FTA (изд-во). Flexography: Principles And Practices. Vol.1-6

Подождите немного. Документ загружается.

closely the source matches D50, with 100

being a perfect match. For color evaluation

in a light booth, a rendering index higher

than 90 should be used. Not all D50 light

tubes are created equal.

Eye Response

The spectrum is divided into the red, green

and blue regions because this matches the

way the human eye sees color. The eye has

three sensors or receptors that detect the

three primary colors. All colors perceived

are a mixture of these primary colors. The

spectral-response functions of the eye are

shown in Figure . They are based on

experiments conducted by the CIE and rep-

resent the standard observer. Because each

person’s eyes are not the same, each person

will see color differently. This component of

color represents an uncontrollable variable.

CIE Color Space

Each of the three components of color

(source, object, observer) has a specific

spectral response curve. These combine to

give the final response curve. Rather than

specifying color in terms of this final spec-

tral curve, it is more useful to combine them

mathematically and create a three-dimen-

sional color space called the CIE perceptual

color space (Figure ). In this space, a

color is uniquely specified by three numbers,

making specification, tolerancing and com-

munication about color feasible. The system

is not perfect however; a unique color in CIE

perceptual color space can be formed by

more than one combination of source,

object, observer. This will be explained in

more detail in the section on metamerism.

CIE perceptual color space is the basis of

quantitative color. There are different math-

ematical algorithms for combining the spec-

tra leading to different numbers, but all have

the general appearance of the model shown

in Figure . In 1976, the CIE standardized

on the model called L*a*b* and the model is

commonly referred to as the CIELab color

space. The additional descriptive term “per-

ceptual” means that this color space is based

on how the eye perceives color. This is in

contrast to name-based description of color

such as “warm red”.

Every color an observer can see can be

represented by its location in CIE perceptu-

al color space, which is commonly described

as L*a*b* and L*C*h°

L*a*b*

L stands for lightness and is the vertical

dimension in color space. Every color has a

lightness or L value. Unlike L, a and b do not

PROCESS COLOR 119

Spectral-response

functions of the CIE

standard observer.

CIELab perceptual color

space model. Hues can

be arranged in a “color

circle”. This “map” or

color space provides

the ability to specify

colors in numerical

terms (L, C, h), which

can be accurately

measured using a

spectrophotometer.

400 500 600 700

Wavelength (nm)

Relative Intensity

Red

Yellow

L=100

White

L=0

Black

Hue

-a

Green

-b

Blue

2 L, a, b without the (*) refers to another color model. Throughout this book,

the L*a*b* model is implied, even if, for clarity, in some of the equations and

diagrams the (*) is omitted.

120 FLEXOGRAPHY: PRINCIPLES & PRACTICES

stand for a perceptual attribute. Instead,

they are the x, y coordinates of the chromat-

ic plane. The chromatic plane is a cross sec-

tion of perceptual space viewed from the top

as a two-dimensional plane. Every color has

a location in this plane (Figure ). The red

color indicated by the circle is located 75

units in the “a” direction and 33 units in the

“b” direction.

L*C*h°

Referring to the same red color as in

Figure , L*C*h° is simply a different way

of navigating to that color (Figure ). This

time, however, the color is reached by going

out 82 units (c) at a 24° angle (h). Geometri-

cally, the difference between L*a*b* and

L*C*h° is the difference between a cartesian

and polar coordinate system. The much more

important difference is that L*C*h° repre-

sents the perceptual attributes of color.

These attributes are described as follows:

L , or lightness, is the lightness or darkness

of the color. The scale goes from 0 for black

to 100 for white.

C, or chroma, refers to the saturation of

the color; zero along the central vertical

axis. A color with a C of 0 is neutral or gray.

The more saturated or pure the color, the

higher the C value. Another descriptive word

used is a strong color as opposed to a weak

color. Values are not capped at any particu-

lar value but rarely exceed 100.

h, or hue, is the perception of the “color”

attribute of color. This may seem like a cir-

cular definition but the best way to describe

hue is to say it determines whether the color

is red or green or purple. The hues are

arranged in a circular fashion so that a par-

ticular direction represents a specific hue.

Color Difference

Once a color is described in terms of a

point in space, the concept of a color differ-

ence follows naturally. It is the geometric

distance between two colors (1 and 2 in

Figure ) and is called delta E (∆E).

