Rochester Institute of Technology

RIT Scholar Works

Theses Thesis/Dissertation Collections

5-1-1989

A Study on doubling in the offset printing process

Amal A. Ba'adarani

Follow this and additional works at: http://scholarworks.rit.edu/theses

Recommended Citation
Ba'adarani, Amal A., "A Study on doubling in the offset printing process" (1989). Thesis. Rochester Institute of Technology. Accessed
from

This Thesis is brought to you for free and open access by the Thesis/Dissertation Collections at RIT Scholar Works. It has been accepted for inclusion
in Theses by an authorized administrator of RIT Scholar Works. For more information, please contact ritscholarworks@rit.edu.
 ROCHESTER INSTITUTE OF TECHNOLOGY

A STUDY ON DOUBLING IN THE OFFSET PRINTING PROCESS

by

AMAL A. BA'ADARANI

A Thesis submitted in partial fulfillment of the requirements of

the degree of Master of Science in the SchoolPrinting and Management Sciences

of

in the College of Graphic Arts and Photography of the

Rochester Institute of Technology

May 1989

Thesis Advisor: Mr. Franz Sigg

 School of Printing
Rochester Institute of Technology
Rochester, New York

CERTIFICATE OF APPROVAL

MASTER'S THESIS

This is to certify that the Master's Thesis of

AMAL AHMAD BA'ADARANI

with a major in Printing Technology has been

approved by the Thesis Committee as satisfactory
for the thesis requirement for the Master of Science
Degree at the convocation of May, 1989.

Thesis Committee:

Franz Sigg
Thesis Advisor

Cliff Frazier
Thesi'S Co-Advir,~r

Joseph L. Noga
6raduat\1 Program Coordinator

Miles Southworth
Director Or Designate
 ACKNOWLEDGMENTS

I would like to thank the following for their time, help and support:

Mr. Franz Sigg of the Technical and Education Center of the Graphic Arts for technical

advice and help in experimental planning; Professors Cliff Frazier and Charles Layne; Mr.

William Eisner of the Technical and Education Center of the Graphic Arts for donation of

materials; Mr. Dave Conn of the Technical and Education Center of the Graphic Arts for

allowing the use of his lab and equipment; Ms. Marie Freckleton for her support and use

of her equipment and the faculty in the School of Printing Management and Sciences.
 TABLE OF CONTENTS

LIST OF TABLES ii

LIST OF FIGURES iii

ABSTRACT 1

CHAPTER ONE. INTRODUCTION 2

FOOTNOTES 7

CHAPTER TWO. THEORETICAL BASIS OF DIRECTIONAL DOT GAIN 8

FOOTNOTES 15

CHAPTER THREE. REVIEW OF THE LITERATURE 17

FOOTNOTES 21

CHAPTER FOUR. HYPOTHESES 22

CHAPTER FIVE. DESIGN OF THE EXPERIMENTAL RIT DOUBLING TEST TARGET 23

CHAPTER SIX. METHODOLOGY 29

FOOTNOTES 35

CHAPTER SEVEN. RESULTS 36

CHAPTER EIGHT. SUMMARY, CONCLUSIONS AND OBSERVATIONS 45

CHAPTER NINE. RECOMMENDATIONS FOR FURTHER STUDY 48

BIBLIOGRAPHY 50

APPENDIX A. CALCULATION OF THE CHANGE IN PRINT LENGTH

CAUSED BY A CHANGE IN PACKING 53

APPENDIX B. DERIVATION OF THE MURRAY DAVIES EQUATION 55

APPENDIX C. TABLE OF REGRESSION EQUATIONS AND ANALYSIS OF VARIANCE 57

APPENDIX D. PRESS RUN DOCUMENTATION 77

APPENDIX E. FILMS FOR LAYOUT 81

 LIST OF TABLES

TABLE PAGE

1 Percent Dot Area Increase due 1 Unit Various Angles 27

.
to Doubling of at

2. Table of r and t values for Magnitude of Doubling versus Doubling Angle 36

3 . Table of r and t values for Dot Gain on 50% tint versus Average
Dot Gain on Concentric Circles 43

4 . Mean Value and Standard Deviation for Magnitude of Doubling

and 50% Tint 43

5 . Table of r and t values for Magnitude of Doubling and Lateral-

Circumferential Register 44
 LIST OF FIGURES

FIGURE PAGE

1. Hierarchy of Dot Gain Classification 3

2 . Enlarged Dots Showing the Effects of No Dot Gain. Fill-in. Slur.

Fill-in and Slur 3

3. The Appearance of Doubling in Three Tones 5

4. The Effect of Light Scattering on Grained Aluminum and Opal Glass 6

5. The GATF Star Target 18

6. The GATF Dot Gain Scale and Slur Gauge 19

7. The GATF Ladder Target 20

8. The RIT Experimental Doubling Target 23

9. 50% Dot and line Tint 26

10. Plot of % Area Change due to Doubling Angle for a 50% Dot and Line Tint 27

11. Layout of Test Form used on Web-Press Run 32

12. Plot ofVertical and Horizontal Register on a Cartesian System

of Coordinates 32

13. Scatter Diagram and Straight Line Approximation between

Magnitude of Doubling and Doubling Angle for Black 1 and 2 38

14. Scatter Diagram and Straight Line Approximation between

Magnitude of Doubling and Doubling angle for Magenta 1 and 2 39

15. Polar Graphs of Doubling for Black and Magenta 1 and 2 40

1 6. Scatter Diagram and Straight Line Approximation between Dot Gain

on 50% Tint and Average Dot Gain on Concentric Circles

for Black 1 and 2 41

17. Scatter Diagram and Straight Line Approximation between Dot Gain
on 50% Tint and Average Dot Gain on Concentric Circles

for Magenta 1 and 2 42

 ABSTRACT

The purpose of this study was to test the usefulness of the Experimental RIT

Doubling Target in identifying directional dot gain in the offset printing process.

Doubling, a form of mechanical dot gain, contributes greatly to color variations while

printing. Slur, another form of mechanical dot gain, is often confused with doubling. Dot
Gain Test Targets currently in use often do not distinguish between the two. A slur test

was carried out and the results showed that slur is not a contributor to directional dot gain.

