1

Colour & Quality

Printed in the Federal Republic of Germany.
All rights reserved. No part of this publication may be
translated, reproduced, or transmitted, in any form or by
any means, without the written permission of the copyright
holder. Any offense will be prosecuted.

Photography
Heidelberger Druckmaschinen AG
Tony Stone, Page 8
Bavaria. Page 31

Copyright © 1995
Heidelberger Druckmaschinen AG

Kurfürsten-Anlage 52 — 60
D-69115 Heidelberg
Telefon + 49-62 21-92-0
Fax + 49-62 21-92-69 99

Fee: DM 15.00
TABLE OF CONTENTS

Light and colour

1.1 Light is colour
1.2 Visual perception of colour
1.3 Colour mixture
1.4 Systems of colour classification

Colour reproduction in printing

2.1 Ink film thickness
2.2 The significance of the halftone value in printin
2.3 Contrast
2.4 Colour balance
2.5 Ink trapping and colour sequence
2.6 Print control strips

Densitometry

3.1 Measuring principle of a reflection densitomete

3.2 The use of filters in densitometry
3.3 Measuring values in densitometry .
3.4 Measurement
3.5 Evaluation
3.6 Standardisation in printing
3.7 Limits of densitometry

Colorimetrics
4.1 Measuring colour
4.2 Tristimulus values/white reference
4.3 Standard illuminants
4.4 Standard observer/colour matching functions
4.5 Evaluation with a spectrophotometer
4.6 Colour difference AE
4.7 Munsell
4.8 Tristimulus method
4.9 Spectral colour measurement
4.10 The measuring principle of the Heidelberg
CPC 21 spectral quality control
4.11 Proof and colour control strips
4.12 Ink control with Heidelberg CPC 21
4.13 Advantages of colorimetrics for offset printing
 1
Light and colour

1 Light and colour

1.1 Light is colour 8
1.2 Visual perception of colour 10
1.3 Colour mixture 12
1.3.1 Additive colour mixture 12
1.3.2 Subtractive colour mixture 13
1.3.3 Autotypical colour mixture 15
1.4 Systems of colour classification 16
1.1 Light is colour

We live in a colourful world. With the help of colours we

brighten up our surroundings to make us feel good.
Interior design and colour schemes directly influence our
impressions and feelings. Colours that go well together
create a harmonious balance, putting us in a good mood.
The printing industry also uses colours to make presenta-
tions more effective. The quality demands on printed
materials supplied to customers are increasing all the time.
To meet these new demands, new standards of quality have
to be introduced.
In order to judge colours we have to "see" them.
For this purpose we need light.
The sun emits light — it is a primary light source.
Most objects in our environment, however, do not emit light
themselves. They are so-called secondary light sources.
We can perceive them and their colours only if they are
illuminated by light.

88
Light is radiation which propagates very quickly — at a speed
of 300,000 kilometers per second. Strictly speaking, light
consists of electromagnetic oscillations spreading out from
their source like waves. Like a water wave, each light wave
consists of a crest and a trough.

wave crest

wave troug I

Waves are classified on the basis of either their length or the

number of oscillations they perform per second.
Wavelengths are given in units such as kilometers, meters,
centimeters, millimeters, nanometers or picometers.
The number of oscillations per second — the frequency — is
measured in Hertz.

Waves of different lengths possess different properties.

X-rays, for example, are used in medicine for diagnostic
purposes, while many households are already equipped with
microwave ovens. Other kinds of waves serve to transmit
telephone calls as well as radio and television programs.

Only a very small range of electromagnetic waves is seen by

us as coloured light. The visible portion of the wave spec-
trum lies between 380 nm (ultraviolet light) and 780 urn
(infrared light). By means of a prism, light can be broken up
into its colour components. White light, being composed of
all the colours of the spectrum, is broken up into all the
colours of the rainbow.

The following illustration shows how the wavelengths from

red to green to blue become shorter and shorter.

99
1.2 Visual perception of colour

It is only in conjunction with light that colours become

"visible" - but why?

Colour can not be regarded as a characteristic feature of an

object, as can its shape. Yet it is a property of objects to
either absorb or reflect light of certain wavelengths.
We can only perceive the colours that correspond to the
reflected wavelengths.

10
10
If white light reaches an object, one of the following may
occur:

All the light is absorbed. In this case we perceive the

object as black.
All the light is reflected. In this case the object appears as
white.
All the light is let through the object. In this case the
colour of the light does not change.
Part of the light is absorbed, the rest is reflected.
We perceive a colour whose hue depends on which
wavelengths are reflected and which are absorbed.
Part of the light is absorbed. the rest is transmitted.
We see a colour whose hue depends on which
wavelengths are absorbed and which are transmitted.
Part of the light is reflected, the rest is transmitted.
Under these circumstances the colour of the reflected and
that of the transmitted light changes.

The properties of the illuminated object determine which of

the above-mentioned effects is likely to occur.

Light reflected or transmitted by an object is received by our

eyes and transformed into nervous impulses, which trigger
the colour sensation in our brain.

11
11
 The retina of the human eye
contains light-sensitive cells.
There are two kinds of cells: rods
and cones. The rods distinguish
between bright and dark,
whereas the cones react to
colours. There are three kinds of
cones, each of which is sensitive
to certain wavelengths. Part of
them reacts to light within a
range of 400 to 500 nm and is therefore sensitive to blue
light. Other cones can "see" only within a range of 500 to
600 nm, i.e. only green light. The third kind of cone is recep-
tive to red light, which lies within a range of 600 to 700 nm.

This composition of rods and cones renders the huMan eye

so sensitive that it is capable of perceiving and distin-
guishing millions of colours.

1.3 Colour mixture

1.3.1 Additive colour mixture

An additive mixture of colours is a superimposition of light

composed of different colours. If all the colours of the spec-
trum are added together, the colour white results.

Red, green and blue are the additive primary colours.

They are so-called one-third colours because each
represents one third of the visible spectrum.
The principle of the additive colour mixture can be illustra-
ted very well with three diascopes, each of which produces a
light spot on a screen in one of the three additive primary
colours.

12
12
 Within the overlapping areas of the three
green red yellow light spots the following secondary
green blue cyan colours emerge:
blue red magenta
blue red + green white
no light black

The principle of additive colour mixture is used in colour TV and

in the theater to produce all the colours of the visible spectrum.

1.3.2 Subtractive colour mixture

For subtractive colour mixture individual colour components

are taken from white light. If all the colour components are
removed, the colour black results.

Cyan, magenta and yellow are the subtractive primary

colours. They are two-third colours because each represents
two thirds of the visible spectrum.

They can be produced by subtracting an additive primary

colour from white light (for example by means of a filter), or
by superimposing the light of two additive primary colours.

Printing inks are transparent substances that function like

colour filters. Which colour results if a blue-absorbing sub-
stance is printed on white paper?

Blue is removed from white light;

the other components (green and
red) are reflected. The additive
superimposition of these two
colours produces yellow.
This is the colour we perceive.
The printing ink has thus subtrac-
ted one third (i.e. blue) from the
white light (consisting of red,
green and blue).

13
13
Let us assume that two transparent substances are printed
one upon another, for example, the printing inks "yellow"
and "cyan". The substances successively filter the blue and
red portion from the white light. As a result, we perceive
green light. Together, the printing inks have subtracted two
thirds of the colour components.

When cyan. magenta and yellow are printed in layers one

upon another the incident light is totally absorbed, (i.e. there
is no reflection); we perceive the colour black.

14
14
 In subtractive colour mixture the follow-
cyan + yellow = green ing secondary colours will result when
yellow + magenta = red cyan, magenta and yellow are printed
magenta +cyan = blue one upon another.

cyan + magenta + yellow = black

no colour = white

1.3.3 Autotypical colour mixture

Colour images are printed using the four printing inks cyan,
magenta, yellow and black. The black printing ink improves
the sharpness and depth of pictures. This is because, due to
the properties of the pigments of chromatic colours. the
black colour subtractively mixed from cyan, magenta and
yellow, is never really dark black as such.

In offset printing the size of the dots depends on the desired

hue. When printed, the dots of the individual colours are
partly juxtaposed or partly or totally printed one upon an-
other. If we look at the dots through a magnifying glass (see
illustration), we perceive colours which — except for the
white of the paper — are the result of subtractive colour
mixture. However, without a magnifying glass and from a
normal distance, the human eye cannot discern the indi-
vidual dots. In this case the printed colours are mixed
additively.

The composition of additive and subtractive colour mixture

is called autotypical colour mixture.
[Illustration: diagram of additive and subtractive colour
mixtures.]

15
15
1.4 Systems of colour classification

Each individual perceives colours in a different way.

A description of hues by several persons will there-
fore lead to very different results.

Printers, however, need standardised criteria for

colour identification. For this purpose different
systems of colour classification have been established.
Some printing ink manufacturers produce sample books
and give the colours names such as Novavit 4F 434. Others
use colour fans like HKS and Pantone. The colour circle is
another aid. It may consist of 6, 12, 24 or more parts.
All these systems set out examples of the individual colour
hues and give them names. Yet they are never complete and
mostly unsuitable for calculations.

