|
Understanding Lighting in the Java 3D API
By Richard G. Baldwin
Java Programming Notes # 1540
Preface
This is the first part of a multi-part lesson designed to help you understand the use of lighting and illumination in 3D scenes produced using the Java 3D
API.
Optical illusions
Creating scenes that appear to be three-dimensional on a flat two-dimensional screen is all
about creating optical illusions. There are many important aspects to
creating these optical illusions in 3D
programming, not the least of which is lighting and illumination.
A 3D simulator program
In this lesson, I will teach you about, and show examples of many of the important
features of scene
illumination in the Java 3D API. I will also provide the source code for a complete Java 3D lighting simulator program that you can compile and run to experiment
with light in the Java 3D API. I will explain how that program works in
subsequent parts of this multi-part lesson.
Viewing tip
You may find it useful to open another copy of this lesson in a
separate browser window. That will make it easier for you to
scroll back
and forth among the different figures, tables, and listings while you are
reading
about them.
Supplementary material
I recommend that you also study the other lessons in my extensive
collection of online Java tutorials. You will find those lessons
published
at Gamelan.com.
However, as of the date of this writing, Gamelan doesn't maintain a
consolidated index of my Java tutorial lessons, and sometimes they are
difficult to locate there. You will find a consolidated index at www.DickBaldwin.com.
The program that I will present in this lesson can be used to experiment with most
of the lighting and illumination features of the Java 3D API either individually
or in combination.
In order to compile and run this program, you will need to download and install the
Java 3D API. As of the date of this writing, it was available at:
http://java.sun.com/products/java-media/3D/. The online documentation
was available at:
http://download.java.net/media/java3d/javadoc/1.4.0/.
In addition, you will need to download and install either
Microsoft
DirectX or OpenGL.
Our perception of 3D visual objects
From time to time, I will make references to Chapter 6 of a document
published by Dennis J. Bouvier at the following URL.
http://java.sun.com/developer/onlineTraining/java3d/.
For example, here is a quotation from Bouvier:
"In the real world, the colors we perceive are a combination of the
physical properties of the object, the characteristics of the light sources,
the objects' relative positions to light sources, and the angle from which
the object is viewed. Java 3D uses a lighting model to approximate the
physics of the real world."
The mathematical 3D model
According to Bouvier, Section E.2 of the
Java 3D API Specification presents the mathematical equations of the Java 3D
lighting model. Still quoting Bouvier,
"The lighting model equation depends on three vectors: the surface
normal (N), the light direction (L), and the direction to the viewer's eye
(E) in addition to the Material properties of the object and the light
characteristics."
This program makes it easy for the user to vary the surface normal
(indirectly), the light direction, and the Material properties of the
object. The program does not make it easy for the user to modify the
direction to the viewer's eye. However, the primary experimental object is
a large sphere which, because of its curvature, inherently results in a wide
range of directions from the object's surface to the user's eye.
Light sources
Returning to Bouvier,
"The lighting model incorporates three kinds of real world lighting
reflections: ambient, diffuse, and specular. Ambient reflection
results from ambient light, constant low level light, in a scene.
Diffuse reflection is the normal reflection of a light source from a visual
object. Specular reflections are the highlight reflections of a light
source from an object, which occur in certain situations."
This program makes it easy for the user to experiment with all three types of
sources, plus another form of illumination referred to as emissive color.
Two user interfaces
This program produces two user interfaces. One interface is the display
of a 3D scene containing five spheres as shown in
Figure 1.
(Note that the importation of the image into this HTML document introduced
some visual artifacts that were not present in the original image produced
by the program.)
The four small spheres
Four of the spheres are small with fixed reflective surface properties and a
fixed number of facets. When illuminated with white light, or colored
light corresponding to the color of the sphere, these spheres appear
as white, red, green, and blue as shown in Figure 1.
The larger fifth sphere
The fifth sphere is larger. Almost everything
about this sphere (other than its size and location) can be modified by the user
through the second user interface shown in Figure 2.
