer returns, notice that the
Grabber tool has been turned on.  Now select a Tri-View
window where the curves can be viewed edge-on.  If you
move the cursor in this window, you will see the original
patterns divide into two.  An unselected (purple) copy of the
original pattern remains fixed, while the selected (yellow)
copy can be dragged with the cursor.  New edges appear,
joining the two outlines.  These edges have been created by
the extrude tool.  When you are satisfied with the extruded
shape's form, click the Grabber icon to turn it off.

    If you had traced out the outline of a building in plan
form you could use the extrude tool to raise the walls with
one stroke of the mouse.  It can also be used to build three
dimensional alphabetic characters, any number of stamped or
extruded forms, and many other objects.

    One useful property of the extrude tool is that if you use
it on a curve that has been filled, the interior edges do not
extrude to form interior partitions.  This prevents useless
faces from being created within extruded objects, where they
would waste memory and processor time.

                                                                        5-14

                                CHAPTER 6

                                MORE TOOLS

    In this chapter we will look at more ways that we can
manipulate and build objects.

Saving and loading objects

    If you build an object, then you can save it to disk with
with the PROJECT SAVE SCENE command.  A file
requester will pop up to enable you to specify the name that
you wish to use for the file.  This requester was described in
detail in chapter 2.  It lets you move around among the
volumes and drawers in your Amiga's file system.  When
you have found the right drawer, you may either select a pre-
existing filename or type in a new one in the provided box.
If you pick a name for saving that has already been used, the
old file will be destroyed.  You will be warned before this
happens, though, and allowed to cancel the save if it is not
what you want.

    A scene is made up of four parts: lamps, the observer,
the world, and objects.  When you have finished picking a
filename to save the scene under, another requester will
appear, listing these parts.  Next to each part's entry is a
'toggle gadget' which may be switched from YES to NO and
back just by clicking within its border.  Items marked 'YES'
will be saved, items marked 'NO' will not.

    'Lamps' is the top item in the list.  If simply specifies
whether lamps will be saved.  You amy not want to save the
lamps in a scene if you are simply saving its objects to be
used in another scene.  On the other hand, you might want have a
particularly pleasing arrangement of lamps that you wish to
save for use in other scenes.

    'Observer' is the next item.  If its gadget is set to NO,
the observer's location and target will not be saved.  Neither

                                                                        6-1

will any of the parameters associated with the observer
menu, such as rendering mode and image size and type.
These things are saved with a scene in the Observer part.

    The 'World' is described in more detail in the next
chapter.  It includes details about the sky, ground, and
background illumination for a scene.

    'Objects' is the final and most important entry.  The
objects in a scene are its vertices, edges and faces.  Color,
texture and smoothness are also stored as part of the objects.
If this item is set to 'NO', no objects will be saved at all.  If
it is set to 'YES' they will be saved according to the gadget
marked 'Which?' at the bottom of this requester.

    The 'Which?' gadget is usually set to 'ALL', which
means that all objects will be saved if 'Objects' is set to
YES.  If the gadget is set to 'SELECTED', however, only
those objects in the scene that are selected (with yellow
vertices will be saved.

    By carefully selecting which parts of a scene to save or
load, you can combine the best parts of different scenes--
lamps from one, objects from another, etc.  Be careful,
though!  If you leave some of these items set to 'NO', then
the next time you save a scene you might not save all you
wanted to.

    PROJECT LOAD SCENE behaves very much like
PROJECT SAVE SCENE, except, of course, that it loads
the scene data back into the computer.  LOAD SCENE also
allows you to choose which parts of the scene to load.  The
selection states of objects are not stored, of course, so all
objects that were saved in scene will be loaded in, but
you may choose whether the objects that are loaded will be
selected or not.

    The LOAD SCENE command does NOT erase any
objects that are already present in the Sculpt universe.  This
allows you to load in different parts from different scenes, as
we have seen.  If you want to clear the slate and start over

                                                                        6-2

with a scene you've loaded, do an EDIT ERASE ALL before
re-loading the scene.