As we have seen, our chromatic sensation depends on the

stimulation of the receptors in our eye, which are sensitive
to red, green and blue. Thus, for an unambiguous classifica-
tion of various colours three values are needed.

With the help of such a system green, for example, could be

described as: green = 0 x red + 1 x green + 0 x blue or, even
shorter,G=0xR+1xG+0xB.

If one draws the primary colours as the axes of a coordinate

system a so-called colour space is obtained.
Many experts have discussed systems of colour classification
and established different notions of how a colour space
should be designed. All of these colour spaces A

have advantages and disadvantages.

16
16
17
Visually perceptible colours within a
lightness plane of the CIE colour space
(standard colour chart).

The saturation of all colours increases from the central

trichromatic unit towards the spectrum locus.

The Euroscale DIN 16 539 describes the position of the

colour locations for cyan. magenta and yellow for three-
colour and four-colour offset printing. It also defines the
colour locations for the subtractive secondary colours red,
green and blue.

The following chromaticity diagram shows the colour loca-

tions as laid down in DIN 16 539 as well as the range of
colours that can be produced in printing. This distribution is
very similar for all lightness values.
Colour hues located within the hexagon can be reproduced
in four-colour offset printing using the colours of the Euro-
scale. Colours outside this area can only be produced with
the aid of additional special colours.

18
 Range of reproducible colours of Euro-
scale DIN 16 539.

In Euroscale DIN 16 539 the following values for coated

paper have been defined for specified printing and measur-
ing conditions:

The values for x, y and Y are measured using spectrophoto-

meters. They exist as handsets or as computing centers with
on-line machine control (as in the Heidelberg CPC 21).

19
19
20
Colour reproduction in printing I
2

2 Colour reproduction in printing

2.1 Ink film thickness 22
2.2 The significance of the halftone value in printing 24
2.2.1 Halftone value shifts 24
2.2.2 Dot gain 31
2.2.3 Print characteristic 32
2.3 Contrast 34
2.4 Colour balance 35
2.4.1 Chromatic composition 36
2.4.2 Chromatic composition with
under colour removal 37
2.4.3 Achromatic composition 37
2.4.4 Achromatic composition with
chromatic colour addition 38
2.4.5 Five-, six- and seven-colour printing 38
2.5 Ink trapping and colour sequence 40
2.5.1 Ink trapping 40
2.5.2 Colour sequence 42
2.6 Print control strips 43
2.6.1 Solid patches (fields) 44
2.6.2 Overprint patches (fields) 44
2.6.3 Colour balance patches (fields) 44
2.6.4 Halftone patches (fields) 45
2.6.5 Slur/doubling patches (fields) 45
2.6.6 Plate exposure control patches (fields) 45

21
Quality assurance in printing aims at a correct and constant
colour reproduction throughout the whole print-run.
In addition to the printing ink and the colour of the printing
stock the most important parameters are ink film thickness,
halftone value, colour balance, ink trapping and the
sequence of colours.

2.1 Ink film thickness

For technical reasons the maximum ink film thickness in

offset printing is about 3.5 pm.

For coated paper and process colours according to

DIN 16 539 the correct colour locations should be achieved
with ink film thicknesses between 0.7 and 1.1 pm.

If unsuitable lithographies, inappropriate printing stock or

unsuitable printing ink are used, however, it may happen
that the standardised points at the corners of the CIE chro-
maticity diagram are not reached.

The range of reproduceable colours also decreases if the

saturation is insufficient. In the illustration the white area
shows how the range of colours narrows with the under-
inking of each of the three process colours.

In terms of physics the influence of the ink film thickness on

the optical appearance can be explained as follows.

Printing inks do not cover the paper; they are, rather, trans-
parent. The light penetrates the ink. In passing through the
ink it encounters pigments which absorb to a greater or
lesser extent certain wavelengths.

22
22
The higher the pigment concentration and the thickness of
the ink film, the more pigments are hit by the incident light
and, consequently, the more of it is absorbed.
Finally, the light rays reach the (white) surface of the print-
ing stock and are reflected. On its way back the light has to
pass through the ink film again and only then can it reach
the observer's eye.

A thick printing ink film absorbs more light and reflects less
than a thin one; the observer therefore perceives a darker,
more saturised, colour hue. The portion of light reaching the
eye thus serves as a basis for the assessment of each colour.

23
2.2 The significance of the halftone value in printing

Next to the printing ink the halftone value is the most impor-
tant parameter for the optical appearance of a colour hue.
The halftone value indicates how much of the printing stock
is covered by ink. The brighter the colour hue to be repro-
duced the smaller the covered area will be.
For the reproduction of different colour hues in classical
scanning with a constant scanning frequency halftone-dots
are used, whose size will depend on the desired hue.
In contrast to this, in frequency-modulated scanning differ-
ent dot spacings are used to produce different hues (all the
dots having the same size). Generally, halftone values are
expressed in percent.

2.2.1 Halftone value shifts

When a halftone dot is transferred from the film via the

plate and the blanket to the printing stock, several factors
may change its geometric size, and thus also its halftone
value.

The alterations in halftone value caused by processing can

be compensated in the pre-press stage. A curve describing
the transfer characteristics is obtained by measuring print-
ing patterns and comparing them to the originals. If during
the entire printing process (from scanning to the finished
print product) the same (standardised) parameters are
always used one can expect the print product to be true to
the original.

Halftone value shifts caused by difficulties in printing, how-

ever, cannot be foreseen. Special attention has to be paid to
them during the printing process. The most important ones
are:

24
24
Transfer of the halftone Influencing factors Features of the halftone dots
in

Film
Mounting Film tape edges, adhesives
Exposure

Development Chemicals,
duration of development

Printing plate Material, abrasion due to

printing
Plate exposure exposure time, vacuum,
side lighting

Dampening Amount of dampening solution

pH value, surface tension,
water hardness,
temperature
Inking Thickness of ink coat,
consistency, temperature
Printing
plate/blanket Rolling/packing

Blanket Material, condition,

surface
Printing blanket/
printing stock Rolling/packing

25
25
Transfer of the halftone Influencing factors Features of the halftone dots
dots

Printing Stock Surface, paper quality

Sheet transport Transfer register
Delivery Offset

The large magnification

clearly shows the first-rat e

result

Increase/decrease of the halftone value

Dot gain means that there is an increase in halftone value

during the printing process as compared to the dot on the
film. The increase is in part due to processing, material, and
machines, and thus cannot be influenced by the pressman
(this aspect is also called enlargement of halftone dots).
To a certain degree, the pressman can counteract dot gain,
especially by manipulating the inking.

Filling-in means the reduction of non-printing areas up to

their complete disappearance. Filling-in may also be caused
by slur or doubling.

Sharpening means a decrease in halftone value as compared

to the dot on the film. In practice, the term sharpening is
often used to describe the reduced increase in halftone
value, though the print may still be fuller as compared to the
film.

26
26
Deformation of halftone dots

Slur means that the form of the halftone dot is changed

during the printing process due to relative motions between
plate and blanket and/or blanket and printing sheet, that is,
a circular point may become oval. Slur in the printing direc-
tion is called circumferential slur, and slur at right angles to
the printing direction is called lateral slur. Diagonal slur
results if both forms of slur occur at the same time.

In offset printing, doubling means that a shadow-like and

unintended, in general smaller, ink dot appears besides the
intended halftone dot. Doubling is caused by an incongruen
retransfer of ink by the subsequent blanket.

The term offsetting refers to those deformations of halftone

dots that are caused by mechanical factors after the printing
process. The term offset is also used to describe the transfer
of ink from freshly printed matter on to another surface.

Points for the pressman to observe

With the help of control strips, dot gain can be monitored

visually and measured in size. For purely visual checks,
signal strips are particularly well suited. Filling-in can be
best monitored with the help of screen measuring elements
with high halftone values.

Dot gain and clogging are mostly caused by excess ink feed-
ing not enough water feeding, too much pressure between
plate and blanket, or by a poorly clamped blanket.
Furthermore, the inking and damping form rollers may not
be well adjusted.

27
27
Under normal printing conditions and precise plate expo-
sure the print is generally fuller than the film. Defects such
as blind plates and build-up of ink on the blanket may cause
sharpening. Remedies might be: more frequent washing of
the blankets and inking units; changing the inks and the
colour sequence; checking plate rollers, printing pressure,
and printing process.

Slur is most obviously signalised by line screens. The paral-

lel lines often indicate the direction of slur. Circumferential
slur normally indicates printing differences that have
emerged between plate cylinder and blanket cylinder, or
excessive printing pressure. This is why each step of the
printing process and the printing pressure should be moni-
tored very carefully. Slur may also be caused by a poorly
clamped blanket or by excessive inking. Lateral slur is
mostly associated with other problems. In this case, the
printing stock and the blanket should be carefully checked.