The slider section
The user interface in Figure 2 consists of three main sections. The
section at the top contains three sliders that are used to adjust the values of
selected scene properties. The values indicated by the sliders are tied
directly to the display shown in Figure 1, and the display changes in real time
as each slider is adjusted to indicate different values.
A grid of radio buttons and text fields
The large middle portion of the interface contains a grid of radio buttons
and text fields. The radio buttons are used to select the scene properties
that will be modified by adjusting the sliders at the top of the user interface.
When a radio button is selected, the label above each slider changes to
appropriately reflect the value of the scene property currently assigned to the
slider. In addition, the values displayed in the text fields to the right
of the selected radio button display the values indicated by the sliders, and
the column headers for the text fields change to appropriately reflect the
values of the scene properties currently associated with the selected sliders.
Shading and size
The four radio buttons at the bottom of the user interface are used to select
between:
- Gouraud and flat shading
- Large display and small display
Figure 1 was produced as a large display using Gouraud shading. You
will see examples of both types of shading as well as large and small displays
later in this lesson.
More than 50 scene property values
The user can vary the surface reflective properties of the large sphere and
can also modify various properties of the lights shown in Figure 1.
The user interface in Figure 2 provides more than 50 scene property values that can be
modified by the user to change the reflective properties of the large sphere and the
various properties of the lights. The ability for the user to modify the
reflective surface properties of the large sphere makes it possible for the user to view
the effects of those modifications in conjunction with varying light
conditions.
Surface properties of the large sphere
The user can control almost all aspects of the following
surface properties of the large sphere:
- The number of facets on the surface of the sphere.
- The type of shading:
Gouraud shading versus
Flat shading.
- The shininess of the sphere.
- The emissive color and intensity of the sphere.
- The ambient
color and intensity of the sphere.
- The diffuse color and intensity of the sphere.
- The specular color and intensity of the sphere.
Five independent light sources
In addition, the
user has control over the following five lights along with the properties shown:
- One AmbientLight object (color and intensity).
- Two independent DirectionalLight objects (color, intensity,
and direction).
- One PointLight object (color, intensity, location, and
attenuation).
- One SpotLight object (color, intensity, direction, location,
attenuation, spreading angle, and concentration).
An almost infinite number of combinations
Given the many properties of the five light sources and the many reflective
properties of the spheres, the lights can be combined with the spheres to
illuminate the scene in an almost infinite number of ways.
For example, having specified the various
surface properties of the large sphere, the user can vary the properties of the lights
and observe the manner in which the lights interact with the five spheres to
illuminate the scene.
Real-time control
The properties of the large sphere and the lights are
varied by selecting a radio button (see Figure 2) that corresponds to the property
of interest and then
adjusting the three sliders at the top of Figure 2. The display of the scene changes
in real time as the user adjusts the
sliders, making it possible to easily view the effects of fixing all of the properties
but one and then varying that property across a range of values by moving a
slider.
Large or small display
The user
interface shown in Figure 2 also makes it possible for the user to select between a large and a
small display. This was done mainly to support my publishing effort as I
import the images into this tutorial lesson. Figure
1 shows the large display. You will see examples of the small display
later.
Testing
This program was tested using J2SE 5.0, the
Java 3D API version 1.4.0, WinXP and
Microsoft
DirectX. Note that it is also possible to use
OpenGL instead of DirectX, but I haven't
tried it.
Normally we would like to think that the silhouette of a sphere is a perfect
circle. In the ideal case, that would be true. However, 3D computer
graphics do not often reflect the ideal case.
How to draw a circle
When you were in elementary school, you probably learned how to use a compass
to draw a circle. If you did it correctly, the circles were nice and
smooth. Unfortunately, however, there are no compasses inside computers. Typically
when we draw a circle with a computer, we
compute the values that represent the circle at a set of regularly spaced points around the
circumference of the circle and then connect those points with straight line
segments.
The closer the points are spaced, the better will be the representation of the
circle.
How to draw a sphere
Similarly, when we draw a sphere with a computer, we
typically compute the values representing the surface of the sphere at a set of regularly spaced points on the surface
of the sphere and then connect those points with straight line segments. Those lines
form the edges of flat polygons and it is this structure built from flat polygons
that constitutes our representation of the sphere. Just as in the case of the
circle, the closer together the points are on the surface of the sphere, the better
will be our representation of the sphere.
