Introduction
to Color Management Part 1: Device Dependent and
Independent Color
Color can be of the most important
elements of a successful digital image. This makes
accurately communicating color essential to the imaging
process. While this might seem to be a straightforward
process, in practice we often find that maintaining the
appearance of a digital file across different printers and
monitors can be difficult or even impossible. Sometimes
this is due to the inherent limitations of the devices in
question. In most cases however, the file can be adjusted
to print or display correctly on each device. If this is
true then the issue is not a device limitation but one of
inadequate color communication. One would expect a file to
retain it’s general appearance across a variety of
devices without adjustment, but this is rarely the case.
This problem is the reason that digital color management
exists.
The same file
displayed on a monitor and printed
on an inkjet printer. Each device is capable of
accurately reproducing the file's color, but due to
poor color communication, niether is depicting
its color correctly.
The system for digital
color management that is in use today began as an Apple
Macintosh project called Colorsync. Because a color
communication system is of limited use if it is used by
only a limited number of computer systems, Colorsync was
eventually spun off into a non-profit
platform–independent standards organization. This
organization is called the International Color
Consortium,
or ICC. The group currently develops and maintains
virtually all standards related to digital color
management, and today “ICC” is nearly
synonymous with digital color management.
The same files
from the same devices
with proper ICC Color management.
At its essence, color management is about
color communication. Colors are “translated”
from one method of display or printing to another. In order
to use color management, we must understand some basic
digital color communication terms. The most basic of these
concepts is the idea of using color modes, which are also
called color models. Since almost any color can be made
from a mix of three well chosen primary colors, color modes
typically describe colors in terms of three attributes.
This is known as tristimulus color and the three attributes
used by a color mode may be actual color primaries or
theoretical axes in a three-dimensional space or graph.
RGB is by far the most common color mode and is the native
color mode for nearly all of today’s input and
display devices. RGB color in the digital world is defined
by a set of three numbers that each range from 0-255. These
numbers refer to the amount of red, green, and blue
respectively used to make the color. The higher the number,
the more of that primary (red, green, or blue) the color
mix contains. If all three primary numbers are equal, say
R50, G50, B50, then the mixed color is neutral- a dark or
light grey. As with all RGB colors, higher numbers mean a
lighter color. R255, G255, B255 is pure white and if you
haven’t already guessed, R0, G0, B0 is black.
The idea of sending color information as red, green, and
blue signals goes back to the early days of color
television. At that time, no one defined exactly what hue
or shade of red, green or blue would be used. Since the RGB
primaries are undefined, a color that is mixed from them
also lacks definition. This is easily demonstrated at your
local electronics store. A wall of televisions each gets
the same RGB signal, but each one displays color a little
bit differently. This is due to each television using its
own version of red, green, and blue phosphors.
Unfortunately this limitation has survived into the digital
age. Without color management each device displaying or
printing RGB values is free to interpret the color in terms
of whatever version of red, green or blue is available.
Because the colors that RGB numbers refer to are somewhat
ambiguous, we call color modes like RGB “device
dependent,” meaning that the color specified
literally depends on the device. RGB numbers by themselves
do not refer to a specific “real” color.
Each television
gets the same RGB signal
but displays color in a slightly different
way.
Obviously this presents
problems for those of us concerned with accurately
communicating color. As you might guess, device independent
color modes exist. These modes use color numbering systems
that have specific color meanings under a given lighting
situation. The most commonly used of these is Lab. Lab
describes a color’s lightness (lower “L”
values are darker), red verses green attributes (positive
“a” values are “red”, negative
“a” values are green), and blue verses yellow
attributes (positive “b” values are
“yellow”, negative “a” values are
blue). Lab color is based on how humans actually perceive
color and can be measured in the real world by instruments
such as spectrophotometers and colorimeters.
These real-world strengths of the Lab color mode are offset
by some significant challenges in the way that Lab is
handled digitally. The hue axes of Lab, a and b, work as
opponent pairs. To make a color more yellow for instance,
you would add “b,” and to make it bluer you
would subtract “b.” Many people do not find
manipulating color in this way to be very intuitive. And
since no digital devices can print or display a Lab signal
directly, color information must be converted out of Lab
and into a device dependent color mode in order to be used.
Three-dimensional
graph of the Lab color model.
Click here to go to Part 2 of the introduction