Graphics Reference
In-Depth Information
Responses of
three types of
cones
Spectral
distributions
of light
Color
percepts
External light source
Eye
Retina/optic nerve
Brain
Figure 28.5: Light, described by its spectral power distribution, enters the eye; the three
types of cones each respond and their individual responses are conducted by the optic
nerves to the brain, resulting in a perception of color. These correspond to three distinct
areas of study: physics, physiology, and perceptual psychology.
700 nm
I
(
λ
)
f
(
λ
)
d
λ
.
(28.2)
400 nm
In short, the total response is a linear function of the incoming light
I
, with the
linear operation being “integrate against the response curve.”
With this in mind, we can consider a system diagram (see Figure 28.5) tracing
how a physical phenomenon (the spectral power distribution of light) becomes
a perceptual phenomenon (the experience of color). Notice that this diagram is
slightly simplified, in that it treats the incoming light without considering how
the pattern of light is organized (i.e., what the person is actually seeing). This
omission makes it impossible for this model to account for phenomena like spatial
comparison of colors or color constancy, but the simplification—we can imagine
that all the light arriving comes from a single, large, glowing surface surrounding
the viewer—makes it easy to discuss the basic phenomena of color.
Given the three types of cones, it's not surprising that color perception appears
to be three-dimensional. We begin with the examination of the aspect that's least
related to color, which is
brightness
—the impression we have of how bright a
light is, independent of its hue. By the way, the brightness we are referring to
is not a quantity that has physical units; it's a generic and informal term used to
characterize the human sensation of the amount of light arriving at the eye from
somewhere (a lamp, a reflecting surface, etc.).
To determine relative brightness of light at different wavelengths, imagine an
experiment in which you are shown two lights: a 555 nm reference monospec-
tral light source, and a second monospectral light source whose wavelength
λ
will
be varied over the range 400 nm to 700 nm. We fix a particular wavelength
, and
you are given a knob with which you can control a multiplier for the reference light
source; you adjust it until it has the same brightness as the one at wavelength
λ
λ
.
We record the setting
g
(
to a new value and repeat. When we are
done, we have a tabulation of how effective light at frequency
λ
)
and reset
λ
is at seeming
bright, compared to light at the reference wavelength 555 nm. For each value
of
λ
λ
, the number
g
(
λ
)
tells how much less effective light at wavelength
λ
is in