The result is that our computer generated sphere will have facets much like the
facets on a diamond.
Divisions in the Sphere constructor
One of the overloaded constructors for the Sphere class in the Java 3D
API has a parameter that Sun refers to as divisions. This parameter
seems to specify the number of flat polygons that would be encountered in making
one trip around the equator of the sphere. This is illustrated by the top
two images in Figure 3.
Silhouettes of a sphere
The upper-left image in Figure 3 shows the silhouette of a Sphere object
constructed with a value of 4 for divisions. Granted it doesn't look much
like a sphere. It looks like a square instead. However, that is what
you get if you draw a circle by using straight line segments to connect only
four points on the circumference of a circle.
The silhouette in the upper right of Figure 3 looks more like what we would
expect to see for a sphere. This is what you get when you use straight
line segments to connect eight equally spaced points on the circumference of the circle.
It should be apparent that if I continue to increase the number of points on the
circumference of the circle, the result will look more and more like a circle.
The Facets radio button
Returning now to Figure 2, the top radio button is labeled Facets. The text box immediately to the
right of the radio button contains the value 50.0. That is the value that
was used to draw the large sphere in Figure 1. In other words, during a trip
around the equator of the large sphere in Figure 1, you would encounter 50
flat
spots. The two top images in Figure 3 were produced with values of 4 and 8
respectively for the number of facets.
How to adjust the number of facets
To adjust the number of facets, you first select the Facets radio button.
This causes the label above the top slider to read:
Facets on Surface of Large Sphere
You then move the slider to indicate the number of facets that you want
your sphere to have, reading the output in the text field to the right of the
radio button. The scale on the slider ranges from 0 to 100.
However, if you attempt to specify a value less than 4 with the slider, the
number of facets is clamped to 4.
An example display that shows the facets
Now consider the bottom two images in Figure 3. These are the same two
spheres shown as silhouettes in the top two images. However, the spheres
in the bottom two images were not rendered as silhouettes. Rather, they
were rendered in a three-dimensional form instead.
Vertices and polygons
Hopefully you can see some of the regularly-spaced points on the surface of
the spheres (commonly referred to as vertices) as well as the connecting
straight lines and the planar polygons that are formed by the vertices and the
lines that connect them. It is these planar polygons that I refer to as
facets.
Note, however, that even though the lower right sphere in Figure 3 was
created with a Facet value of 8, the sphere actually has more than eight
actual facets. When creating a sphere, the value that is specified for
Facets in Figure 2 is the number of planar polygons that you would encounter
in one trip around the equator of the sphere.
(Once again, I see some visual artifacts in the images in Figure 3 that were
caused by importing them into this HTML document. I don't know if you
will see them or not. However, they aren't in the original images so
if you compile and run the program, you should not see them.)
How to use the program
Now you basically know how to use the program. To
specify the number of facets for the large sphere in Figure 1, select the
radio button labeled Facets and then use the top slider to specify the number
of facets that occur during one trip around the equator of the sphere.
This same procedure is used to adjust all of the scene properties except that
many of the scene properties have three property values instead of only one
property value.
Now that you know about the planar polygons (facets) that comprise the
surface of the sphere, the next thing that you need to learn about is the
shading model. In a nutshell,
the shading model is used to determine the colors that will appear at the
different points on each facet.
Two techniques for shading
Java 3D gives you two choices for determining those
colors:
I'm not going to attempt to explain the technical details of these two
shading techniques. Rather, I will simply suggest that you click on the two
links provided in the above list to read what
Wikipedia has to say about them. If you feel that you need more
information on the topic after you read those explanations, go to
Google and search for the keywords.
I'm confident that you will find more material on the topic than you will have the
time to read.
Example of Gouraud and Flat shading
What I am going to do, however, is to show you an extreme example of the
difference between the two shading techniques and show you how to select one or
the other when you are running the program.
The image on the left in Figure 4 was produced using Gouraud shading.
The image on the right was produced by applying Flat shading to the same 3D
sphere.
