Introduction to 3D Modelling and Rendering, Part 1 â

An introduction to 3D for those new to it.

What is 3D?

For those of us used to working in Photoshop and Illustrator it is
important to realise that all that work is 2D, or two-dimensional.
Photographs of real objects or painting them from scratch in Painter,
they are still 2D. This is because we are either working with a pixel
representation or flat objects, like lines, text, paths, etc. This is
true even if we are attempting to simulate a 3D look.

In 3D work, or three dimensions, we are producing a description of real
objects with depth, scenes comprising many objects and the spatial
relationships between them, along with the required lighting
arrangements and viewing characteristics. The end result of 3D work is
still usually 2D. This is either a still image or an animation, but
it’s still made up of pixels. In an ideal world our output would be
three-dimensional too, as in a holographic projection or even a
sculpture. This is a limitation of the output technologies that we have
to work with at present, rather than an inherent characteristic of 3D
work. Since 3D printers exist (they are actually more like a
numerically controlled milling machine in some ways), as do using LCD
shutter glasses for direct 3D display, working completely in 3D is
possible, just not the normal use.

Deep down, usually buried deep inside the software, our 3D work
consists of rather mathematical descriptions of our scenes, such as
place a sphere of radius k, with it’s centre at x,y,z point in space
with a surface texture like stone. Thankfully, we rarely have to deal
with the numerical level unless we choose to. There are good reasons to
dive down to the numerical level at times, such as exact placement. 3D
software is largely click and drag operation these days for most common
operations. It is important to remember that we are trying to represent
things in the three-dimensional world that we are used to living in.
Just as navigating around the real world can get you lost, so is it
easy to become disoriented in 3D software.



Keeping oriented in 3D

In 3D software the convention is to use a set of three coordinates, x,
y and z. Co-ordinates can be absolute or relative. Absolute coordinates
apply to the entire world that we are creating in the computer.
Everything is specified relative to a universal origin, the centre of
your digital universe, with coordinates of 0,0,0. Positive x values may
lie to the right, negative ones to the left. Positive y values may be
up and negative ones down from the origin. Positive z may be in front
of and negative ones behind the origin. Absolute coordinates are used
to position objects in our scene, to place cameras and lights, etc.
Relative coordinates have their origin somewhere other than the world
origin. For instance, in creating an object made up of many parts it
may be more convenient to think in terms of positions relative to what
you wish to consider the centre of the object.

How the software works can have an impact on how easy it is to keep
oriented. Some programs, like Bryce, display only one window, so you
only have one view of your objects/scene at a time. Other programs,
like Vue d’Esprit or Lightwave, by default give you four views: a
front, left and top view plus the view through the main camera. This
last solution is generally preferred but does tend to work best when
you are using a large, high-resolution screen. This is why most of the
consumer level programs use the one view approach, assuming home users
have small screens, whilst professional software takes the four-view
approach.


The stages of 3D work

The following are the main stages of creating a 3D work:

1.    Create objects;

2.    Place objects in relation to each other in scene;

3.    Place light sources;

4.    Place the camera or observer;

5.    Add textures to objects;

6.    Add atmospheric effects;

7.    Render to produce a final image or animation movie.

The exact order of this sequence is partly up to you and partly a
function of the software that you are using. For instance, some
software separates the creation of objects and their placing in the
scene (as in Lightwave), others combine this into one step (as in
Bryce). Likewise, sometimes the textures are placed on objects when you
create them. But they can also be added at the scene creation stage.
Each person gradually finds their own order of working that suits their
needs and the needs of the specific project. For projects involving
many people there may be different order, or indeed some stages my be
performed in parallel, than for projects where you are doing the whole
thing. The order of steps can affect the performance of your software.
The sequence given tends to produce the least delays with most
software, for reasons that will become clear as we progress through
this series.

Creating objects and placing them in the scene is often called
‘modelling’. This is because in creating an object and then a scene we
are building a ‘model’ of it in the computer. Some software even
separates the modelling function from the rest of the software by
splitting the process into two programs. It is quite possible to do the
modelling in on manufacturer’s program and the rest of the process in
another. I quite frequently use three different programs for this
process, making use of the strengths of each, these being Poser and
Byrce and Lightwave.

Light sources and a camera are necessary if you are to see anything of
the wonderful model you have created. Light sources and cameras can be
treated in much the same way as any other object. Light sources will
have their own, special characteristics though, like the type of light
source, whether it casts shadows, its colour, etc. The camera also has
special characteristics, like its field of view, resolution of the
resulting image(s), etc.

Rendering is the process of determining what the scene looks like from
the camera position taking into account all the characteristics of the
objects, light sources and their interaction. Rendering is usually a
time consuming process for any scene of reasonable complexity. This can
vary from a ‘go get a cup of coffee’ to ‘lunch’ up to a whole week, or
more. This is one reason why high complexity rendering of still images
or animations tends to require fast computers and lots of memory. One
reason that the order with which you create your image(s) is important
is that you will usually do lots of little test renders along the way.
Thus you want to leave the details which really slow the rendering down
to as late in the sequence as possible.



Why would we want to use 3D?

We need to represent solid objects, whether in a still image for an ad
or an animation to go in a movie. Since real world objects are 3D,
there will be times when a 3D representation is needed. Sure, we can
paint or airbrush a 3D approximation but it will have a particular
look, assuming that we have the skill level to create it.
Working with
3D software creates a different look. This can vary from one with a
very computer feel to a photorealistic one, depending on the software
and what we do with it. The major advantage of working with 3D software
is that it is easy to produce changes. To change the viewpoint only
requires that we move the
camera and render. To change the lighting or
reposition objects is equally easy. So having created a scene once, we
can produce many different images from it. This is like photographing a
real scene in everything from wide-angle to close-up, and from
different positions. 3D software gives you flexibility. This very
flexibility allows you to re-purpose images. You may do an illustration
for a magazine ad and then the client comes back and wants an animation
for a TV ad, or the web. Once you have built the models, you can re-use
them repeatedly.


This screen grab of the old Metacreation’s Infiniti-D 4.5
shows a four window, working environment. Three windows give front, top
and side views whilst the fourth shows the camera view. This type of
display, common to most of the higher-end 3D packages, works best on a
high resolution, large screen.


The single view at a time display, like this one from Bryce,
works well on smaller displays. Usually keyboard shortcuts or buttons
allow you to switch between views. Whilst not as convenient as the
four-window display it is quite workable. It seems natural once you get
used to it.


This simple cartoon bird was created out of basic
object types and rendered in Infiniti-D 4.5. A background image was
used.

Tags:

No comments yet.

Leave a Reply