Doubling is monitored on the basis of the same elements as

slur. In addition, halftone dots have to be examined with a
magnifying glass, since the line screens alone do not allow to
distinguish between doubling and slur. Doubling may be
caused by various factors, most of which are related more or
less closely with the printing stock.

Offsetting problems hardly ever occur with modern sheet-fed

presses. Those areas of a sheet-fed press that support the
freshly printed side of the sheet mechanically are most likely
to cause offset. Stiff printing stock makes offset problems
more likely. Offset may also emerge in the pile or in per-
fector machines.

28
Printed signal elements such as the SLUR strip are a valua-
ble tool for the quick optical evaluation of halftone value
alterations. Signal elements such as the SLUR strip optically
amplify faults in the printing process.

Faults such as dot gain, sharpening, slur and doubling affect

fine screen elements more than coarse ones. The reason is
that small dots are reduced or enlarged by the same values
as large ones. A large number of small dots, however, has
many times the total circumference length of coarse dots of
the same tone value. This is why, during printing, more ink
is deposited around the fine dots than around the coarse
dots. As a consequence, the fine screen image areas will
appear darker. This phenomenon is the basis on which
signal and measuring elements function.

To give an example, the structure and the function of the

SLUR strip will be briefly explained. In the SLUR strip,
coarse screen elements (surroundings) and fine screen ele-
ments (numbers) are combined.

Compared with the uniform halftone value of the coarse

screen, the fine-screen numbers from 0 to 9 have increa-
singly sharp halftone values. When during the print run of a
properly printed sheet the figure 3 and the coarse-screen
field display the same halftone value, then the figure 3 can
no longer be recognized. If dot gain occurs during printing,
however, then the next-higher figure with a sharper screen
approaches the halftone value of the surroundings.
The more dot gain takes place, the more the halftone value
equality shifts towards higher numbers.

With sharpening, this process is reversed. Here the figures

2, 1 or even 0 may become illegible.

29
Yet the figures merely indicate whether dot gain or sharp-
ening occurs. The causes have to be looked for with a
magnifying glass on the plate or on the printed sheet itself.
The SLUR section to the right of the numbers shows whether
there is dot gain, slurring, or doubling. With dot gain in
printing, the word SLUR is not more legible than with good
printing, although the entire field appears somewhat darker.

Halftone dots, however, are less suited for detecting slurring

and doubling. The typical, direction-related, widening in
cases of slur is easier to detect in the SLUR field. In the case
of circumferential slur, for example, the horizontal lines
forming the word SLUR (parallel to the image start) will be
widened. With lateral slur, on the other hand, the area sur-
rounding the word SLUR, which consists of vertical lines,
will be darker.

The illustration shows how dot variations affect the print

result, using dot gain as an example. Even if the dots of only
one colour are larger than desired, a different hue will
result.

30
30
This, of course, is also important for superimposition.
The transfer process used in offset printing usually causes
the dots to become larger. This effect is called dot gain.

Signal strips help to assess the quality of the print result, but
they do not provide information on absolute values and
errors. To assess the quality of halftone values with objec-
tively verifiable numbers an objective measurement method
is therefore needed.

2.2.2 Dot gain

Dot gain is the difference between the halftone values in the

screen film and in print. Deviations result both from geomet-
rical dot variations and from the effect of light entrapment
(see Chapter 3.4.4).

Similar to the halftone value F, the dot gain Z is generally

expressed in percent. (The formulae for the dot gain Z are
given in Chapter 3.5.1).

The dot gain is the difference between the halftone value in

print F , and the measured halftone value in film F .
1 F

31
Since the dot gain is different in the various halftone value
ranges, the figures on dot gain should also include the
halftone value in film. Example: "15% dot gain with
= 40 %" or, shorter, "Z 4 „ = 15%".

Advanced measuring instruments display the dot gain directly.

N.B.: The dot gain Z(%) indicates the difference be-

tween the halftone value on film F, and the
halftone value in print F„ in absolute figures.
It therefore does not refer to the film value!

2.2.3 Print characteristic

The deviation of the halftone value in print F„ as against the

halftone value in film can be clearly represented for direct
use in repro work in the form of what is called a print char-
acteristic.

To determine the print characteristic, screen step scales

with a minimum of three, or even better five or more screen
steps and a solid patch element are printed. A densitometer
is used to measure the ink densities in the solid patch and in
the screen steps, and subsequently the halftone values are
calculated. When the values thus obtained are drawn in a
diagram versus the corresponding film values the transfer
characteristic is obtained. Having used plate making this is a
print characteristic.

It is valid only for the particular combination of ink, paper,

printing pressure, blanket and plate for which it was deter-
mined. If the same job is printed on another press, with
different ink or on different paper, then the print characteris-
tic will be slightly different.

The illustration shows how characteristic 1 runs at an angle

of 45°. It represents the ideal case in which print and film
are optically identical, but which is unattainable under nor-

I 32
32
mal conditions. Characteristic 2 reproduces the halftone
values actually measured in print. The marked area between
the two lines represents the dot gain.

For determining the dot gain in print, the middle-tone range

is the most significant. The print characteristic shows that
here the halftone value shifts reach a maximum.
By means of characteristic 2 the screen film can be adjusted
in such a way that in print (with a normal dot gain) the
desired tone values are attained.

In practice, however, this can only partly be achieved.

33
2.3 Contrast

As an alternative to dot gain the relative print contrast K(%)

is often determined, particularly to check the screen in the
three-quarter tone.

A print should have a contrast as high as possible.

This means that the solids should have a high ink density,
but the screen should still print open (optimum halftone
value difference). When the inking is increased and the ink
density of the dots rises, the contrast is increased. However,
the increase in ink feed is only practicable up to a certain
limit. Above that limit the dots tend to exhibit gain and,
especially in three-quarter tone, to fill in. This reduces the
portion of paper white, and the contrast decreases again.
If there is no measuring device available with a direct con-
trast display, the relative print contrast can be calculated or
determined on the basis of the FOGRA PMS. (The formulae
are given in Chapter 3.4.3).

If the contrast value deteriorates during a production run in

spite of constant ink value in solid DV, this may be a sign
that the blankets need washing.
If the solid density is correct, the contrast value can be used to
assess various factors which influence the print result such as

■ rolling and printing pressure,

■ blankets and underlays,
■ dampening,
■ printing inks and additives.

Since the contrast value, unlike the dot gain, depends to a

large extent on the solid density it is not suitable as a vari-
able for standardisation. This is why in the recent past its
importance has decreased significantly.

34
34
2.4 Colour balance

As already explained, colour hues are reproduced in four-

colour printing by different portions of cyan, magenta,
yellow and black. If their proportions change, the resulting
colour changes. To avoid this the colour portions for the
desired colour hue must be balanced correctly and reliably.

If only the black colour portion changes, the hue becomes

brighter or darker, a phenomenon we do not regard as
disturbing. The same is true when the chromatic colours all
change relative to their portions and in the same direction.

However, we react critically to shifts in the colour hue.

Such shifts occur if the individual colour components do not
change together, or, at worst, if they change in opposite
directions.

Such impairments of the colour balance can be recognized

most clearly on gray balance fields; the colour balance there-
fore is often termed gray balance.

The extent of the inevitable variations in each of the printing

inks during the printing process largely depends on the
principle of image build-up chosen in pre-press.

In the following, the most important types of image build-up

will be described. The schematic diagrams shown refer to
ideal inks, which, however, do not exist in reality. In addi-
tion, there are chromatic changes due to ink trapping in
" wet-on-wet" printing (see Chapter 2.5). This is why, in
practice, the halftone values deviate from the given, theoreti-
cal values. In order to achieve equal hues, the overprint
patches have been corrected accordingly.

35
35
2.4.1 Chromatic composition

The chromatic composition consists of the chromatic

primary colours cyan (C), magenta (M) and yellow (Y).
Black (K) is merely used to intensify the image depth and to
improve the contour accentuation. Dark colour hues are
produced by mixing the three chromatic primary colours.

If, for example, the printing ink cyan becomes darker, equal
portions of magenta and yellow are added; yet their propor-
tion must remain lower than that of cyan. The portions of
magenta and yellow mix with an equivalent portion of cyan
to make black and thus darken the rest of the cyan.

This will be illustrated with an example.

70% C 80% M 90% Y 0% K 240%

The brown colour shown in the illustration was built up in

chromatic colour structuring from 70 % cyan, 80% magenta
and 90% yellow. All in all surface covering amounts to
240%. Black was not used. Due to the high portions of chro-
matic colours, however, this colour balance is difficult to
maintain during the print process.

In addition, with such a large surface covering the consump-

tion of powder, the drying time and the ink consumption
increase.

The chromatic colour structure of the colour brown, as

shown in the illustration, consists of a chromatic and an
achromatic portion. The achromatic portion is composed of
70% cyan, 70% magenta and 70% yellow, which produce
gray if superimposed in printing. The remaining 10%
magenta and 20% yellow are the chromatic portion.

36
36
2.4.2 Chromatic composition with under colour removal

Under colour removal (UCR) — is a variation of the chromatic

composition with part of the achromatic portion being
replaced by black.