(Note that I made some minor color adjustments to the rightmost image in Figure 4 to cause the two dark brown polygons with the Flat shading
to be distinguishable from the black background.)
Another example
I confess that I chose the example in Figure 4 specifically to emphasize the
difference between Gouraud and Flat shading. However, that is not the case
for Figure 5.
Figure 5 shows two renderings of the same large sphere. This is simply a red
sphere with a Facets value of 50 being illuminated by a directional-light
source located somewhere above my right shoulder and aimed directly at the center of
the large sphere. The image on the left was rendered using Gouraud shading and
the image on the right was rendered using Flat shading. Even with this
relatively large number of facets, the Flat shading approach causes the facets
to be very visible on the large sphere.
Not a high quality shading technique
As you can see, Flat shading is probably not what you would want to use for
high quality 3D representations of visual objects. However, depending on
your hardware, it may be much faster to compute Flat shading than Gouraud
shading, so you may need to use it on some objects for improved speed if speed
is an issue in your 3D applications.
Switching between the two shading techniques
Once you have established the scene properties and your scene is visible in
the format of Figure 1, all you need to do to switch between Gouraud
shading and Flat shading with this program is to select the appropriate radio button at the bottom of the control panel
shown in Figure 2. In a subsequent installment of
this lesson, I will teach you how to write the Java code to switch between the
two types of shading.
Figure 6 gives us another look at the four small spheres.
Calibration spheres
I placed these spheres in the scene to serve a calibration function.
Each sphere has fixed surface reflection properties (except that they do
switch between Gouraud shading and Flat shading in accordance with the selected
shading button).
As I add new light sources to the scene for the purpose of
illuminating the large sphere shown in Figure 1, I can check to see how
the small spheres are being illuminated by those same light sources. The
results should be very predictable.
Locations of the
spheres
The large sphere in Figure 1 is centered at the origin in the
three-dimensional space having coordinates of x, y, and z. The x-axis is
the horizontal axis with positive values to the right. The y-axis is the
vertical axis with positive values going up the screen. The z-axis is
perpendicular to the screen with positive values coming out of the screen toward
the viewer.
The location coordinates for the four small spheres are given in
Table 1 later in this lesson. The red and blue spheres in Figure 6
each have their centers on the z-plane. In
other words, their z coordinate value is 0.0. A line drawn through their
centers would lie on the z-plane and would go through the origin of the z-plane.
However, as you can see, their x and y coordinates are not zero. Rather,
the line connecting their centers would be at an angle of 45-degrees relative to
the x-axis.
The locations of the white and green spheres is such that they lie on a tilted
plane that goes through the origin and also goes through the line connecting the
centers of the red and blue spheres. The white sphere is closest to the
viewer (positive z-coordinate value) while the green sphere is further away from the viewer
(negative z-coordinate value). If you were
to view the four small spheres from a vantage point perpendicular to that plane, they would lie on
the four corners of a square inscribed on that plane.
Location and nature of the light source
While the surface reflection properties and the locations of the four small
spheres are fixed, the locations of light sources that can be used to illuminate
them are not fixed.
The light source that is illuminating the four spheres in Figure 6 is an
invisible omnidirectional source of white light located at the origin in the
three-dimensional space. This places the light source in the middle of the
four spheres. As you can see, the appropriate portion of each of the
spheres is illuminated for an omnidirectional-light source at that location.
I'm unaware of any source of light in the physical world that behaves this
way. The light source itself is invisible, but the impact of that light
source on the spheres is visible.
A heat analogy
This is sort of like having an omnidirectional source of heat
at the origin that causes the spheres to glow in their own distinctive colors
when they become warm. (It would probably be possible to build a system
in the physical world that behaves in this fashion.)
For the record, this light source is an object of the PointLight
class, which I will discuss in more detail later.
