Internal Process Controls - Control of Color Imaging Systems: Analysis and Design - page 505

Linearizing curve of DMA on
photoconductor or Δ E from paper
from which TRC curve for single
separation is determined

100%

D

DMA on Photoconductor
or Δ E from paper normalized
to a maximum value of
100% or 255

C

B The TRC of DMA on

photoconductor or
Δ E from paper

E

0 100%

E A

Area coverage scaled to a
maximum value of 100% or 255

FIGURE 9.21 Schematic diagram illustrating graphically the static TRC inversion process
shown for one measurement point at area coverage A when the measured tone value is greater
than the reference tone value.

input area coverage A at which the measured tone value on the TRC is equal to D and
the output tone value on the reference TRC is calculated as B. The tone value
on the reference TRC is calculated by simply drawing a vertical line until it intercepts
the reference curve. Now, to

find the tone value on the inverse TRC for area
coverage A, draw a horizontal line (toward right in Figure 9.20 and toward left in
Figure 9.21) until it intercepts the actual TRC. The output tone value at this intercept is
indicated by C (this is also equal to B), whose input area coverage (or tone value) will
be equal to E. Thus, E becomes the tone value on the inverse TRC at area coverage A.
This process is repeated until all the input area coverages have tone values on the
inverse TRC.
This type of inverse TRC has bene ts for full range of tone control from 0% to

100%. In a digital printing system, since every pixel in the image that contains
input contone values is processed through the inverse TRC of the level 3 controller,
the

final curve should be smooth and free of undesirable curvatures or artifacts
introduced during the inversion process. If not, image quality defects or
image noise in the output images may occur, and these are not acceptable for
high-quality color printing. While the process of constructing static inverse TRC

''
looks simple, many practical considerations require attention. For example,
the measurement noise could be different for low area coverage (highlight) patches
as compared to high area coverage (shadow) patches, which can induce undesir-
ably high-frequency structure in the inverted TRC. The rounding of values
between 0 and 255 can also introduce errors which can further sacri

''

ce inversion
accuracy. Since the real goal of level 3 controller is to linearize the tone response
on paper to control the appearance of images on paper, paper-based measurements
may be more desirable than those on the photoconductor. The area coverage
selection for paper-based measurements may be different from those used for
measurements on the photoconductor. We will show later how an optimal