Three-Dimensional Control of Color Management Systems - Control of Color Imaging Systems: Analysis and Design - page 398

using soft gamut mapping with ray-based control method, we de ne x c equal to zero

(e.g., 0,0,0) x 0 along the neutral axis (e.g., 50,0,0) and ne tune the parameters a and

b . After that, each node color in the inverted LUT is updated with the new input L*

to output L* map (see Problem 7.12).

7.7 EVALUATION OF PROFILES

7.7.1 G AMUT U TILIZATION AND R OUND T RIP A CCURACY
Color gamut and its utilization depend on various factors. Colorants of toner (or ink),
process physics, type of GCRs, inversion methods, and node spacings are some of
the factors that contribute to the gamut of colors that a printer can produce. The
reproducible gamut is generally smaller than the available process gamut.
The printing industry has access to numerous commercially available pro ling

packages. It is not impossible to find that colors that can be produced using one

pro le may be impossible with another. Hence, it is important to assess the gamut

utilization for a given pro le from an awareness and competitive point of view.

Otherwise, the printer may not reproduce all the colors in an RGB image even though
the printer

is CMYK gamut is capable of producing those colors.
One way to quantify the gamut utilization is through computational procedures
by using A2B and B2A tags of the ICC Pro

'

le. A2B tag represents the printer
model and B2A tag contains the gamut-mapped inverse LUT. Using them in
sequence for a set of in-gamut test colors described in L*a*b* space can sometimes
give a good indication of the quality of inversion. By computing D E a * ,

D E ( D E 2000

or any other D E) values for a set of in-gamut test colors between their input

L*a*b* values for the B2A tag and the output L*a*b* values from the A2B tag,
we can quantify the round trip accuracy. At the same time if the test colors are
suf ciently dense and well distributed, we can also generate the gamut volume.

Ideally, the inversion round trip error for a colorimetric pro

le must be zero and
the round trip gamut volume should be close to 100% when compared to the native
CMYK gamut. If the gamut volume is not 100%, then one should produce a vector
plot between the two sets of L*a*b* values to identify the regions of colors
(e.g., chromatic greens ciencies are present.

The gamut volume and round trip accuracy is computed as follows. Create a
large number of in-gamut test colors in CMYK space (e.g., 5000). Obtain their
estimated L*a*b* values for each of the test colors using the A2B tag. Now use
these estimated L*a*b* values as inputs to the B2A tag. The output of B2A tag is
then fed as inputs to the A2B tag to produce second set of L*a*b* values.
The reproducible color gamut of the printer represented in the printer model is the
volume covered by the second set of colors. The gamut volume can be obtained
using the methods described in Ref. [135]. The roundtrip accuracy is equal to the

= reds) where the de

D E

between first and the second set of L*a*b* values.
Another way to quantify the gamut utilization is by using the printer instead of
the A2B tag of the ICC pro le. First a set of L*a*b* values are measured using a

sensor, such as an in-line or of ine spectrophotometer for a set of CMYK test colors.