Graphics Programs Reference
In-Depth Information
direction, like a one-way street. Input profiles describe how an input
device sees color. Such a profile can be used only to describe the numbers
in the file and can move data only in one direction to the PCS. Figure 1-
13 shows an example of a scanner and camera profile. Note there are
AtoB tags, which convert image data to the PCS, but no BtoA tags. Output
profiles need both tags so we can move in both directions. The AtoB tag
in the output profile allows us to soft-proof our image data. The soft proof
is the preview on-screen showing us how the file should appear when
output. The BtoA tag allows us to convert source data into the printer's
color space.
Profiles can also be very simple or complex in terms of their abilities
to describe a device. There are two basic types of profiles:
Matrix
and
Table-
based
. Table-based profiles are also called
LUT-based profiles
(LUT stands
for
Look Up Table
). Matrix-based profiles are quite simple and need only
a few elements of data, namely a TRC (Tone Response Curve), a white
point, and primary colorants values (see the various sidebars in this
chapter regarding TRCs, gamma, white point, and chromaticity). Most,
but not all display profiles are Matrix-based profiles. RGB working spaces,
which we will cover in the next chapter, are always Matrix-based. Table-
based profiles require much more data and take up more space in kilo-
bytes. The lookup table is three-dimensional, much like a cube of points
throughout color space. Each point contains a correction value that
shows where the point should map in the destination space. Later in the
topic, when profile editing and profile utilities are discussed, we'll look a
bit deeper into the anatomy of ICC profiles. At this point in our journey,
there's little need.
Sidebar
Tone Response Curves and Gamma:
The term
gamma
is used throughout imaging and color
management. The use of the term often describes the relationship between input values and
output values, but it's a bit more complicated than that. The correct term to use is Tone Response
Curve. In a perfect world, all devices would be linear. If you double the input value, the output
value doubles. Most devices are not linear. Instead we require a complex curve to describe the
relationship between input and output. We call this a Tone Response Curve (TRC). The TRC
describes the relationship of the input, such as a digital value, voltage, or light energy, to output.
An example of a TRC is the response of a CRT display. This TRC describes input amplitude
(voltage) and the corresponding light output (brightness). For example, a scanner's TRC would
be the curve that describes the relationship between the light energy that strikes the sensor
and the resulting digital value. Any TRC can be plotted on a grid; this curve describes the entire
range of the device. Imagine a device where the input value and output value are indeed linear.
An input value of 1 produces an output value of 1, 2 produces 2, and so on. If you plotted this
on a graph, you would see a straight line, hence the term linear. Since most devices are not
linear, we instead get a curve on this grid. Curves should be familiar to anyone who has used
Photoshop; it's got a curve dialog. This dialog operates on another type of input/output curve.
Note how an input value maps to an output value as seen in Fig. 1-14.