Let us assume that a 30% under colour removal from the

brown colour of our above example is required.
The achromatic portion consisting of cyan, magenta and
yellow is reduced by 30% and replaced by the correspond-
ing portion of black. As a consequence, the surface cover
does not amount to 240% but to only 180% with the colour
hue unchanged.

This is a great help for the pressman, since the danger of

blotting is reduced and the colour balance can be main-
tained more easily.

2.4.3 Achromatic composition

In contrast to chromatic composition, in achromatic compo-

sition all achromatic contents are replaced by black.
Chromatic colour hues are thus no longer darkened with the
complementary colours but merely with black. The achromatic
structure of the example thus only consists of magenta, yellow
and black. Altogether, the surface covering amounts to not
more than 100%. This allows the colour portions of cyan,
magenta and yellow to be considerably reduced in all images
and colour hues; the printing process as such is rendered
more reliable and the ink film trapping is significantly im-
proved.

37
37
2.4.4 Achromatic composition with chromatic colour addition

The chromatic colour addition is a variation of the achromatic

composition. If the density of the neutral ink black turns out to
be insufficient, portions of cyan, magenta and yellow are
added to the achromatic colour structure again to improve the
neutral image depth (25 percent in the given example).

This kind of image build-up is widely used today and has

proven to be a valuable method in practice, allowing a good
coordination of image and print qualities.

2.4.5 Five , six and seven colour printing

- - -

Advanced four-colour printing even meets high quality

standards. Nevertheless special sets of colour plates may be
needed with some originals and for highest quality require-
ments.

The reproduceable range of colours can be expanded by

using special colours (besides the four primary colours).
If, for example, an additional red is used in addition to the
four process colours, for example, the range of red can be
enlarged. If necessary, more than one of several special
colours may be used.

38
38
The following illustration shows where the colour values
measured for a seven-colour print are located in the CIE
chromaticity diagram.

0. -
8

o. -
7

0. -
6

0. -
5
590 - •
0. - 600
4 Y+R
•,• 610
o. - • 620
0R
3 R+M

0. -
2

0.
0 0. 0. 0. 0. 0. 0. 0. 0. -110- X
0 1 2 3 4 5 6 7

The hexagon located within the diagram's borders shows

the range of the process colours cyan, magenta and yellow
(measured values). The dodecagon surrounding it indicates
how the range of colours can be enlarged by the additional
colours green (G), red (R) and blue (B).

39
2.5 Ink trapping and colour sequence

2.5.1 Ink trapping

Another variable influencing the reproduction of the colour

hue is the ink trapping characteristic. It indicates how well
an ink is accepted when printed onto another ink as com-
pared to when it is printed onto the printing stock.
A distinction has to be made between wet-on-dry and wet-
on-wet printing.

The term "wet-on-dry" printing is used when an ink is printed

directly onto the printing stock or another, dry ink. If, on the
other hand, an ink is superimposed on a wet colour, one
uses the term "wet-on-wet". For multicolour presses, the
term "wet-on:wet " printing is generally used.

If the coverage is uniform and if the hue is located at the

correct coordinates, then one speaks of good ink trapping.

If, on the other hand, the desired hue can not be attained,
the ink trapping is faulty. This may be the case with all
mixed colours. As a consequence, the range of colours is
reduced and certain colour shadings can no longer be repro-
duced.

If the ink film thickness is correct and if the colour locations

of the primary colours cyan, magenta and yellow are situ-
ated at the correct reference locations, it may nevertheless
be the case that the reference locations of the mixed colours

40
40
red, green and blue cannot be attained due to faults in
superimposition during printing.

The following CIE chromaticity diagram demonstrates the

effects of a faulty ink trapping or an unfavorable colour
sequence on the print result. The white area illustrates the
extent of halftone value reduction caused by faulty ink trap-
ping.

41
41
2.5.2 Colour sequence

The schematic illustration shows the results of three differ-

ent superimpositions of the colours cyan and magenta.

In the first example a layer of magenta was printed on a

single-colour press as the first colour. Then a layer of cyan
was superimposed after drying ("wet-on-dry"). The ink film
thickness of both colours is identical. The ink film trapping
is good and the desired colour location has been reached.

The second example was produced on a multi-colour press.

First a film of magenta was printed on the dry paper (wet-
on-dry). Then a layer of cyan was printed onto the still wet
magenta ink (wet-on-wet). Whereas the magenta ink film
was accepted well by the paper, the ink trapping for cyan
was worse (due to the colour separation during superimpo-
sition in printing). The result is a red-cast blue.
In the third example the wet-on-wet printing method was
also used but with reversed colour sequence (cyan onto
magenta). The result is a blue-cast red.

42
42
In four-colour-print the colour sequence black-cyan-
magenta-yellow has been generally accepted as standard.
This colour sequence is also the basis for the adjustment of
colour consistency in printing ink manufacturing.

To reduce the effects of faulty ink trapping that might occur

in special cases, typon and printing plates should be
carefully checked before mounting. For solid fields, for
example, it might be advantageous to print the lighter form
before the more solid one.

In particular, this applies to the superimposition of screens

and solid ink films. The screen should first be printed on the
white paper and then the ink film onto it.

2.6 Print control strips

For controlling print quality on the basis of measured data,

print control strips are printed with the image. They are
available from various research institutes and suppliers.
However, only originals may be used, since deviations might
occur during copying onto a duplicating film that impair the
measurement results.

Print control strips are available for four- to eight-colour

presses. With print control strips for more than four colours
the number of halftone and slur patches is reduced in favor
of the solid and colour balance fields necessary for the con-
trol of the ink fountain zones.

All print control strips consist of several elements. In the

following, the most important patches of the Heidelberg CPC
colour measurement strip, the FOGRA and Brunner print
control strips, will be described.

43
43
2.6.1 Solid patches (fields)

Solid patches enable the uniformity of the inking to be

checked. It is advisable to use one solid field per printing ink
spaced at the distance of the ink fountain zone width (32.5
mm for Heidelberg). This makes it possible to use solid fields
for the automatic colorimetric control of solids.

2.6.2 Overprint patches (fields)

These elements are designed for the visual and densitomet-

ric assessment of the ink trapping performance.

2.6.3 Colour balance patches (fields)

One has to distinguish between solid and halftone colour

balance fields. In solid patches, the superimposition of cyan,
magenta and yellow must result in an approximately neutral
black. For purposes of comparison, a black solid field is
printed next to the overprint field.

Given correct ink film thickness, standardised colour

sequence and normal dot gain, the superimposition of cyan,
magenta and yellow produces an approximately neutral
gray. Different halftone values are used by manufacturers
for the typons of the various colours.

Heidelberg: 70% cyan 60% magenta 60% yellow

FOGRA: 28 % cyan 21 % magenta 19 % yellow
Brunner: 50 % cyan 41 % magenta 41 `)/0 yellow

Colour balance patches are also used for the automatic gray
balance control of cyan, magenta and yellow.

44
44
2.6.4 Halftone patches (fields)

Depending on the manufacturer, halftone fields may contain

different typon halftone values.

From the measured data of the halftone and solid patches

the dot gain and print contrast are calculated.

Heidelberg Fogra Brunner

70% 40% and 80% 50 % and 75 %

2.6.5 Slur/doubling patches (fields)

Line gratings of different screen angling allow the pressman

to visually and densitometrically check foe slur or doubling
faults (see Chapter 2.2.1).

2.6.6 Plate exposure control patches (fields)

Plate exposure control fields are designed for visual moni-

toring of the plate exposure. The control elements shown
contain microlines and micro reverse lines as well as fields
with dots.

45
45
46
 3
Densitometry

3 Densitometry
3.1 Measuring principle of a reflection densitometer 48
3.2 The use of filters in densitometry 50
3.2.1 Colour filters and brightness filters 50
3.2.2 Polarisation filters 52
3.3 Measuring values in densitometry 53
3.4 Measurement 56
3.4.1 Zeroing on paper white 56
3.4.2 Solid density 56
3.4.3 Halftone density 56
3.4.4 Optically effective area coverage
(halftone value in print) 57
3.5 Evaluation 58
3.5.1 Halftone value in print 58
3.5.2 Dot gain 58
3.5.3 Contrast 58
3.5.4 Ink trapping 59
3.6 Standardisation in printing 60
3.7 Limits of densitometry 62

47
Densitometry is the cheapest and most widespread measur-
ing method in the field of repro work and printing.
Densitometers are used as hand-held instruments or in the
form of automatic measurement devices (scanning den-
sitometers).

There are two kinds of densitometers, used for different

purposes:

■ The transmission densitometer is used in repro work to

measure the blackening of the film (transparent
substrata).
■ Reflection densitometers are used to measure printed
images (opaque substrata).

In the following, the working principles of reflection den-

sitometry will be described in more detail.