The target proved to be a visual and quantitative measure of Directional and

Non-directional dot gain. The concentric circles, one of the main components of the target,
were made with the intent of matching non-directional dot gain similar to that of a 50% tint

made with a 150 line/inch screen-ruling. A slur test was carried out and the results

showed that slur is not a major contributor to directional dot gain. The values for dot gain

obtained on the concentric circles were found to be higher in value than the values for dot

gain obtained on the 50% tint. Nonetheless, there was a strong correlation between the

two values of dot gain on both targets. Direction, another aspect of directional dot gain,

was studied in relation to the doubling magnitude. It was found that a definite preferred

angle for doubling existed. The Pearson Product Method of Analysis, the t test for

statistical significance and graphing were used to analyze the data.

 CHAPTER ONE

INTRODUCTION

The offset lithographic method is one of the major printing processes in use. The

term offset denotes a process where the ink is not transferred directly from the plate to the

paper (as is characteristic of direct lithography) but to another intermediate cylinder covered

with a rubber blanket and from there to the substrate. As a printing process, offset

lithography is capable of excellent print quality, both in fine line and tone work at a lower

cost than other printing processes. This is primarily due to the fact that the rubber blanket

can transfer the ink to rough surfaces so that higher quality printing can be done on them.

However, offset printing is a delicate process. Inspection of the production run is

mandatory to maintain stringent quality requirements. Traditionally, some of the different

quality criteria that have been observed are solid ink density, tone reproduction (dot gain),

trapping, slur and doubling, register and color balance

To measure and control print quality, it is best to measure and control each

separate component which contributes to print quality and its variation. In evaluating a

press sheet, we look at the print quality of halftones and that is observed by noticing the
feature of how sharp and clear the detail is. One fault that is either unique to halftones or

most troublesome when halftones are being printed is dot gain. Dot gain is the increase

which takes place in a dot's area during the transfer stages from film to the printed sheet.

Dot gain in itself is not regarded as a printing fault since a certain amount of inkspread,
when transferring from blanket to paper, is
unavoidable.1

There are, however, different

kinds of dot gain with distinct features and causes. Some can be compensated for in the

prepress area; others cannot. It is important to distinguish between the various types of dot

gain because they call for different remedies. The following is a hierarchy of dot gain
classification (figure 1):
 Total Dot Gain

Mechanical Dot Gain Optical Dot Gain

Non Directional Directional

Fill-in Slur
Doubling

Figure 1 .
Hierarchy of Dot Gain Classification

Mechanical Dot Gain: There are basically two kinds of mechanical dot gain:

non-directional and directional.

Non-Directional Dot Gain is called fill-in (figure 2). The halftone dots suffer a

general increase in size in that the dots get uniformly larger due to spreading of ink over the

edge of each dot. This can be caused by too much pressure between the blanket and

impression cylinder, or by printing with too much ink (also taking into consideration the

rheological properties of the ink) and possibly by the characteristics of the blanket (whether

conventional or compressible).

Fill-in
No Dot Gain

Slur
Fill-in and Slur

Figure 2. Enlarged Dots Showing the Effects of No Dot Gain. Fill-in. Slur. Fill-in and Slur
 Directional Dot Gain occurs as slur and doubling. They are regarded as faults
and their occurrence and effect on the print is unpredictable. Slur is a directional increase
in dot size noticeable as an elongation or at the edge of a halftone dot. A
smearing trailing
distinct feature of slur is that it usually occurs in the direction of cylinder rotation on the

press (figure 2).

Slur reduces shadow contrast. Halftone highlight areas and fine lettering are

usually not
affected.2
A slurred impression will often be uniform from the front to the

back of the sheet. Mechanically, the major cause behind slur is attributed to a difference in
surface speed between the two printing cylinders (plate-blanket or the blanket-impression).
This difference in surface speed is caused by a difference in diameter of the plate, blanket

and/or impression cylinder. A slurred impression will show up as a result of excessive

impression pressure. It can also be caused by the ink acting as a lubricant in the heavy ink
slip in impression Other factors that
nip.3

coverage areas permitting the sheet to the

produce other variations of slur in the form of streaks are a loose or slipping blanket,

printing with a soft ink (not enough tack); or the use of too much ink especially when

printing on coated stock. Defective paper in the form of wrinkled or bulged paper can

result in slur. On a single color press, the defective paper will cause slur, but on a

multi-color press, it will show up as a double.

Doubling of a halftone dot refers to a weaker, or ghost dot whose position is out

dot.4
of register relative to the full strength true A double impression rarely prints at the

dot.5
same density as the original It does not occur as streaks or elongation like in slur.

In multi-color work, the blanket picks up a faint impression from the preceding sheet and

fails to transfer it in exact register to the next one being printed. If the transferred dots are

not in register with the true, full impression dots, doubling occurs (figure 3). Doubling
increases the highlight tone areas and gives muddy reproductions. What results is an

increase in the tonal values with an obvious distortion of the dots and a resulting loss of

sharpness (of the dots, not of the image).

 Doubling is an unpredictable and critical defect in offset printing. It is critical in
that it can appear at random and
usually can be corrected only by extensive mechanical

adjustments on the press. Mechanical inaccuracies such as press and the quality of the

paper are its primary causes. It is a problem of register variation between the printing

units of a multicolor press but it can conceivably also happen on a single color press due to

varying positioning of plate and blanket cylinders. It can also be caused by an unstable

automatic register system in the case of a web

press.6

On sheetfed presses, doubling

"slapping"
would sometimes occur along the back edge of a sheet of paper due to the paper

the blanket prematurely. A premature contact of the paper with the blanket unit produces a

weak transfer of ink from the blanket of dots that are out of register with the true dots

transferred to the paper during impression. Sheets with areas that are not flat (wavy or
tight-edged sheets) and static electricity can cause premature contact of the paper and

blanket. This premature contact will cause the paper to prematurely touch the surface of

the blanket causing a faint impression next to the full impression dot. Paper slippage in

the grippers, and excessive wear in the gear train or bearings of the press are other possible

causes of doubling.

5% 1 5% 45%

Figure 3. The Appearance of Doubling in Three Tones

Optical Dot Gain is the apparent expansion of the dot when light is reflected from the

paper through the ink. The dots behave as if they are larger than they really are. For

example a 40% dot pattern does not absorb 40% (reflect 60%) of the incident light. In

fact, less than 60% is reflected suggesting that a 40% dot is behaving apparently like a

larger dot size. This is basically due to light scattering within the paper. Two factors tend
to enhance this effect: of the substrate printed upon.
screen-ruling and surface translucency
These two factors exaggerate the effect of light scattering once tight enters the substrate.

Dots made with fine screen-ruling suffer more light scattering than dots made with coarse

screen-ruling.

Grained Aluminum Opal Glass

Figure 4. The Effect of Light Scattering on Grained Aluminum and Opal Glass.

