Lighting and Textures
in 3D Max
Lighting and texturing a muppet-style
In this tutorial, well cover
some texturing and lighting techniques that I have found useful
in my own work. Some of the techniques discussed come from various
readings while others are the result of trial and error. As
another note, this tutorial was written for Max 2.x, but most
of it will also work with Max 1.x as well other packages. Ready?
Here we go
Meet the cast
Figure 1 shows the model that well
be using (although any simple model (or even a scene with a
strong focal point) would work fine). I chose a simple model
because it is much easier to see changes from one rendering
to the next in a simple model; as such, its hoped that
by choosing a simple character youll be able to more readily
see the desired changes from rendering to rendering.
The character is currently illuminated
with the default Max lighting and has no textures applied. Lets
fix that now.
The chicken and the egg
Before we start we need to decide
how to deal with the "chicken and egg" problem of
texturing and lighting: the effectiveness of a texture is (unfortunately)
dependent on the lighting in the scene; additionally, lighting
a completely "blank" model doesnt really give
one a good feel for how effective the lighting is. Because of
this dependent behavior, starting with texturing usually requires
additional tweaking once the lighting has been added; conversely,
starting with lighting usually means tweaking it after the texturing
has been applied.
Most of the dependency between
lighting and texturing arises from how bump maps, specularity,
reflectivity, etc. rely on the light source for their effectiveness.
Therefore, it is usually best to add them after the lighting.
So, in this tutorial were
going to apply the diffuse coloring to the model, switch over
to setting up the lights, and then switch back to assigning
the appropriate specularity, bump maps, etc.
Staying with in the lines (simple
The first step is to decide what
type of texturing you want: realistic, cartoonish, stylized,
or other. Well try to create a more "realistic"
looking texture with the goal of creating an image which looks
like its a "muppet."
As muppets tend to be a uniform
color and bears tend to be brown, Ive chosen a ruddy brown
color (167, 91, 49) for the ears and face (Ive also locked
the ambient and the diffuse coloring together). The eyes are
going to be white (255, 255, 255) with blue (90, 50, 220) colored
pupils (technically, there should be blue irises with black
pupils, but since this is a muppet, we can get away by having
a pupil/iris complex) and the nose will be red (223, 49, 49).
Figure 2 shows how the "colored" bear looks with the
Step into the light
Once youre happy with the
coloring (including whatever bitmaps/noise/etc. you want to
use for diffuse and/or ambient coloring), its time to
light the model. We start by changing the ambient color from
black to a dark violet color. In "the real world"
pure black is a very rare color; as such, removing it from your
3D scenes helps to add to their realism. Since the main lights
will be yellowish (218, 218, 192), I have chosen a complimentary
purplish hue (18, 11, 25) for the ambient light. Choosing complimentary
light and shadow colors in this manner helps to add "depth"
(and hence realism) to your image.
For lighting the character well
be using techniques borrowed from "standard" cinematography.
We start with the key light. As this is the main light source
for the model, great care should be taken in its placement.
Place it below the character if you want him/her to seem large
and imposing, place the light way above the character if you
want to make him/her seem smaller and less imposing, place it
a little above the camera to achieve a "neutral" amount
At this time you should also decide
whether or not the light is "motivated" or "unmotivated."
A motivated light has an apparent (or at least strong implied
(such as the sun)) source in the scene. An unmotivated light
simply exists. Use unmotivated lights carefully in "realistic"
scenes because they can make your audience wonder (either consciously
or unconsciously) where the light is coming from and therefore
can break the viewers suspension of disbelief.
In this scene, Im using a
motivated light which could either be the sun or a strong spot
light. Because real world light is seldom white or a shade of
gray, Ive set the lights color to a yellowish hue
(218, 218, 192).
As for light placement, Ive
placed it off to the right above the character (See figures
3a, 3b, and 3c). As you can tell from these figures, Ive
attenuated the key light. Real world lights do not go on forever,
they all have finite ranges. Adding attenuation to your lights
helps to produce more gradations of colors, which, in turn,
make the images look more 3 dimensional. I have also turned
on shadows for the light. The result of this light (and the
change to the ambient light) can be seen in figure 3d.
Figures 3a, 3b & 3c
Filling out quite nicely
As you might have noticed in figure
3d, even though the ambient light isnt black, the left
side of the model tends to blend into the background. To help
address this issue, we add another light to the scene: the fill
light. The fill light provides illumination to the model on
the side "opposite" of the key light. If the key light
is on the right, the fill light will be on the left; if the
key light is above the model, the fill light will be below the
model (see figures 4a, 4b, and 4c). Usually, the fill light
is 1 / 2 to 1 / 5 the strength of the key light. In this case,
Ive set its multiplier to 0.22. For a bit of added color
contrast and depth, Ive also set the fill light to have
a bluish/violet color (205, 192, 218). Just as I did with the
key light, I have also attenuated the fill light (again, refer
to figures 4a, 4b, and 4c). Figure 4d shows the affects of all
of the lights we have added so far.