The scene properties
The scene properties for the four small spheres are shown in
Table 1.
|
|
White |
Red |
Green |
Blue |
|
Facets |
50 |
50 |
50 |
50 |
|
Shininess |
128 |
128 |
128 |
128 |
|
Emissive Color |
0.1, 0.1, 0.1 |
0.1, 0.1, 0.1 |
0.1, 0.1, 0.1 |
0.1, 0.1, 0.1 |
|
Ambient Color |
1.0, 1.0, 1.0 |
1.0, 0.0, 0.0 |
0.0, 1.0, 0.0 |
0.0, 0.0, 1.0 |
|
Diffuse Color |
1.0, 1.0, 1.0 |
1.0, 0.0, 0.0 |
0.0, 1.0, 0.0 |
0.0, 0.0, 1.0 |
|
Specular Color |
1.0, 1.0, 1.0 |
1.0, 1.0, 1.0 |
1.0, 1.0, 1.0 |
1.0, 1.0, 1.0 |
|
Coordinates |
-0.5, -0.5, 0.5 |
-0.5, 0.5, 0.0 |
0.5, 0.5, -0.5 |
0.5, -0.5, -0.0 |
|
Table 1 |
The column headers identify each of the spheres according to the dominant
colors of the spheres in Figure 6. The row headers identify each of the
scene properties in the same order as Figure 2 with the addition of the location
coordinates of the spheres in the bottom row.
Color information
For the rows that contain color (surface reflection) information, the
three values in each cell correspond to the values for red, green, and blue
respectively.
(In Java 3D, color values are given in the range from 0.0 to 1.0 as
opposed to the possibly more familiar 8-bit integer range from 0 to 255.)
The White sphere
For example, the information in the Diffuse Color cell for the
White sphere indicates that this sphere will reflect red, green,
and blue light from a diffuse light source equally well.
- red = 1.0
- green = 1.0
- blue = 1.0
Therefore, when this sphere is illuminated by a diffuse source of white
light, it will appear to be white or gray depending on the intensity of the
light.
The Red sphere
On the other hand, the information in the Diffuse Color cell for the
Red sphere indicates that it will reflect red light very well, but
won't reflect green or blue light at all.
- red = 1.0
- green = 0.0
- blue = 0.0
Therefore, when this sphere is illuminated by a diffuse source of white
light, it will appear to be some shade of red depending on the intensity of the
light. The same is true if it is illuminated by a diffuse source of red
light.
Location coordinate information
For the bottom row containing location coordinate information, the values in
each cell correspond to the x, y, and z coordinates values respectively.
These coordinates specify the locations of the four small spheres that I
attempted to describe verbally earlier.
Similarities and differences between the small
spheres
Much of the information in Table 1 will mean more to
you as we proceed through an explanation of the various scene properties.
For now, suffice it to say that many of the values are the same among the four
spheres. The major differences between the small spheres are the ambient color,
the diffuse color, and the location coordinates. The emissive color and
the specular color of all four spheres are the same, as are the number of facets
and the shininess.
That brings us to the
scene property that is probably the easiest to deal with: Emissive
Color.
Here is part of what
Wikipedia has to say about emissive light (not necessarily with respect
to the Java 3D API):
"The materials also specify "emissive light" for the polygons. The
emissive light is light that the polygon itself emits (for example your
monitor). In case the polygon is emissively lit you can think of it as a
visible light source, which is NOT emitting light to the surrounding
objects!"
Here is some of what
Bouvier has to say about emissive color, specifically with regard to the
Java 3D API:
"The Material object allows the specification of an emissive color.
This can be used to create the effect of a
glow-in-the-dark object. Having an emissive color does not make the visual
object a light source; it will not
illuminate other visual objects."
A stand-alone scene property
The emissive color scene property is easy to deal with because, unlike many
of the other scene properties, it stands completely on its own and doesn't
require interaction with other scene properties. As described above, a
visual object can emit its own light. This is controlled by the emissive
color scene property shown in Figure 2. When a visual object emits its
own light, it does not illuminate adjacent objects.
Unlike the physical world
Once again, I am unaware of any source of light in the physical world that
behaves in this manner. Most illuminated objects that I am aware of will
cause adjacent objects to also become illuminated to some extent.
Color and intensity is variable
In any event, given that restriction, the emissive color scene property can
be used to cause an object to be illuminated in any color at any of the
allowable intensity levels. This is shown in Figure 7 where the large
sphere is illuminated with four different colors and intensity levels by
emitting its own light.