Figure (4) shows a comparison of two kinds of surfaces that react differently in

terms of light scattering. A dot on an aluminum surface would have a sharp edge with no

spreading of light. Aluminum is not a translucent surface, which means light cannot

diffuse and scatter within aluminum. A 50% dot would absorb 50% of the light and reflect

50%. A dot on a translucent surface such as opal glass would behave differently. The dot

pattern is completely diffused before it emerges from the surface. If actual measurements

were taken with a densitometer for this 50% dot pattern, it would read a reflection density
of 0.6 instead of 0.3 representing an increase in reflection due to the effect of light
scattering.7

In this chapter, the different types of dot gain have been defined and briefly
described. The following chapter will investigate the problem of doubling and attempt at

describing its major causes.

 7

FOOTNOTES

CHAPTER ONE

^outhworth, Miles F., Dot Gain Causes and Cures. Quality Control Scanner,
Vol.2 No.9 1982 pp.1

2Anon., Make the Halftone fit the Paper. Research Progress Report No. 15,
September -

October 1949, Graphic Atrts Technical Foundation, Pittsburgh Pa.

3IbM,

4Bureau, William H., What the Printer should know about Paper. Graphic Arts

Technical Foundation, pp.190, 215.

5Treff, Ernie H., An Engineer's View of Dot Gain on a Web Offset Press. A

Presentation made at the PIA Annual Meeting in San Francisco on May 10, 1988,
p. 11

6Ibid.

7J. A. C. Yule and W. J. Nielsen, The Penetration of Light into Paper and its

Effect on Halftone Reproduction. TAGA Proceedings 1951, pp. 69-70

 CHAPTER TWO

THEORETICAL BASIS OF DIRECTIONAL DOT GAIN

Slur

There are two kinds of directional dot gain: Slur and Doubling. Slur is

theoretically caused when the surface speed of the plate, blanket and impression cylinders

are not the same. If the diameters should differ either by overpacking or underpacking the
plate and blanket, the surface speed of the cylinders will
vary and cause blanket slip.1
A
change in packing will be needed in order to correct for this problem. An advantage, if it
can be described as such, is that packing has to be quite different before slur is detected
and even if it were, it would not be considered much of a problem as long as it was

consistent throughout the

process.2
The first hypothesis was formulated in an attempt to

investigate the importance of slur (see page 24).

Theoretical Analysis of the Effect of Slur

The effect of packing on slur was theoretically analyzed with the use of an

equation that calculates the change in print length as a result of a change in packing. This
equation can be used to check for an increase in dot area of a 50% square dot of 150 line

screen ruling. The 50% square dot had a side length of 120 microns. The following
3 D=
formula 2PrtC / 100 was used for calculating change in print length resulting from a

change in packing where D denotes change in print length; C is for % circumference

occupied by image; P is for packing change.

Upon obtaining a value for the change in print length, a ratio of increase in print

length to the 50% dot side length of 120 microns was used to show the resulting change in

size as a result of this change in packing. The overall increase in dot size was less than a

micron or approximately .3 microns (see Appendix A for actual calculations). As a rule of

thumb, a 1 micron change in diameter of a 50% dot of a 150 line screen causes a change in

dot area of almost 1% dot area. Since the slur does not occur all around the dot, the

increase in the dot size would be much smaller. This is due to the fact that it is occurring in

only one direction. This change was small enough to warrant it being ignored. Even if
there was a chance that slur dot is totally The
actually occurred, its effect on the negligible.

effect on print length is important because it would ultimately affect register. This,
however is not an issue of concern in this study. This only goes to support the contention

that slur is not a fault in printing and consequently in directional dot gain. The only

problem of directional dot gain that is significant in practical work is doubling.

Doubling
Doubling caused by inaccuracies on the press can best be understood by
identifying the types of presses involved. Presses are classified as sheet-
fed and web-fed

according to the form in which the paper is fed into and through the press. On a sheetfed

press, sheets of paper are fed into the press one at a time, the impression is made and each

sheet is removed or delivered into a pile. On a web press, the paper is fed from a roll and

printing is continuous as the paper passes between the impression cylinder and blanket

cylinder. It is important to understand how doubling is caused on each of these presses

and to understand the variables that interplay to create the phenomenon.

Primarily, a double exists whenever there is a problem in register (however, this is

"ghost"

not a misregister between 2 colors, but between a color and its image). Doubling
caused by misregister within a single unit can be a result of variation in the relative position

of the plate and blanket from impression to impression. There are theories that indicate that

sideways play of the plate cylinder in addition to printing with a loose blanket can result in

a variation in the relative position of the plate causing doubling.

The easiest way to depict how doubling occurs is usually on a multi-color press.

The actual misregister occurs as follows: The press sheet enters the first unit of the press

and the first layer of ink is laid down. The press sheet then moves to the second unit of the

press and the still wet ink of the first unit on the paper, transfers to the blanket of the

second unit. When it is time for the next sheet to enter the second printing unit, this ink

will be transferred back from the second blanket to the second sheet. This secondary

transfer of ink is commonly known as print-back. It is important to understand that this

problem of misregister of backtrapping is within a single color. If the two sheets are not

in perfect register, a double image occurs, a strong one from the first unit and a weaker one
 10

from the second unit: the end result is that the dot increases in size color and
affecting
contrast.

The Effect of Paper on Doubling

Paper is a major source of
variability that is responsible for doubling.
Paper-related problems that promote
doubling on a sheetfed press are in the form of paper

wrinkling. Wrinkled paper is paper that lacks flatness. Paper wrinkling can be caused by
an imbalance of moisture in the paper. An example would be that of paper brought into the

pressroom from a storage area

having a different temperature and/or humidity level than

that of the pressroom. Paper is usually susceptible to wrinkling when unwrapped and

unprotected. The paper quickly releases or absorbs moisture as it adjusts to the new

environment. This moisture adjustment takes place primarily around the edges of the

unwrapped paper. This creates internal stress in the sheets that almost guarantees paper

wrinkling. Moisture imbalance occurs between the protected areas deep inside the stack of

paper and the unprotected outside edges of the sheets. Depending on whether paper is

losing or gaining moisture, a tight-edged or wavy-edged condition develops in the sheet.4

A tight-edged or wavy edged paper does not lie flat on any surface. This affects the way
the paper will travel on the press. Sometimes a slight bulge or wave in the paper would

cause the paper to touch the blanket prematurely resulting in a faint impression alongside

the main impression dot This is another cause of doubling.