Mathematically,

Location of red color in

CIELab color space

shown in the a*b*

plane, which is a slice

through CIELab color

space at a constant

value of L*.

Location of the same

red color of figure 21 at

the same value of L*

and located by a dis-

tance from the center

C* and an angle h.

The distance in CIELab

color space between

two colors is the color

differnce called delta E.

h = 24°

90°

180°

0°

270°

L*C*h°

-a

-b

∆E, CMC (2,1), CIE’94

a=75

b=33

-a

-b

L*a*b

∆E  L

L



a

a



b

b





As a measure of the difference between

two colors, ∆E serves as a specification of

color tolerance. That is, two colors match if

their difference is less than a certain value of

∆E. Unfortunately, specifying an acceptable

∆E value is not a simple matter. Ideally, the

same ∆E would mean the same perceived

color difference throughout color space.

Experience shows that this is not the case. A

small ∆E in a neutral gray would be more

apparent than the same ∆E in a saturated

dark red.

To overcome this deficiency, weighting

factors are introduced into the ∆E calcula-

tion. Currently, the CMC weighting calcula-

tion has widespread acceptance. With some

modification, this has been adopted by the

CIE as CIE’94. When quoting ∆E values or

tolerances, it is essential to know which cal-

culation is being used. Othewise, the num-

bers will be different. Typically, reference is

made to ∆E, CMC or CIE’94 tolerance or

color difference. To complicate matters even

further, there are additional adjustment

parameters used in the CMC and CIE’94 cal-

culations. The usual values for these are 2

and 1 and the CMC color difference may be

quoted as CMC(2,1). Refer to Appendix C

for details.

Metamerism

Every color has a unique point in CIELab

color space—its own set of L*a*b or L*C*h°

values. What is not unique is the combina-

tion of spectral curves of the source and

object which can produce that color. This

leads to the common phenomenon called

metamerism. It means two colors are a

match under one illumination, but not under

a second illumination. Visually, the test for

metamerism simply means looking at the

sample under different illumination sources.

Many light booths provide multiple sources

for this purpose. If no light booth is avail-

able, the sample can be viewed near an open

window to approximate D50 and then under

a standard tungsten filament light bulb (illu-

minant A). A quick, inexpensive and less rig-

orous method to determine if a light source

is the standard D50 is to use the RHEM light

indicator. Available from GATF (Graphic

Arts Technical Foundation), it is an illumina-

tion test target consisting of alternating

patches of two colors that match under D50,

but do not match under different illumina-

tion such as “A” or standard flourescent

lights. Figure illustrates a simulation of

the indicator (actual appearance will be dif-

ferent). Simple visual examination reveals if

the illumination is D50 (or at least close to it).

Similar illumination test targets are available

from other vendors. A more rigorous method

requires a spectrophotometer and will be

described in the measurement section.

Gamut

The range of colors that can be repro-

duced by C, M, and Y inks on a particular

substrate is called the gamut of the system.

Recall that different combinations of the

process colors are used to create all printing

colors. Even if inks of the “perfect” CMY

shown in Figures

and

were avail-

able, one still could not combine them to

PROCESS COLOR 121

A metamerism

indicator, such as a

RHEM Light Indicator,

is used to test if a light

source is D50.

GATF/RHEM LIGHT INDICATOR

IF STRIPES ARE SEEN, LIGHT NOT 5000K.

GATF/RHEM LIGHT INDICATOR

IF STRIPES ARE SEEN, LIGHT NOT 5000K.

D50 illumination

“A” illumination

122 FLEXOGRAPHY: PRINCIPLES & PRACTICES

create all colors that the human eye can per-

ceive. A simple and real life example would

be a red laser, the kind used in the super-

market to scan the bar codes. This has a light

of only one wavelength. The spectrum is a

sharp spike at 633 nanometers. The ultra-

pure red color of the laser beam is consid-

ered an out-of-gamut color, and there is no

way even “perfect” C, M, and Y inks could be

combined to yield such a spectrum.

Every device, including monitors, scan-

ners and proofers have their own specific

gamut; colors they can read or render.

Figure shows the gamut of a digital

proofing system and a flexo press. As is the

case in this example, the proofing system

usually has a larger gamut than the press.