3.1 Measuring principle of a reflection densitometer

In reflection densitometry the ink to be measured is illumi-

nated by a light source. The light ray passes through the
transparent (glazing) ink layer and is partly absorbed.
The non-absorbed content of the light is largely scattered by
the printing stock. Part of this reflected light again passes
through the ink and is absorbed again. The remaining non-
absorbed light reaches a detector, which converts the light
into electricity. The result of the measurement with a reflec-
tion densitometer is given in density units.

In the measurement, lens systems are used to focus the

light. Polarisation filters serve to prevent differences in the
measured values obtained from a shining wet surface and
from the surface of a dry ink. Colour filters are inserted for
measurements of colours (see Chapter 3.2.1).

48
48
The illustration explains this principle, taking a coloured ink
as an example. Ideally, the incident white light consists of
equal portions of red, green and blue. The printed colour
contains pigments which absorb the red portion and reflect
the green and blue portions, which is why we call it "cyan".
Densitometers are intended for measurement within the
absorption range of each colour, where density and ink film
thickness closely correlate. In our example a red filter is
used which allows only red light to pass, whereas blue and
green are blocked.

The density of a given ink mainly depends on the pigmen-

tation, its concentration and its ink film thickness. For a
given ink the density is a measure of the ink film thickness,
yet it does not tell us anything about the hue.

49

49
3.2 The use of filters in densitometry

3.2.1 Colour filters and brightness filters

The colour filters in a densitometer are tuned to the absorp-

tion performance of cyan, magenta and yellow.

Common standards such as DIN 16 536 and ISO/ANSI 5/3

define the spectral transmission bands and the positions of
the transmission maxima accordingly. Of the narrow and
broadband colour filters listed there, referred to as status A
and T in ISO respectively, narrow band filters should be
used since the difference in measurement resulting from the
use of different brands of filters is smaller than in broad
band filters.

Colour filters must always be chosen complementary in

.

colour to the printing inks which are to be measured.

The colour black is measured with a visual filter tuned to the
spectral of brightness sensivity of the human eye.
Special colours are measured with the filter which returns
the highest measurement value.

The following three illustrations show the reflexion curves

for cyan, magenta and yellow, together with the respective
colour filters according to DIN 16 536.

50
51
51
3.2.2 Polarisation filters

Densitometers can be used to measure both dry and wet

printing inks. Wet colours have a smooth, shining surface.

During the drying process, the ink adapts to the irregular

structure of the paper surface, and the reflection effect
decreases. If a given ink is measured first in wet and then in
dry condition, different readings will result.

In order to eliminate this effect, two crossed linear polari-

sation filters are inserted in the path of the rays. Polarisation
filters allow the light of only one particular vibration direc-
tion to pass, while blocking all light waves which are
vibrating in other directions. Part of the light rays polarised
by the first polarisation filter are reflected by the ink surface
specularly,'i.e. without altering their direction of vibration.
The second polarisation filter is aligned at an angle of 90° to
the first so that the reflected light waves are inhibited from
passing.

Light rays, however, which penetrate into the ink film and
are reflected either by the ink or by the printing stock, lose
their original polarisation.

They are therefore able to pass through the second polari-

sation filter and reach the detector.

By thus blocking the portions of light reflected by the wet

colour surface, approximately equal readings for wet and
dry inks are obtained.

Due to the absorption by the polarisation filters less light

reaches the detector; the readings obtained with such
devices are therefore generally lower than measurements
made with other instruments.

52
3.3 Measuring values in densitometry

Densitometers display their readings for the ink density D as

a logarithmic number. It is the logarithmic ratio of the
absorbed light for a "reference white" to that obtained from
the measured ink film. In practice, the ink density reading is
mostly referred to as the "density".

The ink density value is calculated using the following formula:

The reflectance factor 13 is calculated as follows:

where Le, is the light reflectance of the printing ink and

Le w the light reflectance of the reference white.

The reflectance factor 13 gives the ratio between the light

reflectance from a measuring sample (printing ink) and from
a "white" (reference value).

With the 13 - value calculated above the ink density is:

53
53
There is a close correlation between ink film thickness and
ink density. The illustration shows that with the ink film
thickness increasing, light reflectance decreases and the ink
density value increases.

54
54
The diagram illustrates the correlation between ink film
thickness and ink density for the four process colours in
offset printing.

The vertical line marks the ink film thickness range of about
1 pm customarily used in offset printing. The diagram also
shows that the density curves do not start to flatten off until
significantly higher ink film thicknesses are reached.
From these ink film thicknesses upwards there is hardly any
further increase in the ink density; even if the measurement
were performed in a full ink container, the density value
would not be higher. However, these ink film thicknesses
are no longer relevant for offset printing.

55
55
3.4 Measurement

3.4.1 Zeroing on paper white

Before measurement is started, densitometers must be cali-

brated to zero against the paper white (reference white) of
the printing stock in order to eliminate the influences of
paper colouring and surface characteristics on the evalua-
tion of the printed ink film thickness.

For this purpose, the density of the paper white in relation

to "absolute white" is measured, and this figure is set to zero
(reading D = 0.00).

3.4.2 Solid density

The readings of a solid area, are referred to as solid density

(DV). It is measured on a print control strip, which is printed
on the sheet at right angles to the print direction.
Besides other control elements the print control strip con-
tains solid fields for all four process colours and, if
necessary, for additional colours.

The solid density value allows a regular ink film thickness

to be checked and maintained (within a certain tolerance)
throughout the whole sheet width and print-run.

3.4.3 Halftone density

The halftone density is measured in the halftone patches of

the print control strip. In the measurement spot of three to
four millimeters, a combination of dots and paper white is
included, similar to that seen by the human eye.

The measurement value is the ink density in halftone (DR). The

larger the ratio of the area of dots to the total area on a meas-
ured surface and the higher the ink film thickness of a given
printing ink, the higher is the measured halftone density value.

56
3.4.4 Optically effective area coverage (halftone value in print)

When screens are measured with a densitometer, it is not

the geometrical area coverage, i.e. the area ratio between
dots and paper white on the measuring spot, but the "opti-
cally effective area coverage" which is measured.

The difference between geometric and optically effective

area coverage is due to the fact that both in visual observa-
tion and in densitometric measurements, part of the arriving
light penetrates into the paper between the dots at the
unprinted points, but is trapped under the dots during
reflection and thus absorbed.

This effect is called "light gathering". It causes the dots to

appear optically larger than they are in reality. The optically
effective area coverage is composed of the geometric area
coverage plus the optical gain of the area:

57
57
3.5 Evaluation

From the measurement values of solid and halftone densi-

ties halftone value, dot gain and contrast can be calculated.
First, however, all measuring devices must be calibrated to
zero against the paper white.

3.5.1 Halftone value in print

Given the DV and DR readings, the halftone value in print FD

can be calculated using the Murray-Davies formula.

3.5.2 Dot gain

The dot gain Z(%) is obtained from the difference between

the measured halftone value in print FD and the known
halftone value in film FF.

3.5.3 Contrast

The relative print contrast is also calculated from the read-

ings of the solid ink density DV and the screen ink density
DR. The DR value here is best measured in the three-quarter
tone.

58
3.5A Ink trapping

The ink trapping is calculated from the solid density values for
every individual colour in the solid fields, for all two-colour
superimpositions and for the three-colour superimposition in
the solid superimposition fields of the print control strip in
accordance with the colour sequence involved.

The ink trapping calculated with the following formulae

indicates which percentage of an ink is superimposed onto
another. The value is given relative to that of an isolated ink
printed on the paper whose trapping is set to 100%.

3.5.4.1 Two-colour superimposition

mitt

D1+2 is the ink density for the superimposition of both

colours,
D, is the ink density of the colour printed first and
D2 is the ink density of the colour printed last.

N.B.: All ink densities must be measured with a filter

which is complementary to the second colour.

3.5.4.2 Three colour superimposition

-

where
D1+2+3 is the ink density for the superimposition of all
the three colours and
D3 is the ink density of the colour printed last.

N.B.: All ink densities have to be measured with a filter

which is complementary to the third colour.

59
The given formulae are also used in the Heidelberg CPC 21
Quality Control. In addition, there are other methods of
calculating the ink trapping. All these methods are contro-
versial and, for this reason the values obtained should not
be interpreted too stringently. However, for a comparison
from run to run, and especially within the same run, they
are in fact meaningful. The higher the FA value, the better is
the ink trapping performance.

3.6 Standardisation in printing

In offset printing there are many steps between the typon

and the final print product e.g. reproduction (the making of
progressives), proof, plate exposure and print runs.
In each of these processing steps the size of the elements of
the image changes: halftone dots become larger or smaller,
lines become thinner or thicker.

The typical performance for each of these process steps can

be described by transfer characteristics, the most common
of which are the plate exposure characteristic and the print
characteristic.

The whole reproduction process aims at making the print

look like the typon. At the pre-print phase all the transfer
characteristics must be known. It is only then that variations
of the elements of the printed image that result from the
characteristics of the process can be compensated.
For reasons of economic efficiency, however, this is only
possible if the number of transfer characteristics is low.