Paper-related problems that promote doubling on a web press can best be

understood by studying the paper travel. Paper does not flow but is drawn in the case of a

web press. This creates a force in the web known as web tension. Controlling tension on

a press is actually controlling conditions under which the paper is drawn through the press.

This helps to minimize variability in flow behaviour in order to help maintain

register.5

Part of this control takes place at the infeed of a web press. The infeed is a section that

extends from the roll of paper to the first printing unit. It contains a roll stand and a series

of rollers that lead the web into the first unit. The infeed controls the speed, the tension

and lateral position of the web before it reaches the first printing unit. For example, a
 11

poorly set-up roll in the infeed can have length

consequences along the entire of the press.

Any sideways movement of the web (web weave) can cause a register change from one

revolution to the next. This effect is circumferential. Web is

weave considered as the

source of the largest doubles observed in web-offset

machinery.6

In addition, rollers in
the infeed have a tendency to glaze especially with the smooth steel rollers. The urethane

rollers suffer the same displacement as that which occurs between the plate-blanket nip
resulting in tension variations. Web tension has to be held constant and that can only occur

if the steel rollers remain clean. Failure to do in dust both

so will result paper build-up on

the steel and urethane rollers. This will be seen in color variation on the paper due to the

mechanical shifts.

Tension in a web is also affected by temperature fluctuations in the dryer.

Tightening and loosening of the web as the temperature varies will consequently change
register and cause doubling. Water pick-up by the web is another cause for tension

variation between the units. This tension variation can also promote doubling.

Press Characteristics that Affect Doubling

Web and sheetfed presses share certain components in terms of press design. One
of these components are bearers. Bearers are hardened steel rings found at the end of

printing cylinder bodies, and are used on both sheet and web presses. The bearers of plate

and blanket cylinders are in contact and supposedly facilitate smoother rolling between the
cylinders. Bearers found on either sheet or web-fed presses are thought by some to be

essential for high quality / high speed printing, but it is also true that many presses have

been made without them and run quite successfully on similar kinds of
work.7

External factors that further affect bearer performance can be the presence of a

gum coating on the surface of the bearers while printing. Gum coating on bearers is

build-up as a result of cleaning the plate. The eventual deposit of gum

caused by gum

coating on the bearers is hard and can cause enough pressure to cause a bump which forces

the cylinders out of parallel and could eventually flatten a spot on the blanket. Doubling
and Slur would result, depending on the job layout, and due to the absence of parallelism

Doubling or Slur would probably not show on all jobs, but

cylinders.8

between the up
 12

certain jobs would have color variation due due

to the wear and to the cylinders being out
of parallel.

The plate, blanket, and impression cylinders on an offset press are driven by gears

and supported
by bearers. Bearers are a common occurrence on web presses but not a

common design characteristic on sheetfed presses. Sheetfed presses are mostly

gear-driven. Even with the best gear drives, cylinders will vary minutely in speed and as

the mesh of the gears move from tooth to tooth. Dot Gain (the percent of growth over

original dot size) and directional dot gain (slur and doubling) are caused by actions within

the nip areas of the cylinders. The nip areas are between the inking roller and plate, the

plate and blanket, and finally between the blanket and paper .

The plate and blanket nip is considered to be the most critical. At the plate-blanket

nip, friction between the two cylinders is reduced due to the lubricating action of the ink
and water film. As the ink travels from plate to blanket and to paper, this same ink is split

in half between the two surfaces. Less ink between the blanket and paper is a cause of

more friction in that nip area. The resulting transfer occurring at the printing nip, the point

at which the paper is actually taking ink, is so delicate that even a slight variation in speed

can cause problems like gear streaking and doubling. The cylinder bearers help prevent
these problems by smoothing out the drive through rolling friction. For maximum effect,

they are preloaded, that is the cylinder bearers are brought together until there is substantial

contact force between the bearers while the blanket is compressed to provide the printing
pressure.9

The radius of the blanket cylinder is constantly changing as it goes through the

plate-blanket nip, because the blanket is soft whereas the radius of the hard plate cylinder

remains constant. Because of the change in the radius of the blanket cylinder, we can

expect a change or a tendency to change of the surface speed of the blanket. This change

in surface speed coupled with constant speed of the surface of the plate cylinder produces a

situation in which printing problems such as slurring and doubling are likely to occur.

Such is not the case in the blanket to blanket nip on a

perfecting press which prints
 13

both sides of the paper in one pass. The radius of the two blanket cylinders in the printing

nip are
simultaneously changing in the nip area. This change is expected to produce a

change in surface speed of the blankets. Whether the tendency is for the surface speed to

increase or decrease is of little significance in that there will be a similar tendency on both
sides of the nip. Whatever happens in the way of increasing or decreasing speed on the

surface of the blankets, the elasticity of the paper will allow it to track the two blankets

accurately by either stretching or contracting. This makes the blanket to blanket nip a very
stable operation and problems are not expected to to arise there. Instead any
printing
problems involving the transfer of the image such as slurring and doubling are expected to

occur in the plate to blanket nip.

Dr. A. Ghany Saleh found in his investigation into the causes of dot gain and its
effect on color reproduction that the pressure in the NIP area between plate and blanket and

blanket and substrate was an important variable. This pressure will result in an ink squash

which contributes to the dot growth. Machine geometry or design and press conditions

can induce other forms of dot gain or dot defects such as slur and doubling. Dot gain will

ultimately be influenced by press configurations, cylinder diameter, number of gears,

quality, dimensional stability, finishing and shape of the teeth, ink train, cooling system

housing.10
and the type of bearing in

The inking system on a press can be a major source of printing variations. It is

important that ink transfer be consistent in order to ensure consistent color. Consistent ink

transfer invariably involves proper roller settings, maintaining ink and water balance and

the proper chemical condition of the rollers. Rollers in the inking system are normally set

to one another and to the vibrator rollers. The ductor roller setting to the vibrator roller and

the receiving drum is most crucial. During tests run by GATF on instrumented presses,
where the ductor roller is set too hard against the vibrator, the load created by the

acceleration of the ductor was enough to slow the unit down and create vibration and

tension variations in the web before and after the printing unit involved. On a sheetfed

slight vibration that occurs on a press would automatically affect the plate.
press, any
associated with this problem of ductor shock include doubling and slur
Printing problems
and streaks.
 14

Ink and water are two other press variables that must be kept to a minimum to

maintain ink and water balance. However, minimum water should be run without

promoting piling on the ends of the rollers. Excessive amounts of water tend to be used on

longer running jobs; however this can be detrimental to the process. As mentioned earlier,

excessive amounts of water means a greater deposit of coating, sizing and fiber from the

paper coating. This wetness on the paper can cause piling and manifest itself in a grainy

print. In addition this water can accumulate on the inking rollers, vibrators, and
doubling.11
distributors and can consequently promote slur and

Blanket type and quality influence print quality. W. R. Grace & Co. U.S.A., a

leading blanket manufacturer carried out tests of performance of blankets under varying

packing conditions. Its studies revealed that dot gain and slur can be minimized by proper
blanket packing and that compressible blankets can print well under a much wider range of

than conventional blankets. Compressible blankets also have the added advantage
packing
of being compressed greatly at the nip while maintaining relatively light surface pressure

19
resulting in low dot gain and slur.
 15

FOOTNOTES

CHAPTER TWO

1
Porter, A. S., Lithographic Presswork First Edition, Graphic Arts and Technical

Foundation, 1980, p. 244.