This is particularly true for high-end proof-

ing systems used to make the final contract

proof. In Figure , Y, M, C are the points of

100% yellow, magenta and cyan. R,B,G are

the solid overprints of YM, CM and YC

respectively. The colors which are inside the

polygon connecting these six points make

up the gamut of the device. On this same

illustration (Figure ), the gamut of a

color transparency would be larger still.

The dots inside the diagram of Figure

show the C (chroma) and h (hue) locations

for colors produced with different dot per-

centages of the process-color inks. The

points are actual measurements of the

L*C*h° values and clearly demonstrate that

the hue remains constant when printing dif-

ferent tones of the same color. When printing

on white paper, only the chroma and light-

ness should change. Figure illustrates

that this indeed happens in the real world.

Gamut mismatch is one of the great chal-

lenges in printing, and boils down to the

question of what do to with the colors that

are outside the gamut of the printing press.

When reproducing a color transparency, for

example, there are many colors the press

can not reproduce. The gamut of the tran-

parency must be compressed. The methods

to do this is what much of color manage-

ment, scanner setup, and the conversion of

RBG to CMYK is all about.

The gamut of a flexo

press (shown in black)

and a digital proofing

system (shown in blue).

The dotted lines are the

C*, h° values of differ-

ent dot percentages of

the process colors

yellow, magenta and

cyan.

PROCESS COLOR 123

Color Measurement

efore discussing specific mea-

surements in detail, some gener-

al comments regarding metrolo-

gy are in order. Every measure-

ment consists of two parts: the

value and the measurement

error associated with it. There is no such

thing as “exactly” one inch. Measured with a

ruler, the error might be 0.063". Using a

machinist’s micrometer, the error might be

reduced to 0.001". Using even more sophisti-

cated measurement techniques might

reduce the error to the micron level or even

less, but there will always be some error or

uncertainty associated with the measure-

ment.

Because specifications are based on mea-

surements, this means any specification has

a tolerance. Beyond the mere impossibility

of measuring the “exact” value, a tolerance is

needed for economical and practical rea-

sons. A 1" diameter curtain rod with a toler-

ance of ±0.063" would be fine. The same

0.063" tolerance on a shaft for a piece of

machinery would be disastrous.

All specifications given in FIRST have a

tolerance associated with them. They repre-

sent achievable values and tolerances. The

actual values used for a particular process

and job needs to be agreed upon by all the

parties involved.

DENSITOMETER

There are several types of densitometers

(Figure ). A transmission densitometer

measures the amount of light that has been

passed through an object. This type of den-

sitometer is used to measure films. A reflec-

tion densitometer is used to measure the

amount of light reflected by an object, and is

used to measure proofs and press sheets.

The second category of densitometer

addresses the difference between black-and-

white instruments and “color” instruments.

The word color is in quotes because a color

densitometer doesn’t measure color as has

been defined in CIELab color space. It sim-

Typical examples of a

reflection densitometer

and a transmission

densitometer.

Reflection Densitometer Transmission Densitometer

124 FLEXOGRAPHY: PRINCIPLES & PRACTICES

ply measures the amount of cyan, magenta

and yellow present. A black-and-white den-

sitometer has a similar response as black-

and-white sensors in the human eye and

gives only one density value.

Color sensitivity is achieved by filtering the

light. Table 19 shows the filters used. The

red, green and blue filters are the comple-

mentary colors of cyan, magenta and yellow

and the filters are called the complementary

filters or major filters for those printing col-

ors. The filters may be called C, M, Y and

visual, depending on the model of the densit-

ometer.

Note: The filters in Table 19 are specified

for measuring C, M, Y. This means that the

densitometer is specifically designed to

measure those colors. Ideally, when measur-

ing a cyan, for example, the measured den-

sities of magenta and yellow should be zero.

In reality, there will be some density in all

channels. When measuring a color other

than C, M, Y, the densitometer gives the den-

sities of the C, M, Y components of that color.

This can be used as a process control tool to

keep the color at the same density. It cannot

be used to determine if two colors match.

Density

The scale used is called density. The high-

er the number, the less the light. In order to

make the measurement, the densitometer

needs to know how much light there was to

start with. This is part of the measurement

procedure when using a densitometer.