Small spheres are not illuminated by the large
sphere
Note that the small spheres in the scene in Figure 7 are not illuminated by
the light being emitted by the large sphere. However, if you look very
carefully, you may be able to see that each of the small spheres also emits a
small amount of light on its own. This is indicated by the values in the
Emissive Color row of Table 1.
No shading, Gouraud or otherwise
Also note that even though the visual objects in Figure 7 are spheres, there is
no shading to create an optical illusion causing them to look like spheres.
Rather, they look more like colored disks. Visual objects that are
illuminated by emitting their own light are not shaded in the Java 3D API.
Rather, the entire surface of the object is illuminated uniformly.
How to use emissive light
To illuminate the large sphere by causing it to emit its own light, select
the Emissive Color radio button in Figure 2. Then adjust the three sliders
to cause the color mix and the intensity of the illumination to be what you
need.
The color values that you specify by adjusting the sliders will be displayed
in the three text fields next to the radio button in real time as you adjust the
sliders. Also, the color and intensity of the large sphere in the display
will change in real time as you adjust the sliders. Things are about to get a little more complicated. Up to
now, just about everything that we have done has been
done on a standalone basis. However, from this point forward, you will
have to think in terms of the reaction of a surface having specified reflection
properties to a light source having its own set of light-radiation properties.
Light sources and reflections
We saw earlier what Bouvier has to say about the
three kinds of real-world lighting reflections, and the simulation of those
three kinds of reflections in the lighting model of the Java 3D API:
- Ambient reflections
- Diffuse reflections
- Specular reflections
At this point, we are interested in ambient reflections only.
We will take up diffuse and specular reflections later in the lesson.
A link between ambient color and ambient light
Figure 2 contains the following two radio buttons, which
are the main topic of this portion of the lesson:
- Ambient Color
- AmbientLight
The Ambient Color radio button makes it possible for you to control a
set of surface reflection properties that will cause the large sphere to react only to light
radiated
from an object constructed from the AmbientLight class.
The AmbientLight radio button makes it possible for you to control a
set of properties belonging to an object constructed from the AmbientLight
class.
What is ambient light?
In the Java 3D API, ambient light is light that arrives uniformly from all
directions. All visual objects in a scene that is illuminated by ambient
light will be uniformly illuminated according to their surface reflection
properties for ambient light.
A real-world example
As a real-world example, a sphere illuminated by ambient light is similar to what you would
experience if you were to hang a basketball from a clothes line in an open field
on a very cloudy day. To a large extent, the basketball would be uniformly
illuminated over its entire surface (although the top may be a little
brighter than the bottom).
Although the sun would have been the original (highly directional)
source of the light, that light would have been so heavily filtered, scattered,
and reflected by the cloud layer, the ground, and the other physical objects in the
region that the light would appear to come from all directions
uniformly. As a result, there would be no shadows and the entire
surface of the basketball would be almost uniformly illuminated. Light
sources constructed from the AmbientLight class in the Java 3D API
simulate this kind of light with the illumination being completely uniform over
the surface of illuminated objects.
How does ambient color (reflection)
fit into this picture?
Ambient surface reflection properties and light radiated from AmbientLight
objects are uniquely tied together. Both must exist with non-zero states
for either to have an impact on the scene. If there is no ambient light
illuminating the scene, changing the AmbientLight surface
reflection properties of a visual object will have no impact on the scene. Conversely, if
there are no visual objects in the scene having non-zero values for the ambient
surface reflection properties, even high intensity ambient light
sources will have no impact on the scene.
Examples of ambient light
Four examples of the interaction between ambient surface reflection
properties and an ambient light source are shown in Figure 8.
The
ambient light source
For all four images shown in Figure 8, the color and intensity properties of
the ambient light source were:
- red = 0.50
- green = 0.75
- blue = 1.00
This is colored light. It is not white light because there are
different intensity levels from each of the primary colors.
The large sphere is not initially visible
The large sphere is not visible in the upper-left image in Figure 8.