    If you only need to save or load objects, then you may
use the commands PROJECT SAVE OBJECT and
PROJECT LOAD OBJECT.  These operate a little differently
from the commands that work with complete scenes.  You
are not given the choice of saving or loading lamps, the
observer, or the world, and only selected objects are saved.

    There's a more subtle difference, as well.  LOAD
SCENE will always load objects into the same 'absolute'
position in the Sculpt universe where they were when SAVE
SCENE stored them.  On the other hand, SAVE OBJECT
records the position of the saved vertices relative to the
cursor.  When the objects are loaded in by LOAD OBJECT,
they will be placed the same way relative to the cursor's
location, no matter where it is in the scene.

    Suppose that you had constructed a model of a grand
piano and all the vertices were selected.  Then you could
position the cursor under one leg and save the object to a
disk file.  If at another time, you wished to introduce a grand
piano into the Tri-View, you could move the cursor to the
position where you wanted the leg of the piano to be and
execute the PROJECT LOAD OBJECT command.

    If you need numerous copies of the same object, you can
also use the command PROJECT LOAD OBJECT to add
multiple copies.

Rotating objects

    We may not always want our objects to be lined up with
the axes of the Tri-View.  If we are building a model of a
city, for instance, and we want to use the EDIT ADD CUBE
command to produce buildings, we have yet to see an easy
way to produce a curve that faces to the southeast, or any
other 'off axis' direction.

                                                                        6-3

    The Rotate Gadgets are just such a way.  They may be
used to rotate any selected object.  To see how they work,
do an ERASE ALL and make a cube.  Position the cursor at
the center of the cube and select all eight of the vertices (the
Tri-View's Select gadget should do nicely for this).  If you
now click the left mouse button on one of the Rotate Gadgets
at the top left border of the Tri-View window, you will see
the cube rotate 5 degrees for each click.  If you hold down
the right Amiga key as you click, each rotation step is
increased to 45 degrees.  Similarly if the left Amiga key is
held down, the rotation step is only one degree.

    The left and right ALT keys can be used to modify the
rotation gadget even more.  If the gadget is used with the left
ALT key, the rotation is only 0.1 degrees.  This is too small
to see, but, as we will see in Chapter 13, there are ways of
measuring angles with great precision.  The right ALT key
increases the rotation angle to 180 degrees, so this is a quick
way to turn an object about-face.

    If you move to another window, you will be able to
revolve the cube around a different axis.

    To see a different use for the rotate gadgets, expand the
Tri-View to about ten times its original size and position the
cursor some distance from the diminutive cube.  Click the
Rotate Gadget and you will see the little cube move in a
circular orbit around the cursor.

    In summary, the Rotate Gadget rotates all the selected
vertices around the cursor.  Be careful that you do not have
any vertices accidentally selected outside the Tri-View,
because they will be rotated also.

    For now, you will probably want to select all the
connect vertices of an object to be rotated.  Otherwise only
part of the object will be rotated and a tangled mess can
result.  As you begin to create more detailed objects, though,
you may find it convenient to select portions of an object's
surface and alter its shape by clever manipulation of the
cursor and rotate gadgets.

                                                                        6-4

Expansion and contraction

    Each of the EDIT ADD sequence of commands creates
an object just a little smaller than the size of the Tri-View.
This wouldn't do if we wanted an object of a certain size,
since we can't specify the size of the Tri-View precisely.
Besides, if we found that we had created an object of the
wrong size, our only recourse would be to erase it and try
again, or start fiddling with the Grabber.  That would
hardly be practical for a pseudo-sphere with 320 faces.
Fortunately there is a command that enables you to change
the size of an object with great precision.