Measurements are taken either relative or

absolute to the substrate. Relative means the

clear film (for transmission) or non-printed

substrate (for reflection) is the reference.

For absolute, it is no film for transmission

and a white reference supplied by the manu-

facturer for reflection.

The density scale used by the densitome-

ter is logarithmic. This means there is a fac-

tor of 10 between density units. A density of

1 has 10 times the light as a density of 2

(remember, higher density means less light).

Table 20 shows the density for different per-

cent reflectance/transmission values. The

reason to use a logarithmic scale is because,

to a good approximation, it represents the

way the eye responds to light. It means that

a density of 0.2 added to a density of 0.3 will

look very much like a density of 0.5; that is,

densities add.

Using density measurements, other useful

metrics can be calculated: dot percent, trap,

print contrast and hue error/grayness. The

formulas are given in Appendix B.

Dot Percent

One of the key measurements taken by a

densitometer is dot percent. The dot percent

is a calculation based on the measured densi-

ty of the tint and the solid of that same color.

Trap

This is a measure of how well one ink

overprints a second. Again, the measure-

REFLECTANCE OR

TRANSMISSION DENSITY

100% 0.0

10% 1.0

1% 2.0

0.1% 3.0

0.01% 4.0

DENSITY

Table 20

FILTER MEASURES

VIS K

COLOR DENSITOMETER

Table 19

ments are densities. The calculation uses the

relative densities of the overprint, first-down

ink and second-down ink. The measure-

ments are done using the appropriate filter

(Table 19) for the second-down ink.

Print Contrast

This is a measure of the sharpness of the

print and uses the densities of the solid and

a shadow tint (typically 70%).

Hue Error/Grayness

These are calculations using combinations

of the densities applying all three filters. The

metrics were developed to characterize the

purity of process inks; how well they

approach the

“

perfect” cyan, magenta and

yellow. With the advent of spectrophotome-

ters, the recommended metric is the actual

color (i.e., L*a*b* or L*C*h° values) of the

ink. This is what is specified in FIRST.

SPECTROPHOTOMETER

A spectrophotometer is used to measure

the entire visible spectrum of a sample. The

real color curves presented elsewhere in this

chapter, were all taken with a spectropho-

tometer. The key part of the spectropho-

tometer is an optics head that contains a

light source in a fixed geometry, an element

like a prism which breaks up the light into its

discrete wavelengths, and a detector of the

dispersed light (Figure ). The spec-

trophotometer can either display the spec-

trum, or it can send the spectrum to a com-

puter. Physically, a reflection spectropho-

tometer looks very similar to the reflection

densitometer illustrated in Figure .

Recall that the L*a*b* or L*C*h° values are

a combination of the object and source spec-

tra, taking into account the response of the

standard observer. The optics head delivers

the object spectrum and the standard

observer is well defined and fixed. The effect

of different light sources, such as D50 and

D65, can be calculated, and the spectropho-

tometer can display the resulting L*a*b* or

L*C*h° values under these different sources.

PROCESS COLOR 125

Elements of a spec-

trophotometer showing

the optics head which

gathers the data to gen-

erate the spectrum

which is used to calcu-

late L*a*b* or L*C*h°.

Spectral curves of the

two colors that match

under D50 light but

show a definite visible

difference under "A"

light.

Spectral CurveSpectrophotometer

(Optics Head)

Color Data

L = 45

a = 68

b = 39

400 500 600 700

100

Wavelength (nm)

Intensity

126 FLEXOGRAPHY: PRINCIPLES & PRACTICES

The ability of the spectrophotometer to do

calculations enables it to also function as a

densitometer. Recall that in a densitometer,

the light is filtered as was shown in Table 19.

This is nothing more than a modificaton of

the light source. If the spectrum of the filter

is known, all densitometric values can be cal-

culated. The spectra of the filters have been

defined and are known as status T. Using this

standard, all the metrics used in densitome-

try can be calculated by a spectrophotome-

ter. Like a densitometer, the spectrophtome-

ter can make absolute measurements or

measurements relative to the substrate.

The spectrophotometer can be used to

quantify metamerism. As an example,

Figure shows the spectra of the two col-

ors in the RHEM light indicator. Note that

the two spectra cross at several points, a

condition required for two colors to be

metameric. Illuminated with D50 light, the

colors match to a CMC ∆E of less than 1.