60
Standardisation in printing therefore aims at defining only a
small number of transfer characteristics along with their
tolerances in order to obtain low-cost and high-quality
repros without having to take into account the properties of
individual plate-exposure devices or printing presses.

All process steps must aim at this goal, and their constancy
must be continuously monitored. Print control strips, plate
exposure control fields and, in particular, colorimeters at the
printing press are valuable tools in achieving this goal.

3.6.1 Standardisation systems

There are various standardisation systems. Yet all of them
aim at the same goal: producing cost-effective prints of con-
stant high quality.

Instructions for standardisation in printing are available

from various research institutes and suppliers. As an exam-
ple, the reader is referred to the standardisation guidelines
compiled by the FOGRA, the German Research Society for
Printing and Reproduction Technology, on behalf of the
Bundesverband Druck BVD (Association of the German
Printing Industry).

This concept is described in great detail in the illustrated

publication "Manual for Standardisation of the Offset Print-
ing Process — Instructions for "Platemaking and Printing".
This publication (in an A 4 folder) and a video cassette with
the same title are available from the Bundesverband Druck
e. V. in Wiesbaden in English and German.

61
61
3.7 Limits of densitometry

Like the colour separation technique, densitometers work

with filters tuned to the four process colours. They provide a
relative value for the ink film thickness, i.e. they do not
measure the optical appearance of the colour.

62
62
This fact sets certain limits to their application. The table
lists the typical fields of application compared to the tristim
ulus colorimeter and spectrophotometers.

One essential disadvantage of densitometry is that the same

colour densities do not necessarily lead to the same optical
impression. This is the case when the colour substances to
be compared differ from each other. Therefore setpoint
values can not be taken from proofs or other samples.

The restriction to the three colour filters for red, green and
blue is of similar importance. When colour sets are com-
posed of more than the four process colours, the measuring
of the additional colours becomes problematic. In most cases
there are no appropriate filters for the additional colours, as
a result of which the values measured for ink density are too
low and those for dot gain are incorrect.

The use of densitometers is also critical for colour control on

the basis of multicolour halftone patches such as gray
balance patches. If a gray balance patch is measured with
the three-colour filters the resulting ink density values are
different than the values that result when each of the colours
is measured by itself. This is because each of the three print-
ing inks contributes to all ink densities. The reason for this
is that the process colours are not perfect two-third inks and
thus also absorb light from other spectral ranges.

Densitometers are useful in monitoring the print-run of a

four-colour print. In all other cases densitometers are of
limited use.

63
The following two examples illustrate how additional colours
are measured with a densitometer.

The hue (light beige) shown here — has a relatively high

reflectance, slightly decreasing in the blue range
(380 to 500 nm). Accordingly, the highest density value
(0.17) is measured with a blue filter. This low value cannot
be altered easily since changes in ink film thickness only
lead to insignificant shifts in the density. In practice light
pastel colours are therefore assessed visually on the basis of
an OK-sheet and corrected manually.

6464
The additional colours HKS 8 and HKS 65 shown in the
second example have completely different hues as can be
seen from their reflectance curves. For both colours the
absorption within the blue range (380 to 500 nm) is greatest
As a consequence, the highest densities (1.60 for either
colour) are measured by the blue filter. Equal density valu€
measured by the same filter thus do not necessarily mean
that the hues are equal!

The appearance of a colour can thus only be evaluated

colorimetrically.

65
65
66
Colorimetrics 4

4 Colorimetrics
4.1 Measuring colour 68
4.2 Tristimulus values/white reference 70
4.3 Standard illuminants 70
4.4 Standard observer/colour matching functions 72
4.5 Evaluation with a spectrophotometer 74
4.6 Colour difference AE 76
4.6.1 CIELAB 79
4.6.2 CIELUV 82
4.6.3 CIELCH 83
4.6.4 CMC 84
4.7 Munsell 86
4.8 The tristimulus method 87
4.9 Spectral colour measurement 88
4.10 The measuring principle of the Heidelberg
CPC 21 spectral quality control 90
4.11 Proof and colour control strips 91
4.11.1 Proof control strips 91
4.11.2 Colour control strips 92
4.12 Ink control with Heidelberg CPC 21 93
4.12.1 Colorimetric control by gray fields 93
4.12.2 Colorimetric control by solid fields 95
4.12.3 Densitometric control by solid fields 97
4.13 Advantages of colorimetrics for offset printing 98

67
67
As described in the chapter "Systems of colour classifi-
cation", three numbers are needed to unambiguously define
a colour. Colorimetrics describes how these figures are
determined and how they relate to each other.
One prerequisite is, however, that colours are measurable.
Thus colour measuring and colorimetrics are directly
connected with each other.

4.1 Measuring colour

Colours are measured with tristimulus colorimeters or

spectrophotometers. Their function is described in Chapters
4.8 and 4.9.
In principle, the construction of colour measuring devices
follows the visual and sensoric model of the human eye (see
illustration).

The ink (sample) is illuminated by a light source (radiation).

Part of the light is absorbed by the sample, the rest is reflec-
ted. The reflected light is captured by the human eye. There
the red, green and blue sensitive cones (visual receptors) are
stimulated. Via the optic nerve, this stimulation will trigger
off the perception of colour in our brain.

This natural process is imitated in the measuring device.

In the measuring process, the light is sent onto the printed
sample. The reflected light passes through a lens system and
to a sensor, which measures the intensity of the incident
light for each colour and transmits the measured readings to
a computer. There they are weighted with functions that
imitate the weighting functions of the three types of sensitive
cones in the human eye, and which have been defined by
the CIE for the standard observer. The result is the tristim-
ulus values X, Y and Z. These are finally converted into
chromaticity coordinates or coordinates of other colour
spaces (such as CIELAB or CIELUV).

68
68
man measuring
device

69
69
4.2 Tristimulus values/white reference

In the measuring of colours, the identification of tristimulus

values from measured reflections and emissions requires
standardised conditions. Most of them have been laid down
by the manufacturer of the equipment and have been taken
care of in such a way that the user need not pay further
attention to them. In the measuring of body colours,
however, three factors are usually variable and must be
adjusted by the user: the white reference, the type of light
and the observer.

Normally, colorimetric values are determined relative to

"absolute white". The calibration is thus set to the cali-
bration standard of the measuring unit, which is in turn
calibrated against a theoretical absolute white. In contrast to
densitometry, the paper is used as a reference only in excep-
tional cases.

4.3 Standard illuminants

Without light there is no colour. But this also means that the
type of light influences our colour perception. The colour of
light is determined by its spectral composition.

In natural sunlight, the weather as well as the season and

time of day influence the spectral composition. Photogra-
phers and film directors often have to wait a long time until
the lighting conditions are the way they want them to be.

Likewise, there are differences in the spectral composition of

artificial lamp light. Some lamps produce reddish light
whereas others emit a greenish or bluish light.

Spectral reflection and consequently colour perception

change depending on the light conditions. Tristimulus values
must, therefore, be based on standard light.

I 70
70
In standardisation, the intensity distribution has been laid
down for different types of light in the range between 380
and 780 nm (at intervals of 5 nm). The illustration shows the
spectral distributions for the standard illuminants A, C, D50
and D65.

The C, D50 and D65 standard illuminants are similar to aver-

age daylight with a peak radiation intensity in the blue area.
The illustration shows the composition of the type D65 illu-
minant. An A standard illuminant has a peak intensity in the
red area; it thus appears reddish (evening light and electric
light).

71
 4.4 Standard observer/colour matching functions

Each individual has three colour matching functions to

evaluate red, green and blue. In the case of persons with
normal chromatic vision, they will be almost identical.
Thus colours are perceived differently only in boundary
areas. For example, what one individual still perceives as a
bluish green, another will see as a greenish blue.

That is why it was necessary to define, for colorimetric

purposes, an individual with average visual perception, i.e.
the "standard observer". A comprehensive series of tests
with a large number of persons with normal chromatic vision
was carried out in 1931. On the basis of these tests, the
colour matching functions x, y, and z were defined and laid
down as binding by the CIE both in national and inter-
national standards such as DIN 5033 and ISO/DC 12 647.

2.0 2.0

1.5 1.5

1.0 1.0

0.5 0.5

0.0 0.0
400 500 600 700 nm 400 500 600 700 nm

72
The study was carried out for an observer angle of 2°.
The observer angle in the sense of the standards of colori-
metrics is the visual angle at which a colour area is viewed
(see illustration). For example, if an area with a diameter of
3.5 cm is viewed at a distance of 1 m, the visual angle will
be exactly 2°.

In 1964, the same test was repeated for an observer angle of

10°, and the results were likewise laid down in a supple-
mentary standard. Hence the "standard observer 1964"
came into being.

1m

2°A 3.5 cm
400 500 600 700 nm 10°A 17.5 cm

73
73
4.5 Evaluation with a spectrophotometer

The standard colour values are calculated from the radiation

function of the illuminant S(X), the measured degree of spec-
tral reflection of the sample (X) as well as the colour
matching functions 3Z (X), y (X) and 7 (X) of the standard
observer.