This is because all three ambient reflection properties were set to 0 for
the large sphere for this image. However, the four small spheres are visible. This
is because each of the small spheres has non-zero values for its ambient reflection properties as shown in the Ambient Color row of Table 1.
Reflection color properties react to light
source colors
In order for a visual object to be visible when subjected to any light source
in the Java 3D API, the reflection color properties of the object must contain
color components that are also contained in the light. For example, a
visual object having only red reflection properties subjected to a light source
having only green or blue color components will appear to be black.
When there is a color match between a color component in the surface
reflection properties and a color component in the light source, the numeric
values of the matching color components are used to determine the intensity
of that color component in the resulting overall color of the visual object.
Ambient reflection properties react to ambient light
source
The AmbientLight color values for the light source combined with the
ambient reflection properties for each of the small spheres resulted in
the colors shown for the small spheres in each of the images in Figure 8.
For example, the small sphere in the lower left of each image doesn't appear
to be white as was the case in Figure 1. Rather, even though the
ambient reflection properties of this sphere contain the maximum possible values for
red, green, and blue, (and the sphere would appear to be white when subjected
to white light) in this case it was subjected to
colored light and therefore appears to be more blue than white.
Add a red reflection property to the large sphere
The red ambient reflection property value for the large sphere was increased
from 0.0 to 1.0 for the upper-right image in Figure 8. This caused the large
sphere to become visible with a somewhat subdued red color.
(Even though the value of the red reflection property for the large
sphere was increased to its maximum possible value, the contribution of
red
light in the source was only at half its maximum value, resulting in a red
color with subdued intensity.)
Add some green and blue reflection capability
For the lower-left image in Figure 8, the green ambient reflection property
value for the large sphere was increased from 0.0 to 0.75 producing the olive
drab color shown.
For the lower-right image, the blue ambient reflection property value for the
large sphere was increased from 0.0 to 0.5, producing the gray color shown.
Reversed property value order
After increasing the ambient reflection property values for the large sphere
as described above,
the value order of the ambient reflection property values for the large sphere
was exactly the reverse of the value order of the ambient color property
values for the light source as shown below:
- Sphere: red = 1.0, green = 0.75, blue = 0.5
- Source: red = 0.5, green = 0.75, blue = 1.0
This resulted in the large sphere appearing to be colored gray in the
lower-right image of Figure 8. The
color gray suggests that the color property values from the light source were
combined with the color property values from the large sphere in such a way as
to produce nearly equal contributions of each of the primary colors in the final color.
So far, I have been unable to find the
specifications as to how color values in the reflection properties are
combined with color values in the light source. However, I suspect that
they are simply multiplied together. If so, that would result in a final
overall color for the large sphere with the following color component values:
red = 0.5, green = 0.56, blue = 0.5
As suggested above, these color component values are nearly equal, which
would result in a gray color.
This further demonstrates that the color of a visual object depends on a
combination of the reflective color properties of the object and the radiated
color properties of the light source.
No shading
Another important aspect of ambient light is also illustrated by Figure 8.
In particular, there is no shading with ambient light. This is because the
ambient light is assumed to originate uniformly from all directions
simultaneously so that the entire surface of the visual object is assumed to be
uniformly illuminated.
Doesn't contribute to the 3D optical illusion
Because of the lack of shading, neither emissive color nor ambient color
contribute very much to the optical illusion that is required to cause images
displayed on a two-dimensional screen to appear to be three dimensional.
For that, we need some form of directional light, and that is where we are going
next.
Referring back to Figure 2, we have two more kinds of reflective properties
to consider:
- Diffuse Color
- Specular Color
We also have three more kinds of light sources to consider.
- DirectionalLight (parallel rays)
- PointLight (omnidirectional rays)
- SpotLight (cone-shaped rays )
In addition, we have a shininess property that I have been ignoring up
to this point.