Illuminated with “A” light, the colors match

to a CMC ∆E of 2.86, which is clearly visible.

A common measure of metamerism is called

the metamerism index (MI), which can also

be calculated (see Appendix C). In this case,

its value is 3.6. The metamerism index mea-

sures the difference between the colors

under different light conditions. A low value

doesn’t mean the colors are the same, only

that the visual difference is the same under

both light conditions.

PROCESS COLOR 127

Color Management/Workflow

olor management has received a

lot of attention in recent years.

It has become associated with

CIELab measurement and con-

trol of color. In a real sense,

color has always been managed,

it is just the tools and techniques which are

changing. The particular method of how and

where color is controlled and adjusted is

determined by the particular workflow or

specific procedures and programs used to

put the job on press.

Figure shows a highly simplified dia-

gram of the traditional workflow. A scanner

is the input device for the original art, which

is to be printed in process color. The box

labeled “computer” is the source of the rest

of the design. Generally, process work is not

originated in a software program. The design

and scanned images are assembled into a

single job in the workstation. The electronic

file is then output to film for proofing, or to

make plates for printing.

The key to the management of color, are

the functions labeled in the circles of

Figures , and . The circle next to

the scanner, labeled “setup”, is part of the

scanning function. A scanner operator typi-

cally sets highlights and shadows and

adjusts tone curves for the original to be

scanned. On high-end scanners, the setup

will typically include color corrections or

manipulations. When scanning directly to a

CMYK file, the scanner operator must

choose many of the parameters of the RGB-

to-CMYK conversion. Setups for different

types of originals come from experience,

which is based on how the proof looks. If the

color of the proof is not right, there is a color

correction cycle, either to the file or the orig-

inal can be scanned again with a different

setup. In many flexo operations, in order for

the press to print what is shown on the

proof, the process image needs to be modi-

fied. This is done with a cutback curve. This

is because proofing systems have less dot

gain and the films used to make proofs can

not be used to make plates for the press.

Figure a summarizes the color changes

A simplified flowchart

of traditional color

management workflow.

Scanner

Computer

Workstation

(Assembled

Job)

Digital

Camera

PressPlateFilm

ProofFilm

ICC

Profile

ICC

Profile

ICC

Profile

Color

Correction

ICC

Profile

ICC Workflow

128 FLEXOGRAPHY: PRINCIPLES & PRACTICES

that occur during the process of reproducing

a photo as a printed piece.

In order to discuss color management

using CIELab-based metrics, it is necessary

to modify Figure as shown in Figure .

The biggest change as far as color manage-

ment is concerned is the addition of a cor-

rection in the proofing path of the process.

The dashed lines around film indicate that

there may or may not be film produced at all.

While digital platemaking is not yet as com-

mon, digital proofing certainly is gaining

wide acceptance. Color is handled as before

with one important difference. The correc-

tion in the proofing path modifies the proof

to match the press. The correction is not a

simple cutback curve but rather a complex

set of corrections based on L*a*b* measure-

ments. This correction can be applied to

CMYK files and is a CMYK-to-CMYK conver-

sion. This means that the RGB-to-CMYK con-

version is not part of the correction process.

Any separator or scanner operator who has

invested years of learning how to make a

separation can still use that experience to

make the separations. The aim of this work-

flow is to match the proof to the press.

The last case to consider is shown in

Figure . This is what many people have

in mind when they talk about color manage-

ment using ICC profiles. Notice that a digital

camera has been added as an input device.

In this workflow, every device is character-

ized in terms of how it sees or outputs color.

If those characteristics are known, it is pos-

Cutback

Setup

Scanner

Computer

Workstation

(Assembled

Job)

PressPlateFilm

ProofFilm

Correction

Color

Correction

Modified Workflow

Scanner

Computer

Workstation

(Assembled

Job)

Digital

Camera

PressPlateFilm

ProofFilm

ICC

Profile

ICC

Profile

ICC

Profile

Color

Correction

ICC

Profile

ICC Workflow

This modified flowchart

uses CIELab metrics in

the workflow; a correc-

tion step has been

added to the proofing

path.

In this flowchart, full

implementation of color

management uses ICC

profiles for all input and

output devices.