2Ibid.

3Ibid.

4Anon., The Paper. Press or Blanket may Cause Cut Sheets to Wrinkle. American

Printer, Vol. 211 No. 5, September 1988, pp. 102.

5Crouse, David B. Web Offset Press Operating. Second Edition Graphic Arts and

Technical Foundation, 1984, p. 121.

6Treff, Ernest H. An Engineer's View of Dot Gain on a Web Offset Press. A

Presentation made at the PIA Annual Meeting in San Francisco on May 10, 1988,
p. 25.

7Tyma Louis S., Ingo Koebler and Herbert Stoeckle. Bearers A Necessary Evil?
TAGA Annual Proceedings 1982, p. 402.

8Printing Industries of America Web Offset Section Meeting Proceedings: 1987

Web Offset Section Annual Meeting, p. 244.

9Ibid.

10Saleh, Abdel Ghany. The Analysis of the Dot Gain Problem and its Effect on Color

Reproduction. TAGA Annual Proceedings 1984, p. 498.

 16

11Ibid.

12Leslie, Geoffrey. Blankets and the Control of Dot Gain. Australasian Printer Magazine,
Vol.38 No. 10 November /December 1987, pp. 8.
 17

CHAPTER THREE

REVIEW OF THE LITERATURE

A review of pertinent literature reveals that few studies of mechanical dot gain,

particularly doubling have been undertaken. There have been indications in literature that

control of mechanical dot gain through various measures has been attempted. However,
before control is possible, a method of analysis and measurement of directional dot gain is
needed.

Test targets are test images used to determine the quality of printing in the various

reproduction stages. Some test targets are sensitive to dot gain and are capable of

detenriining whether an increase in dot area is a result of directional or non-directional

gain. Doubling is often confused with slur. Directional dot gain, whether doubling or
slur, can cause a regular halftone tint to get darker and it may be hard to assign the real

cause to either of the two. However if a test target is used that is sensitive to directional

dot gain, it becomes easy to determine by visual means whether directional dot gain has

occurred. One of the first papers published where the relation between directional and

Rhodes.1
non-directional dot gain was studied was that by Warren He designed a test

pattern consisting of parallel line tints arranged side by side; one with horizontal parallel

lines and the other with vertical parallel lines. This pattern allows visual and objective

observation of definition; a term he used to describe sharpness of printing. He defined the

term resolution as the ability of a system to resolve or discriminate between closely

spaced elements. Resolution test objects perpendicular to the direction of sheet travel were

affected by slur more than those oriented parallel to sheet travel. If directional dot gain

exists, the two parts of the target will be affected differently and one will print darker than

the other. This is a visual indication of slur and/or doubling. Furthermore, his objective

method of evaluation was conducted with the use of measurements taken with a

densitometer. This method allowed him to distinguish between directional and

non-directional dot gain and to further assign numeric values to directional dot gain in
 18

terms of densitometric measurements. The disadvantage of this test target was that it
would not indicate a directional dot gain if it should occur at an angle of 45 degrees.

The LTF Star Target (or what is known today as the GATF Star Target) is an

adaptation and refinement of a similar target that was developed in 1957 by Robert E.
Wood of the Western Printing and Lithographing Company Plant in Racine Wisconsin.
The WPL target consisted of a solid one-eighth inch diameter center surrounded by ninety
pie shaped solid radial wedges. It is a visual indicator of mechanical dot gain; the center

increases in diameter if ink-spread (fill-in) occurs and elongates or stretches at right angles

to the slur direction if slur

occurs.2

GATF designed the Star Target with only 36 wedges which resulted in a much

smaller center. It proved to be a quick and effective measure of ink-spread, slur and

doubling during a press run. The geometric properties of this arrangement makes its press

sheet image sensitive to ink- spread, slur and doubling especially when stripped in the trim

areas at each corner of the trailing edge of the press sheet. This is because these areas are

usually the most affected by slur or doubling. Doubling will cause a figure 8 to appear in

the centre of the star, while slurring will appear as an oval in the centre with elongations at

right angles to the direction of sheet travel. The problem with this target is that it cannot be

used to make densitometric measurements for dot gain because it is too small. Originally,
designed for
use.3

it was visual evaluation

Figure 5. The GATF Star Target

The GATF Dot Gain Scale was another test target designed for inspecting and
and without the use of the densitometer. It also was a good
evaluating halftones visually
indicator of change in dot size at the various reproduction stages whether in the contact

of negative or positive in platemaking, proofing or in printing. The design of the

printing
 19

dot gain scale is based on the principle that tints made with fine screens are much sensitive
to changes in dot area than are coarse
screens. It consisted of ten numerical steps, starting
from 0 to 9. These numerical numbers were made with a 200-line screen while the

background was of a coarser screen of 65 tine. During the reproduction stage, if a

number has the same

density as the background, it would be an indicator of a numeric

amount of gain. While these numbers are sensitive to directional and non-directional dot
gain, there is also a parallel-line type section which indicates directional dot gain.

This parallel line section is known as the Slur Gauge. It consists of a bar of

vertical line background, and a series of horizontal lines which form the word SLUR. The
lines have the same value, so the word SLUR is invisible when all lines are printed with

equal thickness. But if slur occurs, the horizontal lines thicken and the word SLUR shows

up as darker or lighter than the background.