The lambda in brackets (X) shows that the calculation

depends on the wavelength X, of the light (e.g. in a wave-
length range between 400 and 700 nm at intervals of 5 nm).
In the first step of the calculations, the values of the radia-
tion function of the standard illuminant S( ) are multiplied
by the measured degrees of reflection 13(k) of the sample for
each wavelength X (i.e. for each spectral colour of a specific
type of light). The result is a new curve, the colour stimulus
function p(X ).

In a second step the values of the colour stimulus function

are multiplied by the colour matching functions K (X), y (X)
and 7 (X). This results in three new curves.

Finally, by integrating and multiplying with a normalisation

factor, the tristimulus values X, Y and Z are calculated from
the areas under these curves by integration, which makes it
possible to exactly describe the measured colour.

74
74
X

75
4.6 Colour difference AE

The colour difference is a measure of the distance between

two colour locations in the colour space (e.g. between origi-
nal and printed sheet).

In the chapter "Systems of colour classification", the CIE

colour space was explained. But this colour space has one
major disadvantage: Not for all colours does the human eye
perceive colour location differences of the same value as
being identical.

MacAdam, an American, studied this fact in a long test

series, analyzing and illustrating the results. The illustration
shows the so-called MacAdam ellipses in tenfold enlarge-
ment. Since the CIE colour space is three-dimensional, the
ellipses are really ellipsoids, i.e. ellipse-shaped three-dimen-

76
sional bodies. The size of the ellipsoids is a measure for the
perception threshold of colour deviations (each viewed from
the center of the ellipsoid and for the individual hue).

This system is therefore of no practical use in the evaluation

of colour differences, since it implies that the acceptable
tolerances are different for each hue. For reliable and
powerful calculations of colour differences, a colour space is
needed in which colour differences that are perceived as
being identical have the same numerical values. CIELAB and
CIELUV are two such systems; they were developed by
mathematical transformation from the CIE colour space.

Through this transformation, the MacAdam ellipsoids of

varying size were mapped on spheres of almost identical
size. In this way, the human eye perceives identical colour
differences for all colours as almost identical.

In 1976 the CIELAB and CIELUV colour spaces, the most

commonly used in the printing industry, were standardised
on an international basis.

77
77 I
 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 - X

The illustration shows the location of the a*- and b*-axes of

the CIELAB colour space in the x-y colour table.

Other colour spaces such as the CMC system and the

Munsell colour space, are also used in the U.S.A.

78
78
 L*
4.6.1 CIELAB

The CIELAB colour space is most frequently used for

measuring body colours (printing inks), for example, in
preparing ink formulae or for quality control in
printing. Colour tone and colour saturation are
drawn on the a* and b*axes. The a*-axis runs from
—a* (green) to +a* (red), the b* axis from —b* (blue)
to +b* (yellow). The lightness axis L* runs from 0
(black, at the bottom) to 100 (white, at the top).

The illustration shows the CIELAB colour space for

body colours. Since it is the result of a transforma-
tion, its shape is different from that of the CIE —a
colour space. Likewise, the shape of the individual
lightness levels changes with L*.

h*

—I.,

In the illustration, a cross section through the CIELAB colour

space is shown for body colours at a lightness level of
L* = 50. The scaled down green area and the enlarged blue
area can clearly be seen.

79
L*= 75.3 means that it is a bright colour the location of
which is between yellow and red with a* = 51.2 and
b* = 48.4. The present example is therefore a bright yellow-
red or orange.

Result: The pre-set reference location of the colour and

the measured location of the colour differ.

I 80
80
In accordance with their perceptibility, the colour location
differences can be classified as follows:

Since the transformation used is not linear, the regularities

of the CM colour space cannot readily be applied to the
CIELAB colour space. One argument in favor of its use is its
worldwide application.

81
81
4.6.2 CIELUV

The CIELUV colour space was also derived through transfor-

mation from the CIE colour space but by using other
formulae than for the former. The three coordinate axes
are denoted by L*, u* and v*.
Since the CIELUV and CIELAB colour spaces are
the result of different transformations, they also
differ in shape. Both are used for body colours (see
illustration).

The illustration shows a cross section through the CIELUV

colour space for body colours at a lightness of L* = 50 (see
also Chapter 4.6.1). The green area in the CIELUV colour
space is located closer to the center than in the CIELAB
colour space; furthermore, the blue area is larger.

I 82
82
The CIELUV colour space is often used for the evaluation of
colours on colour monitors (e.g. on scanners or computers).
Its advantage lies in the linearity of transformation so that
all the regularities of the CIE colour space remain unchanged.
(This is not the case with the CIELAB colour space.)

4.6.3 CIELCH

The term CIELCH is used when the polar coordinates C

(distance from the center) and h (angle) are used instead of
the Cartesian coordinates a, b or u, v in the CIELAB or
CIELUV colour space. CIELCH is thus not an additional
colour space.

For CIELUV, the calculation is the same.

Here a schematic representation with the same measured

colour location as in Chapter 4.6.1.

83
4.6.4 CMC

CMC, an evaluation of the colour location differences based

on the CIELAB colour space, was developed in Great Britain
in 1988 by The Colour Measurement Committee of the
Society of Dyers and Colourists (CMC). It does not describe
the perception of colour deviations (as CIELAB or CIELUV
do), but their acceptance by the observer.

In general, colour deviations near the lightness axis are

perceived to be much more disturbing than those in satu-
rated colours. Likewise, deviations in chroma (saturation)
are more readily tolerated than in the hue angle.

The illustration shows the principle underlying the CMC

evaluation of colour location differences in the CIELAB
colour space. Each ellipse shows the locations with constant
colour location differences in accordance with the CMC
formula with respect to the centre of the circle (reference
colour location). It can be clearly seen that the ellipses (toler-
ance ranges in the CMC colour space) are smaller in the
achromatic area than in the region of higher saturation.
In addition their shape is such that the admissible deviations
in the hue angle are smaller than in the chroma (saturation).
The ellipses also make it possible to individually adjust the
evaluation of deviations in lightness and hue. This adjust-
ment is made by means of the two weighting factors land c.
(1 is the lightness weighting factor; the weighting factor c for
the hue is as a rule equal to 1.) The textile industry often
operates with weighting factor ratio of 1 : c = 2 : 1, which
means that deviations in lightness will be twice as accept-
able as deviations in colour hue.

84
84
This ratio can be adapted to the needs of the application in
question. As a consequence, the values for colour location
differences are significant and comparable only in connec-
tion with the weighting factors.

85
85
4.7 Munsell

Munsell developed a system of colour classification with

equidistant location differences as far back as 1905. In this
system, the colours are arranged according to hue, bright-
ness and chroma. The basic hues are red, yellow, green,
blue and purple. The system was published in 1915 as the
"Munsell Book of Colour" for 40 hues, light type C and
brilliant and matt specimens.

The five basic colour hues were subdivided into 100 hues (
even numbers each having 16 chroma and 10 brightness
levels. The illustration shows a cross section of the colour
body for 40 hues. The result is an irregular colour body
since for some colours and lightness values not all fields ar
covered.

The Munsell coordinates cannot be converted into CIE coor-

dinates.

Further systems of colour classification are the DIN colour

card (DIN 6164), the Natural Colour System (NCS), the OSA
system (of the Optical Society of America) and the RAL
Design-System (RAL-DS).

I 86
86
4,8 Tristimulus method

The construction of tristimulus colorimeters is similar to

that of densitometers. Instead of the three colour filters red,
green and blue and the visual filter, combinations of filters
are used which imitate the three colour matching functions
y and 7.

Tristimulus colorimeters, however, have a lower absolute

measuring precision than spectrophotometers since, as a
rule, neither can the colour matching functions be imitated
exactly nor is the required standard illuminant available.
They are suited, however, for determining colour differences
since in this case the absolute values do not necessarily have
to be precise.

In addition, tristimulus devices are considerably cheaper

than spectrophotometers.

The measuring field is illuminated with a Iamp whose spec-

tral composition is close to that of a standard illuminant.
In our example, cyan is to be measured.

The spectral reflection is measured by means of the three

different filter combinations, and the tristimulus value X is
measured behind filter (red), the tristimulus value Y behind
filter (green) and the tristimulus value Z behind filter (blue).

After the measurement the tristimulus values can be conver-

ted into a colour space (CIELAB or CIELUV) in which colour
differences are perceived to be equidistant.

87
87
I Measurement principle of the three-range

1 photometer

4.9 Spectral colour measurement

In the spectral measuring process the total visible spectrum

from 380 to 780 nm is measured. The light reflected from a
printing ink is separated into its spectral components by
means of a diffraction grate and measured by an array of
sensors.

Depending on the required accuracy, the identity of the

incoming light is measured in steps of one, five or ten nano-
meters. The tristimulus values X, Y and Z are calculated
from the measured reflections. For this, the colour matching

88
88

I
functions, are stored in the computer. Since these functions
need not be simulated by filters, the absolute accuracy of
spectrophotometers is very high. However, they are more
expensive than tristimulus colorimeters.