All possible combinations
Any of the three kinds of light sources in the above list will react with either of the two kinds of
reflective properties in the above list. Furthermore, the shininess property comes into play any
time specular reflection is involved. This gives us six combinations of
sources and properties that we need to understand. We also need to
understand how and when shininess impacts the results and to understand the
existence or lack of shadows. This leaves us with the
following major topics to cover in the remainder of this installment of the lesson:
- Diffuse Reflection and the DirectionalLight Class
- The Shininess Property
- Specular Reflection and the DirectionalLight Class
- Shadows
- Diffuse Reflection and the PointLight Class
- Specular Reflection and the PointLight Class
- Diffuse Reflection and the SpotLight Class
- Specular Reflection and the PointLight Class
What is diffuse reflection?
Here is part of what
Wikipedia has to
say about diffuse reflection:
"Diffuse reflection is the
reflection of
light from an uneven or granular surface such that an incident ray is
seemingly reflected at a number of angles. It is the complement to
specular reflection. If a surface is completely nonspecular, the
reflected light will be evenly spread over the
hemisphere
surrounding the surface (2×π
steradians).
The most familiar example of the distinction between specular and
diffuse reflection would be
matte and
glossy paints
as used in home painting. Matte paints have a higher proportion of diffuse
reflection, while gloss paints have a greater part of specular reflection."
What is a directional-light source?
Bouvier has this to say about light radiated from a DirectionalLight
object in Java 3D:
"A DirectionalLight source approximates very distant light sources
such as the sun. Unlike AmbientLight sources, DirectionalLight sources
provide light shining in one direction only. For objects lit with a
DirectionalLight source, the Light vector is constant. ... Since all light
vectors from a DirectionalLight source are parallel, the light does not
attenuate. In other words, the intensity of a DirectionalLight source
does not vary by the distance between the visual object and the
DirectionalLight source."
Assumptions regarding a DirectionalLight source
I will paraphrase much of what I have learned about the DirectionalLight
class. An object of the class is intended to simulate
light sources that satisfy the following assumptions:
- The light source is assumed to be very far away, such as the sun for
example.
- All the light rays travel in parallel paths in the same direction.
- The light wave has a flat wavefront that is at least as wide as the
scene that is being illuminated by the directional-light source.
- There is no attenuation of light energy regardless of the distance to
the light source.
Cannot directly control the actual location of a
directional-light source
As with the other kinds of light sources, you can control the relative
intensity of each of the color components radiated by a directional-light
source.
However, unlike point-light and spotlight sources, you cannot control
the actual location of the directional-light source. The location of
the light source is assumed to be somewhere far outside the 3D space enclosed by
the scene.
Can control the direction to the directional-light
source
However, you can control the direction of the light being radiated by the
directional-light source. Since all light radiated by a directional-light source is parallel, this means that you can also control
the direction to the directional-light source.
Controlling the direction of the light
To control the direction of the light being radiated from a directional-light source, you define a 3D
vector with its
tail at the origin in
the 3D space. You define the vector by providing the coordinates of the
head of the vector.
The direction of the light is then assumed to be parallel to the direction of
that vector. To change the direction of the light, you change one or more
coordinate values that specify the location of the head of the vector.
A lever analogy
In effect, the source is located on the end of a very long lever that goes
through and pivots at the origin in 3D space. The direction of the light
is parallel to the long dimension of the lever.
Your vector can be thought of as representing a very short segment of that
lever
as it emerges from
the origin on the opposite side of the origin from the directional-light source. You can
rotate your vector in 3D space by changing the coordinates of the head of the
vector. This has the effect of causing the other
end of the lever to rotate in the reverse direction in 3D space. This, in
turn causes the location of the
source on the end of the lever to move.
You can think of this as a 3D lever with
the length on one side of the fulcrum (pivot point) being much longer than the length on the
other side of the fulcrum. When one end goes down, the other end must go
up, and vice versa.
Behind, up, and to the right
For example, if you want to cause the directional-light source to be up to the right and
behind the viewer, you cause the vector to point down to the left and into the
screen. This can be accomplished by using the following coordinate
values for the head of the vector:
- x = -1.0
- y = -1.0
- z = -1.0
These are the vector coordinates that were used along with a
d | 
Managing the Modern Network
Business Service Management: Generate Revenue Through IT 
Evaluating Software as a Service for Your Business
Storage Networking: Configuration and Planning
Is Your Disaster Recovery Plan Good Enough? 