It2 .
p. '-(,1
Figure 6. The GATF Dot Gain Scale and Slur Gauge

GATF proceeded to design another test target later to be known as the Ladder
Target. The Ladder Target was designed to show gripper-to tail variation in the various

reproduction stages. Primarily it was designed to give a qualitative measure by visual

inspection, but numerical measures of slur and doubling could be obtained by taking
densitometric measurements in certain areas. The target was three-quarters of an inch wide

and twenty-five inches long. The center portion consisted of a 50% percent horizontal line
screen. The two outer strips are the same except that the lines run vertically. Slur and

doubling are measured quantitatively as the difference between the two neighboring areas.

The Ladder target is very sensitive to directional dot gain and manifests itself by a

perceptible darkening of the center of the target in the direction of printing. Doubling can
also cause the center of the target to darken if the doubling was occuring in the printing
direction. If the doubling was occurring sideways, the sides of the target will be darkened.

This target had the same disadvantage as the parallel line tint target in that it was incapable
 20

of showing any directional image displacement if it should occur at angles of 45 degrees.

Therefore a later version of this target also included a bar with lines at a
45 angle.5

Figure 7. The GATF Ladder Target

The UGRA Plate Wedge is another target that can test for directional dot gain.

Since its conception in 1962, the target has under gone some changes. The 1976 version

of the UGRA-Scale target consists of five elements namely a continous tone gray scale, a

60 and 120 line, a halftone step wedge, circular patches used as resolution targets, a line
target.6
patch and a slur The slur target is used to test for directional dot gain. It consists

of concentric circles whose distance is equivalent to their width (50% area). It is used to

observe slur and doubling on the press. This target was an actual refinement of the parallel

line tint target because it will show directional dot gain at any angle. Slur would manifest

itself as a spreading of lines running across the printing direction; lines running parallel to

the printing direction are not affected. Slur is indicated by two dark segments parallel with
the printing direction. Doubling also manifests itself visually as dark segments as in the

case of
slur.7

However these segments could be in any direction and there can be 2 or

more segments depending on the severity of misregister. Neither parallel line tints nor

concentric circles can distinguish between slur and doubling though they serve the same

purpose of indicating directional dot gain.

Test Targets such as Gretag, RIT Color Control Bars and the Kodak Customized

Color Target use a test patch comprised of concentric circles. Generally, their disadvantage

lies in the fact that they could only give qualitative or visual indication of directional dot

gain and do not lend themselves to numerical analysis because of their small size.
 21

FOOTNOTES

CHAPTER THREE

Rhodes, Warren L., Study of Objective methods for Evaluating Sharpness in Lithography
TAGA Proceedings 1955, p. 109

2Jorgensen, George W., The GATF Star Target for Inkspread and Resolution

Measurements. Research Progress Report No. 52, February 1961, Graphic Arts
Technical Foundation, Pittsburgh, PA.

3Anon., GATF Scales Detect Dot Gain. Canadian Printer and Publisher, April 1986, p. 37

4Ibid.

5Hull, Harry H., The GATF Ladder Target -

A New Test Image. Graphic Arts Technical

Foundation Research Progress Report # 99, October 1973

6Sigg, Franz. A few things about Microlines that most people do not know, TAGA

Proceedings 1988, p. 434

7UGRA -
GRETAG Plate Control Wedge PCW: Technical Description
 22

CHAPTER FOUR

HYPOTHESES

1 . The dot area difference between the dark and light segment on the Experimental RIT

Doubling Target is not significant when 12 mil packing is added to the blanket while

simultaneously removing the same amount from the plate packing.

2. There is no statistically significant relationship between the direction of doubling and

the magnitude of doubling.

3. There is no difference between Dot gain measured on the 50% tint of the Experimental

RIT Doubling Target and the average dot gain calculated from the dot gain
readings of the light and dark segments on the concentric circles.
 23

CHAPTER FIVE

DESIGN OF THE EXPERIMENTAL RIT DOUBLING

TEST TARGET

Design of the Target

The RIT Doubling Target is 1 1/4 inches square in area. The major part of the

target is covered by a series of concentric circles. At the corners are located four 5mm

square patches that consist of:

1 . 3% dot patch used to observe the effect of doubling on small dots

2. Solid patch used to indicate ink-film thickness in order to calculate dot gain

3. 50% tint to correlate between the average of light and dark area values of the

concentric circles and the 50% tint

4. A clear patch to zero out the densitometer.

In each half of the circumference of the circle, a scale, indicating angles from 0 to 180 of

directional dot gain is available to facilitate labeling the direction of directional dot gain.

RITDOU LING TARGET

'*-
160
0
20
1
^.60
120
^^1
100-j^H

30 JH

eo^^H W 12

w.-
140
dfl^^^
20
0
160
j
Rochester In ol Techn ogy
|

Figure 8. The RTT Experimental Doubling Target

 24

Aspects Of new desinn as comnarpri to the parallel line tint :

It was made into circles allowing it to be sensitive to all directions.

Its large size allows one to take minimum and maximum measurements in the
lightest and darkest segments within the patch with a normal densitometer.
Its use of the
Murray Davies equation allowing it to obtain values for directional
gain as opposed to the use of visual evaluation methods or
density measurements.

Small dots may make it easier to differentiate between slur and doubling.

Characteristics of the concentric circles design:

The basic advantage of the new design is the large size of the concentric circles.

This enables one to make dot area measurements at the light and dark segments which

gives an indication of directional and non-directional dot gain. Moreover the design of the

concentric circles is an advantage in that it solves the problem of directional dot gain

occurring at 45 degrees which was impossible to take care of with parallel line tints.

The principle of measuring dot gain in this fashion is not in question and using

Murray Davies is simply an existing method for this specific application. Where the circle

is lightest, there is no directional dot gain because the darkening of the line falls on top of

itself indicating no directional dot gain. The actual reading would only indicate fill-in and

optical dot-gain. At the darkest spot, we measure the worst possible case of directional dot

gain as well as fill-in and optical dot gain. Hence the difference between the light and dark

segments is due to the contribution of directional dot gain.

Problems with the method:

If doubling should occur over several printing units, there may be a complex

moire pattern that has not only one minimum and maximum section. This may make it

impossible to measure. This is, however, a problem for all doubling measurements, and

not just only of this method.

 25

A problem arises in that doubling is a complex phenomenon. It is unpredictable

and can not

only occur in one direction on one printing unit but in a different direction on

another
printing unit. Furthermore, the color printed on the last printing unit on a press,

normally does not have any doubling of dots, since no more printing units for it to go

through, exist.

It was previously stated there is the possibility of having slur on top of doubling
and both of them can occur in different directions. In order for this method to work, we

therefore need to rely on simplified assumptions which are:

1 . Slur is not considered much of a problem; packing has to be off by a sizable

amount before slur occurs.