Apart from the high absolute accuracy, one major advantage

of spectral colour measuring is the fact that spectrophotome-
ters can read out the tristimulus values for practically all
standardised types of light and observers, if their values are
stored in the computer. Furthermore, they can calculate
colour densities for all filter standards.

So far spectral measuring has been applied most consist-

ently in the ink industry.

In ink grinding, ink manufacturers have to comply strictly

with given targets. This is very important in the case of
standardised inks (Euroscale), but also in the case of HKS
inks and all special grindings. In these cases, the specimen
is measured with a spectrophotometer, and the mixture
ratio for the printing ink is calculated on a personal com-
puter with an ink program.

Previously it was not possible to make optimum use of spec-

trophotometers in printing shops. They were expensive and
cumbersome, and it was not possible to use them directly for
process colours. They were therefore only used for the indi-
vidual measuring of special inks and the testing of materials
(such as printing stocks and inks). They were unimportant
for quality control in printing.

89
89
4.10 The measuring principle of the
Heidelberg CPC 21 spectral quality control

At DRUPA 1990, Heidelberg was the first and only manufac-

turer to present a spectral measuring unit for offset printing
directly linked with the offset printing machines via the CPC 1
automatic remote colour control: the measuring unit CPC 21.

During the measuring process, a measuring head scans the

print control strip, making a spectral measurement of all con-
trol elements. Alternatively, the standard illuminants A, C, D50
or D65 and the standard observers 1931 and 1964 can be used.

The measuring principle of CPC 21 is illustrated in the

diagram below.

First, the illuminant is directed to the printed sample via

ring catoptrics at an incidence angle of 45°. The reflected

90
light at an angle of 0° is directed via a deflection mirror and
a fiber-optical light guide from the measuring head to the
spectrophotometer. There it is split into its spectral colours
by means of a diffraction grating which has an effect similar
to that of a prism.

Photodiodes measure the radiation distribution in the entire

visible spectrum (between 380 and 730 nm) and send the
results to a computer. There the measured colour values are
evaluated colorimetrically; the result is given in the tristim-
ulus values X, Y and Z and the chromaticity coordinates
x, y and Y.

The values can be converted either to the CIELAB or to the

CIELUV colour spaces. After the measured values have been
compared to the previously set reference values (allowing for
the pre-set tolerance AE), the required modifications are
transmitted via CPC 1 to the ink ducts of the printing units,
where they are realised immediately.

4.11 Proof and colour control strips

4.11.1 Proof control strips

Off-press tonal proofs are increasingly being used instead of

press proofs. The reason is that an off-press tonal proof is
cheaper and can be produced more rapidly than a press
proof. There are different proofing methods, all of which
operate without offset printing ink. However, the colouring
materials of the proofs (i.e. the toners) and offset printing
inks differ in their composition.

Heidelberg has developed a special proof control strip (see

illustration).

91
91
It has solid patches of the colours black, cyan, magenta and
yellow, plus one halftone patch with 70% area coverage per
colour, ink trapping elements and a gray field consisting of
70% cyan, 60% magenta and 60% yellow. In addition there is
an interface element for the automatic reading of reference
values into CPC 21. When this element has been fitted, all the
measuring elements will be recorded. On this basis, the meas-
ured values can be stored as reference values.

4.11.2 Colour control strips

The colour control strip for spectral measuring with CPC 21

has also been developed by Heidelberg and contains (except
for the interface element) the same measuring elements as
the proof strip. Additionally, exposure measuring elements
for standardised plate exposure are available.

Heidelberg offers three different colour control strips:

The print control strip type 4 GS (Gray field and Solid con-
trol) for four printing inks, print control strip type 6 GS for
five and six printing inks and the print control strip type 8
GS for seven and eight printing inks.

The data of these colour control strips and the print control
strips of the older densitometric measuring unit CPC 2-01
are stored in CPC 21.

The user may also input additional print control strips by hand.

92
92
4.12 Ink control with Heidelberg CPC 21

CPC 21 offers three types of ink control:

■ colorimetric control on the basis of the gray field,

■ colorimetric control on the basis of solid or halftone fields
and
■ densitometric control on the basis of solid or halftone
fields.

4.12.1 Colorimetric control by gray fields

As mentioned in Chapter 2.4, colour balance is a decisive

criterion for the optical impression of a printed image.
Faults in the colour balance are especially obvious in gray
fields. It therefore seems sensible to use gray fields for
measurement-based tuning as well as for monitoring and
controlling the print run stability. Colorimetrics is ideally
ited for this. The inks cyan, magenta and yellow should
t erefore be controlled colorimetrically on the basis of gray
f Ids (if possible with three-quarter tone patches).
As reference values, both in-house standards and values
from proof control strips may be used.

93
93
The illustration shows a monitor display of CPC 21. The loca-
tion of the reference colour is shown in the a-b plane at the
upper left. In our example it is located in the center, i.e. on the
gray axis. The center of the illustration shows an enlargement
around the reference colour location. The three circles mark the
boundaries of the three \6E-tolerance classes close, medium and
wide. The lightness axis is situated near the right border of the
screen, also with respect to the reference colour location.
Here, too, the three tolerances are marked.
Each cross marks a reading. In the example shown, the
measured colour locations of the colour zones deviate
towards yellow-green and brighter.

If the deviations are larger than the allowed \6E-tolerance ,

the processing unit will automatically compute the necessary
corrections for cyan, magenta and yellow. In addition to the
spectral readings of the gray field, those of the single colour
solid patches and halftone patches of cyan, magenta and
yellow as well as the spectral readings of solid superimposi-
tion patches are evaluated. In this way, all relevant factors
involved are taken into consideration.

94
94
Corrections in the press will automatically be made via the
CPC 1 press control unit.

4.12.2 Colorimetric control by solid fields

Colorimetric solid control is generally preferred for black

and for special colours.

Black primarily has an effect on brightness. Since the

human eye tends to tolerate deviations in brightness more
readily than chroma deviations, black can be controlled on
the basis of solid fields. Experience has shown that the influ-
ence of black on the colour balance is thus sufficiently
accounted for.
Additional colours are mostly printed as isolated and solid
areas. It is, therefore, sensible and correct to monitor them
on the basis of solid fields.

But in solid patch control, too, spectral measuring and

colorimetric evaluating have important advantages over
density control: it can be ascertained exactly whether a pre-
set hue has been reached. Furthermore, reference colour
locations can be entered either as numerical values or as
specimen measurements. This is not possible in colour den-
sity measuring.

CPC 21 indicates after the first measurement whether or not

the reference colour location can be reached with a specific
colour. If not, the expected colour location difference AEpossible

is output.

95
The illustration shows a CPC 21 monitor output in solid
measuring. At the upper left, the reference colour location in
the a-b plane is marked. The center of the enlargement in
the middle of the monitor marks the colour location with the
smallest possible colour location difference AEpossible
from the
reference colour location, i.e. the best colour location that
can be obtained with the selected colours.
The theoretical reference colour location is marked by a
circle (in the example below the red axis and at the right of
the yellow axis).

If there are deviations in excess of the allowed tolerance, the

necessary modifications will again be computed.
The spectral readings of the single-colour solid and single-
colour halftone patches are used in the calculation.

96
96
4.12.3 Densitometric control by solid fields
In addition to colorimetric data, spectrophotometers can
also establish density values for any colour filter. As an
aid to the user, Heidelberg CPC 21 therefore also provides
colour density values independent of the type of control.

Especially for repeat orders, for which colour density ref-

erence values have already been calculated, solid density
control may be an alternative.

97
4.13 Advantages of colorimetrics for offset printing
In conclusion, a survey of the essential advantages of
colorimetrics for offset printing:

The measurement readings match as closely as is

possible the subjective perception of colour.

Colorimetrics is a colour evaluation technique

which is independent of the printing process and
can be used throughout the printing process from the
pre-press stage via all proof stages and, finally, for
quality control.

Colorimetric reference values can also be given as

numerical values. An interface to pre-press units is
available.

colorimetrics reference values can be taken from

specimens.

only with colorimetrics is it possible to objectively

tune colours.

Colorimetrics makes image-related colour control

possible (e.g. by means of gray fields) without
colour-specific calibration procedures and without
stored values.

by means of Colorimetrics, all inks, even very

light special inks, can be controlled correctly and
consistently.

98
Dot gains are reliably detected by spectral colour
measurement even if special inks are used.

The production run control is safer since changes in

the printing stock, ink soilings and metamerism can
all be detected.

Halftone printings with more than four colours can

also be controlled reliably.

The printing quality can be defined and verified beter.

There is a measure for colour deviations independent
of the hue.

Spectral colour measurement makes the development

of better colour control models possible.

The printing industry will adapt to the colour measur

ing principle currently used in all colouring indus
tries.

Densitometry is an integral part of spectral colour

measurement.

The tendency toward the use of more than four inks

is accounted for.

Colorimetrics also makes it possible to objectively

compare parts of printed images with the originals.

99
100