        +-----------------------------------------------+
        |                                               |
        |===============================================|
        |                                  +--------+   |
        |     CONTRACT  or  EXPAND  about: | CENTER |   |
        |                                  +--------+   |
        | +---+ +------------------------------+  +---+ |
        | |1/2| |            |X|               |  | 2 | |
        | +---+ +------------------------------+  +---+ |
        |                                               |
        |       +------------------------------+        |
        |       |             ||               |        |
        |       +------------------------------+        |
        |      +---+             +---+           +---+  |
        |  E/W:|YES|         N/S:|YES|       U/D:|YES|  |
        |      +---+             +---+           +---+  |
        |                                               |
        |   +--------+                       +--------+ |
        |   | CANCEL |                       |   OK   | |
        |   +--------+                       +--------+ |
        +===============================================+

                        The Expand Requester

    Let's learn by doing.  First, make a one-subdivision
sphere and select all its vertices, then execute an EDIT DO
EXPAND command.  An expansion control requester will
pop up.  Unlike some other requesters, this one has a drag
bar along the top.  Move the requester so that you have a
good view of each of the Tri-View windows.

    The EXPAND requester has two slider gadgets across its
width.  Moving either slider to the left or right will move the
selected vertices towards or away from the chosen expansion
point.  The vertices move proportionally, more or less so
according to their initial distance, causing a smooth change

                                                                        6-5

in the size of the selected object.  The upper of the two
sliders provides the coarser adjustment.  The lower slider
moves the vertices very slightly.  Chapter 13 introduces a
measuring device that can be used to monitor the size of an
object precisely as you change it.  This makes the fine
adjustment more useful.

    There are two other sizing gadgets, one at either end of
the sliders.  One is marked '1/2', the other is marked '2'.
Clicking these gadgets will make the selected object either
half- or double-sized.  If you make an object much larger, it
will move outside the Tri-View, so give yourself enough
room to work before you call up the EXPAND requester.

    At the top of the requester is a gadget that starts out
labeled CENTER.  In this state, the gadget indicates that the
center of expansion is at the center of the object (as defined
on the next page).  If you click the gadget, it will switch to
CURSOR, and then to CENTROID.  This lets you choose
what part of the object will remain fixed as the rest expands.

    You can use the gadgets near the bottom of the requester
to limit the expansion to one or two dimensions.  Click on
the boxes and see them toggle between YES and NO.  Turn
off two of the directions and then try contracting the sphere.
You will see it flatten out like a pancake.

    This is an important principle.  EXPAND can be made to
work in any combination of the three dimensions.  For
instance, a simple cube can not only be lengthened, it can be
resized in any dimension to make any kind of rectangular
block.  The compass marks on the Tri-View windows can be
helpful in choosing which dimensions you want to resize.

    Even more variety can be achieved by selecting only
portions of an object to be resized.  By selecting the top four
vertices of a cube and shrinking them, you can create a
trapezoidal pedestal (a pyramid with its top cut off).  You
can select individual vertices of a sphere and resize them,
creating a number of spikes or dimples on the sphere's
surface.  Starting with a few simple objects and some
imagination, you can quickly create any number of shapes.

                                                                        6-6

    Note that expanding relative to the cursor gives you
some special options.  Placing the cursor away from the its
center will make the object move, as well as change the shape.
This makes it possible to fit an object into a given space.
The CENTROID setting is similar in its effect to CENTER.
It is provided primarily for completeness, but it may come in
handy for some applications.

Finding the center of an object

    There are times when it is desirable to have a quick and
precise way to find an object's center.  The command EDIT
SNAP CURSOR TO CENTER can be used to do jus that.
It moves the cursor to the center of the selected vertices.  For
our purposes, the center of a group of vertices is considered
to be the center of a rectangular box that is just big enough
to hold all the vertices.  The faces of this imaginary box are
parallel to the Tri-View sides.  This may sound a bit more
complicated than necessary, but there are many ways to
define the 'center' of an object.  This definition will usually
give us the location we want.