2. Doubling occurs more at the first and second units following the original

impression than between the second and third units. The farther away the

sheet is from the original impression, the weaker will be the print-back and

therefore the doubling and the effect of doubling. The ink at this point is

diluted and weak and would not contribute greatly to the problem.

Screen Ruling:

1) What screen ruling should be used for making the concentric circle in order to

obtain, on the average, the approximate same value of dot gain as would be

obtained on a 150 line/inch 50% dot tint ?

How is dot gain a function of the doubling angle for a dot tint and a line tint ?
2)
 26

50% Dot of 150 Line/inch Screen

Ruling Line Tint

ml siwulOOSS wu

10 units
5uiii5

10 units

14.14 units

Units2
100% Dot Area = (14.14 units)2= 200 =
Unit Square

50% Dot Area =

(10.0 units)2
=
100 Units2= 50% Dot

Figure 9. 50% Dot and Line Tint

If we chose the geometric relationships between dot and line screens such that

45
both show the same area increase for a double that occurs at angle, then the
45
relationship of figure (9) results. Note that the dot screen is shown at a angle. Under

these conditions, the screen ruling for the line tint has to be 1.41 times finer than for the

dot tint. In other words the equivalent screen ruling for a 150 lines/inch dot halftone is a

line tint with 150x1.414 = 212 lines/inch screenruling.

Below is a diagram of two sine functions that indicates the percent area change due

to doubling. In other words, if a double is a function of a certain direction (or angle), we

can calculate the expected % change in area for both the dot and line tint. Tables 1 and 2

are the calculations that were used to illustrate graphically the sine functions for both the

dot and line tint. The average for both sine functions happens to be almost identical.
 27

10 n

Line Screen

3 8 -

O
Q
Average

3
o
a>
D)
c 4 -

m
-C

O
(0
4) 2 -

T"

T
90 180 270 360

Angle of Double

Figure 10. Plot of % Area Change due to Doubling Angle for a

50% Dot and Line Tint

Table 1. Percent Dot Area Increase due to Doubling of 1 Unit at Various Angles

Dot Screen
Percent Increase in Dot Area due to Doubling Ancle Doubling Ancle
0
l*10*(sin(0) + cos(0)) =
10
units2
=
5% area double
22.5
l*10*(sin(22.5) + cos(22.5))
-
13.06
units2
=
6.533% double
45
l*10*(sin(45) + cos(45)) = 14.14
units2
=
7.07% double
67.5
l*10*(sin(67.5) + cos(67.50))
=
13.06
units2
=
6.533% double
90
l*10*(sin(90) + cos(90)) = 10
units2
= 5% double
 28

Line Screen

Percent Increase in l ine Area due to Angle Doubling Angle

Doubling
2* 10*
1 *sin(0) =
0 units =
0% area double
0

2*10*l*sin (22.5) =
7.65 units =
3.837% double 22.5

2*10*l*sin (45) =
14.14 units =
7.07% double
45

67.5
2*10*l*sin (67.5) =
18.48 units =9.24% double
90
2*10*l*sin(90) =
20 units = 10% double

The average of a half sine curve is defined by the equation: l/2*Tt*amplitude

Dot Screen =
(2*2.07)/tc =
1.317 : 1.317 + 5.00 =
6.317 % which represents the average

value of the dot screen

Line Screen =
(2*10)/rc =
6.366% which is the average value of the line screen
 29

CHAPTER SIX

METHODOLOGY

The analysis procedures utilized in this fulfill

experiment were chosen to the following
tasks:

1 . To show whether slur is a major part of directional dot gain

2. To understand whether the

doubling magnitude is a function of the doubling angle
3. To determine whether register variation on the two dimensions of the web

(circumferentially and laterally) is related to the problem of doubling.

4. To examine whether a
relationship exists between the concentric circles and the

50% tint and determine if the concentric circles can be used to represent a 50% tint

in terms of non-directional dot gain.

The First Experiment:

In order to study the effect of slur, we can purposely produce it by removing the

packing from the plate and replacing it underneath the blanket. This will result in a change

in diameter in both the plate and blanket cylinder. The plate cylinder diameter was reduced

by .3mm of packing and the blanket cylinder was reduced by the same amount. A

condition as such would theoretically create slur. The Experimental RIT Doubling Test
Target was used to visually verify whether slur had occurred.

The Second Experiment:

Since it was suspected that the automatic web alignment mechanism might cause

doubling and registration problems, two web press runs were carried out with the

following two conditions:

Conditions:

1 -

Web Aligner and Register Control were on automatic.

2- Web Aligner and Register Control were turned off.

 30

A web-steering device known box is found

as the tilt basically at two locations on
a press
namely at the infeed controls and before the folders. This mechanism is
responsible for the sideways placement of the web through the units and the
printing
folders. Running two tests where the tilt box was activated and deactivated would permit

studying its effect on register and doubling. Two test targets sensitive to register variations

were used to
visually detect and measure any register fluctuations. One was positioned

horizontally and the other

vertically on the test form.

Several test targets sensitive to dot gain were incorporated in the layout of the test

form. They were used as a visual reference for comparison between the new test target

and the common ones in use. The color sequence that was run on the press was black,
cyan, magenta and yellow. Yellow being the color on the last unit was run with a blank

plate because it was felt that it would not throw any light on doubling since the last unit on

the press suffers the least or no doubling.

Once makeready was complete and the required density levels were attained, 1000

consecutive sheets were taken off the stacker and numbered in proper sequence. The two

press runs were given code numbers 1 and 2 signifying the above mentioned conditions.

Magenta and black were the two units studied for doubling since they represent the extreme

cases on the press. Black being the first unit is expected to have more doubling than
magenta. Cyan was not analysed because it was assumed that its response will be

intermediate between black and magenta.

Two visual readings were taken from the RIT Visual Registration Scale target.

One reading was for lateral register and the second was for circumferential register. The

remainder of readings were taken off the RIT Doubling Target. One was a visual reading

of the angle of the double. The densitometric readings were taken with a GRETAG D186

densitometer which was interfaced to a computer where the data was accumulated. The
 31

following patches were read: the solid patch, the 50% patch and the dark and light
segments in the concentric circles. It is important to note that the same targets at the same

location on the press sheet was used in the collection of data.