Figure 4a, 4b & 4c
Backing it up
Although we have filled out the
models left side, it still tends to blend into the background.
To help address this issue, we add another light to the scene:
the back light. The primary purpose of the back light is to
provide a small sliver of illumination along the back edge of
the model to help separate it from the background. I usually
create this light by cloning the key light and moving it around
behind the model. Figures 5a, 5b, and 5c show the placement
of the back light in the given scene. Since I based the back
light on the key light, it has the same color, employs attenuation
(which I have modified appropriately) and also casts shadows.
Figure 5d shows the added illumination offered by the back light.
Figure 5a, 5b & 5c
Kick it in the eye
There are two more lights traditionally
used in lighting characters: the kicker and the eye light. Much
in the same way the back light is used, the kicker light is
used to help pull the model out of the background by adding
some more bright illumination to the side of the model currently
in shadow. Traditionally it is placed low if the key light is
placed high; however, Im trying to achieve a "spot
light" feel so I placed the kicker higher in the scene.
Figures 6a, 6b and 6c show the placement of the kicker light
for this scene. Figure 6d shows the resulting render.
The other light traditionally used
with characters is the eye light. This light is used to make
the characters eyes "pop" thereby making the
characters emotions easier to read. It is especially important
with characters that who have deep-set eyes. As our character
has eyes which are flush with the face and we are going to apply
a texturing trick to help make them pop, Im leaving out
the eye light.
Note: if I were going to animate
this scene, I would probably include the eye light and animate
its values over time to make sure that the proper illumination
was achieved through-out the course of the animation.
Figure 6a, 6b & 6c
Now that all of the lights have
been added, take the time to move them around to try to find
the best placement. For me, this meant moving the back light
closer to the kicker light (thereby making it a back-kicker
combo light (see figures 7a, 7b and 7c). The result can
be seen in figure 7d.
Figure 7a, 7b & 7c
So now that we have a colored,
lit model, its time to apply surface texturing to the
model. Surface texturing is what makes the model stop looking
so flat and uninteresting.
Because theyre a key focal
point, well start texturing with the eyes. First, we want
to make the iris/pupils look less flat. To do this, change the
shading type to metal; make the ambient color (26, 26,
26) much darker than the diffuse color (90, 50, 220); make the
shininess 90 and the Shin. Strength 100. Now the iris/pupils
seem more lifelike, see figure 8. Next we do the sclera (whites
of the eyes).
First, make sure the whites are
suitably shiny (Shininess 61 and Shin. Strength 90). Now comes
the trick which should enable us to avoid using a key light:
turn on reflection, select ray-tracing, and set the amount to
71%. The result is Figure 9.
You probably noticed that the ray-tracing
caused the eyes to pop more, but now they look really bad. The
problem lies in the fact that the rest of the face is being
reflected in the eyes. What we really wanted was just the area
around the eyes to be reflected. So go into the ray-tracing
settings (click on materials reflection map button (which
should say raytrace)) and turn on attenuation setting it to
exponential (start: 0 end: 100 exponent: 2). Now only a small
portion of the face is reflected in the eyes and they still
pop off of the head (see Figure 10).
Now its time to texture the
head and ears. I have chosen to apply a box coordinate system
to the head (I then "acquire" it to apply it to the
ears (this keeps the size of the texture the same on both parts)).
I know that I want these parts to be textured like felt, so
I look through my textures for an appropriate choice. For this
image, I settled on using TREEZ86.JPG for the texture. This
may seem like an odd choice, but TREEZ86.JPG is fairly noisy,
very organic and has subtle changes in color which lend themselves
well to the texturing we want (see Figure 11).
Since the model is using "box"
coordinate system, we want to mix some noise (which is 3D and
not dependent on the coordinate systems used) with the texture
to break up the seams which would otherwise arise. This also
helps to break up any noticeable tiling/mirroring artifacts
that might occur. I mix the noise with the base texture using
a uniform 40% mixture method. The result is shown in Figure
For more depth
And were done! For further
experimentation, you could try to increase the amount of realism
in this scene through the use of gels. In cinematography, a
gel is a lens or colored film added to a light to change its
characteristics. In 3D work you can do this through the use
of projection maps. For instance, use a noise-based projection
map on a light to help break up the uniform flatness the 3D
lights typically have and add even more depth/realism.
Yet another way to increase realism
is through the use of volumetric light. Real world light (especially
spot lights) tend to illuminate the dust particles in the air.
Adding a little bit of volumetric lighting to a scene can help
to mimic this behavior and again increase a scenes believability.
For other examples of the work of Shawn Lewis, see DPM's Gallery