A
lighting algorithm is the specific mathematical operation performed on
a surface to shade it to produce the desired lighting effect. Lighting
algorithms can be used to represent the different values of the
lighting equation. Commonly this includes, but is not limited to, the
following lighting values.
Ambient Light
The
ambient value of the lighting equation is used for various purposes. In
simple lighting algorithms, it is often used to add a fill color to
lighten the scene. This fill color is a way to simulate basic area
light by using nothing more than a single color value. This does not
lead to realistic lighting conditions and is just a value used to
brighten the scene.
On the other hand, the
ambient value can be used as the occluding factor. This means the
ambient value is used to store a percentage of how many surfaces in the
scene can possibly block light rays from reaching the surface. This
technique is known as ambient occlusion.
In
ambient occlusion you essentially trace many rays from the surface
being evaluated into the scene. The percentage of those rays that hit
some surface in the scene is recorded and used as the ambient occlusion
value. This value is multiplied by
the other lighting terms to create shadows in the scene. These shadows
appear soft when many rays going in many different directions from the
surface are used. This is because some points on a surface might be lit
differently than others, creating a smooth gradient of gray values
across the surface.
Regardless of whether
the algorithm used involves nothing more than adding a static color
value to the scene or is as complicated as ambient occlusion, the
ambient value itself is generally used to simulate light coming to the
surface from the environment. Global illumination algorithms go a step
further to account for light bouncing off of a surface and evaluate the
resulting color bleed onto the surface. An example of this is to take a
bright blue ball and place it on a plain sheet of paper. If you place a
bright light on top of the ball so that the light is shinning down on
it, the white piece of paper should have some tints of blue on it
because the blue reflecting off of the ball is reaching the paper. This
color bleeding effect is common in global illumination.
Diffuse Light
The
diffuse lighting value is light that has reached a surface and has
reflected off of it evenly, or at least it seems even. When light hits
a surface, it is scattered back into the scene. The smoother the
surface is, the more light reflects in the direction from which it
came. Surfaces such as mirrors are extremely smooth, and lights reflect
back so that a mirror image appears on the surface. For surfaces that
are not as smooth, light often hits microscopic grooves along the
surface that cause the light to scatter in different directions. Since
so much light is hitting such a surface from so many different angles,
diffusely lit surfaces look as if light is being evenly reflected in
all directions.
Specular Light
Specular
light is light that hits a smooth surface and reflects sharply in
specific directions. With diffuse light, the millions of light rays in
nature strike the surface and scatter in so many different directions
that objects look evenly lit regardless of the viewing angle as far as
light reflection is concerned.
With
specular light the sharp reflection of light, instead of the scattering
reflection, in many angles is what creates a highlight on the surface.
Mirrors are so smooth that the light can create clear reflections on
them. For objects that are smooth enough to reflect light more sharply
than others, highlights appear that can be seen from certain
orientations. Since light reflects more sharply in some directions than
others, the appearance of the highlight can change depending on your
viewing angle. This differs from diffuse light, because a diffusely lit
object looks the same regardless of the viewing angle (aside from
shadows, but we’ll assume shadows are not
present), but an object reflecting specular highlights looks different
as the object’s orientation or the viewing orientation changes. This is
similar to rotating a plastic soda bottle while standing in sunlight.
The highlights from the light hitting the object shimmer and change as
the orientation changes since the reflection isn’t evenly dispersed in
all directions.
Lambert Diffuse Model
The
Lambertian reflection model is used to simulate light rays striking a
surface and reflecting back into the scene, where the brightness of a
point on that surface looks the same regardless of where the observer
is located. In other words, Lambertian reflection is ideal for diffuse
light, where the nature of diffuse light is to illuminate a surface so
that it looks the same around all angles.
The
Lambert equation for diffuse light is fairly straightforward and very
popular. To calculate the diffuse contribution, you need the light
vector and the surface normal. With these two pieces of information you
can calculate the dot product between them, and the resulting
floating-point value is what you use as the diffuse contribution. This
is commonly referred to as N dot L, where N is the normal and L is the
light vector. You take this dot product value and multiply it by the
surface color, material color (e.g., textures etc.), and so forth.
Blinn-Phong Specular Model
The
Blinn-Phong reflection model is used to perform real-time lighting in
computer graphics. The Blinn-Phong reflection model is actually a
modification of the Phong reflection model. A reflection model can be
thought of as an algorithm that is used to describe how light reflects
off of objects.
The Blinn-Phong reflection
model is used to create the specular contribution of the lighting
equation. The steps to create the specular contribution using the
Blinn-Phong model are as follows.
1. | Retrieve the normal vector.
|
2. | Compute the light vector as the light’s position minus the vertex position and normalize it.
|
3. | Compute the view vector as the light’s position minus the camera’s position and normalize it.
|
4. | Compute the half vector that is necessary for the Blinn-Phong algorithm as the light vector plus the view vector.
|
5. | The
specular value is the dot product of the normal and the half vector
raised to a specific power (this power is known as the specular power
and is used for shininess).
|
To
use the Blinn-Phong reflection model, the first pieces of data
necessary are the normal, the view vector, and the light vector. The
normal is the surface normal. The view vector is the vector that
describes the direction from the camera’s position to the point being
shaded, which allows the specular contribution to be view-based. The
last vector, the light vector, is a vector from the light’s position to
the point being shaded. If these vectors are calculated in the vertex
shader and sent to the pixel shader, the results can be interpolated to
point not from the original vertex position but from the pixel’s point
being shaded. You could also calculate them in the pixel shader, but
the interpolated values from the vertex shader work just as well.
Once
you have these vectors, you calculate a new vector called the half
vector. This vector is used instead of finding the reflection vector as
is done in the Phong reflection model. It is faster to calculate than
the method used in the Phong reflection model. Like the diffuse
contribution that is calculated using N dot L, the specular value is
found using N dot H, where N is the normal and H is the half vector.
Raising this value by a power, known as the specular power, allows us
to control the amount of shine an object has.
Materials
Material
is a term used to describe the representation of a surface. For
example, a brick wall in a video game might have as part of the “brick”
material a texture image of bricks, a normal map used for bump mapping
to increase the detail of the bricks, a diffuse modifier, and anything
else the artist creates to make the wall look like a brick wall.
These
days a material is a collection of pieces of data used to create the
look of a surface. In the early days of 3D graphics, most games used
textures as the main or sole source of the material. Materials can
include, but are not limited to, the following.
Decal color texture map
Normal map
Alpha map
Diffuse color
Specular color
Emissive color
Ambient color
Vertex shader
Pixel shader
Geometry shader
In
today’s games, materials are complex assets, and some game engines have
their own material systems built into the rendering system. With the
amount of content that goes into games, materials will most likely
continue to grow in complexity as the amount of data necessary to
represent certain surfaces increases.