    The command EDIT SNAP CURSOR TO CENTROID
is sometimes useful as an alternative way to find the middle
of a group of selected vertices.  The centroid is the 'center of
gravity' of the vertices, assuming that each vertex has the
same weight.  The centroid will often differ from the center,
as defined above, but not always.  For example if most of
the vertices lie on the left side of an object, then the centroid
will be offset from the center in that direction.  For some
shapes, like a subdivided sphere, the center and the centroid
are the same, while for others, like the hemisphere, they are
quite different.

Manipulating shapes

    If you were to EXPAND a sphere along one dimensional
axis, you might expect that it would lengthen out into an
ellipsoid (oval solid).  This is indeed the case.  The Grabber

                                                                        6-7

Tool can be used to perform an EXPAND function of
sorts, by selectively lengthening middle sections of an
object.  This distinction will come in handy over time, as
you build more diverse shapes.

    Suppose that you wish to create a long cylinder with
spherical ends.  First construct a sphere with the EDIT ADD
SPHERE command.  Use a subdivision value of one or two.
Then move the Tri-View so that exactly half of the sphere is
sticking out of the Tri-View.  Now is you use the Select
Gadget, you will find that you have selected all the vertices
on one hemisphere.  Click the Expand Tri-View gadget to
make room for what will follow.  Turn on the Grabber and
you will find that you can pull the two haves of the sphere
to make the cylinder with rounded ends.

    With tools such as EXPAND and the Grabber, you can
build a great variety of objects just by altering pre-existing
ones.

The magnet tools

    If you need to pull or push several vertices, then the
Grabber tool will do that for you.  However it moves them
all by the same amount.  Sometimes you may want to move
some further than others.  This is what the magnet tools do.

    Let's try an example.  Erase all objects and create a few
vertices.  Ensure that they are selected.  Now execute the
menu command TOOLS MAGNET ATTRACT and you will
see the cursor turn into an icon for a horse-shoe magnet.  As
usual, if you hold down the left mouse button you can drag
the tool around.  If, while you are holding down the left button, you
click the right mouse button, then all the selected vertices
will move a small distance towards the magnet.  If a vertex is
far from the magnet, the vertex will only move a small
distance.  Just like a real magnet, the attraction gets weaker
at increasing distances.  When a vertex gets very close to the
magnet, the attraction gets weaker again in order to prevent
the vertex from overshooting the magnet.  Try it and see
how the vertices move.

                                                                        6-8

    The strength of the magnet can be varied.  If the left
Amiga key is held down (as well as the left mouse button),
when the right button is clicked, the strength is reduced.  If
the right Amiga key is held down, the strength is increased.
If you select TOOLS MAGNET REPEL, everything works
as before, except that the selected vertices move away from
the magnet.  The magnet tool is released by clicking the right
mouse button without holding down the left button.

    Erase everything and make three vertices arranged in a
large equal sided triangle.  Join the vertices with edges.
Ensure that the three vertices are selected and execute the
EDIT DO SUBDIVIDE command a few times until you have
made a triangular grid.  Position the cursor at the center of
the grid, and then move to another Tri-View window.  Move
the cursor away from the grid and select the MAGNET
ATTRACT tool.  Operate the magnet a few times and see
what happens.  The net of points will deform and a spike
will be pulled out towards the magnet.

    If you repeat the experiment with a new grid, but use the
magnet to repel vertices, the grid is deformed, but the
bump is rounded rather than spiked.  By moving the magnet
between pulses and by switching from attraction to
repulsion, any number of unusual shapes can be built.

    If you operate the magnet in a Tri-View that is large, the
strength of the magnet is increased, so that its effect appears
to be the same.  If the Tri-View is contracted, the magnet gets
weaker and it can be used to make very delicate
modifications to a small portion of an object.

    There will be a tendency for a selected object to grow,
overall, as it is deformed by MAGNET REPEL, and shrink
as it is shaped by MAGNET ATTRACT.  Don't worry too
much about this as you shape a set of vertices, because you
can always use EDIT DO EXPAND to resize the resulting
shape to fit your scene.