Measurement of Vertical and Horizontal Register

The terms vertical and horizontal register signify circumferential and lateral register

variations on the two dimensions on the web. Sideways movement of the web caused by
the web alignment and the automatic register system can result in register variations. Two
RIT Visual Registration Scale targets were incorporated into the design of the test form to

relate these two variables magnitude of doubling. One target was positioned in the

direction of web travel in order to indicate any circumferential variation that might occur,

and the other was positioned perpendicular to web travel to show any lateral or sideways

change in register. Another additional variable was calculated from the values obtained

from circumferential and lateral measurement and was indicated as Magnitude of

Misregistration. This variable is defined by the equation:

Magnitude of Misregister= V Vave)2

+
Have)2

(Vreg (Hreg
-

Vrfl: Vertical or Circumferential Register Variation

I
eg

Vave: Average or mean value of Vertical or Circumferential Register Variation

H.^: Horizontal or Lateral Register Variation

Have: Average or mean value of Horizontal or Lateral Register Variation

Hypothesis 2 was an attempt at relating the magnitude of Doubling to the

corresponding doubling angle simultaneously taken from the same target. However, the

process of to relate these two variables did not end at this point. It was felt that
trying
and horizontal register changes would provide
relating magnitude of doubling to vertical
some additional information that could be of noteworthy interest. Furthermore the register
 32

angle derived from the two coordinates, vertical and horizontal register can also be studied

in relation to magnitude of misregister. The value for magnitude of misregister can be

calculated from the two variables vertical and horiontal register variation. The register

angle and magnitude of misregister would define the two coordinates on a polar system of

coordinates.

RIT Visual Registration Lead Edge RIT Visual Registration Scale

positioned vertically
Scale positi on et
(circumferential effect)
horizontally
(lateral effect)

Operator Side Direction of Paper Travel

Figure 1 1 Layout
. of Test Form used on Web-Press Run

Circumferential Register

Magnitude of Misregister

? Lateral Register

Angle of Misregister

Tangent of the angle = x_= lateral register

y circumferential register

Vertical and Horizontal Register on a Cartesian System of Coordinates

Figure 12. Plot of
 33

Determination of
doubling angle

A degree of
uncertainty was noted when for
trying to read the value doubling angle off the
doubling target:
1 It was felt that
.

reading the doubling angle off the target did not need to exceed an

10
accuracy of + 5 degrees.
2. Because there are two segments, there is a of
180
rather than
symmetry
360. Therefore an angle of
0
degrees corresponds or is the same as an angle of

180.

3. In addition, if there was no moire or specifically no directional dot gain that was

evident on the target, it was impossible to assign a doubling angle. We chose in

such a situation to give it a value of zero.

Analysis:
A special program written in BASIC language was used to collect the numeric data that

was taken visually and densitometrically. The Statistical Method of Analysis used to test

Hypotheses 2 and 3 were:

1) Regression and Correlation Analysis in order to develop a linear equation relating the
two variables, namely Doubling and Doubling Angle and 50% Dot Tint and 50% Line Tint

(concentric circle). It is important to point out that this method of analysis does not imply
the establishment of a cause-and-effect relationship between the two variables.

2) The Pearson Product Moment Correlation Coefficient r was used to investigate the

nature of the relationship between the two variables. Values for r were defined between -

and +1 . A value of +1 indicates that x and y are perfectly related in a positive linear sense

that all the points in the scatter diagram lie on a straight line with a positive slope.
meaning
A value of -
1 indicates that the two variables are perfectly related in a negative sense. An r

value close to zero indicates that x and y are not linearly related.
3) A t test was used to test for a significant relationship between the dependent variable and

the independent variables. The critical value for a two-tailed t distribution for a sample size

of 500 and a confidence level of 99% equals 2.58. Any t-value in the experiment above

significant relationship.
2.58 was considered a statistically
 34

4) Graphical Representation was carried out

using the Macintosh Cricket Graph software.

All calculations and computations were done using Minitab Statistical

the Computing
System.

Test for Significance:

In the statistical analysis of this study, the regression equation
relating any two
variables that were studied was obtained. The regression equation is

E(y) =80 + BjX E(y) y for

where = mean value of a given value of x; B0 =y-intercept
of the regression line and
Bj =
slope of the regression line.

The estimated regression equation was obtained and it is represented by the following
equation: y =
b0 + bjx
If a relationship exists between x and y, the coefficient BjX should differ from zero.

A conclusion regarding the significance of the regression relationship can be tested

using the following hypothesis:

HQ: Bi = 0

Hj.B^O

One of the properties for the sampling distribution for b: is that

B11
Mean: E(b1) =

bi Bi / sbl has a t distribution with n-1 degrees of freedom

The t test regarding Bj =

(n representing the sample size and n-1 representing the degrees of freedom).

t distribution table, a value for t corresponding to a level of

By looking under a

confidence of 99% and (n-1) degrees of freedom is obtained. If this t value is less

hypothesis be
than the calculated value B, =
bx Bx / sbl ,
then the null will rejected

and one can conclude that there is a significant relationship between the two variables.

deviation of the coefficient x)

(sbl is the standard
 35

FOOTNOTES

CHAPTER SIX

iDavid R. Anderson, Dennis J. Sweeney, Thomas A. Williams. Statistics Concepts

and Applications. Second Edition. West Publishing Company: St. Paul,

1986. pp.525-527
 36

CHAPTER SEVEN

RESULTS

First Hypothesis
Visual examination of the target revealed that no slur or
doubling had occurred.
There was no moire indicative of the phenomenon, however there was the expected change

in the print length of the image as a result of the shift in packing from the plate to the

blanket (Refer to chapter 2 under slur). The image printed longer because of the decrease
in plate diameter. What is of importance in a situation like this is that a change in print

length is expected to happen but not that of slur. The Null Hypothesis was accepted due to

the obtained results.

Second Hypothesis
There was no strong correlation between magnitude of doubling and doubling
angle. This was obvious from the low values for the correlation coefficient for both black

and magenta. The values obtained from the t test however signified that there was a

significant relationship between the two variables. This significant relationship is an

indication that we are 99% sure that the variables involved for the test for the correlation

did not occur by chance.

Table 2. Table of r and t valuesfor Magnitude of Doubling
versus Doubling Angle

Magnitude of Doubling

Black 1 Magenta 1 Black 2 Magenta 2

r Doubling Angle .38

.46
.43 .49

9.17 11.65 10.59 12.58

t Doubling Angle
significance at alpha level of .99
-99 -99 -99
 37

What was interesting to note in the graphical analysis, was that a relationship can

be shown to exist between the two doubling

variables magnitude of and doubling angle by
simple observation of the polar plots. One of the features of the scatter and polar plots as

shown in figures (13 and 14) and (15) was the orientation of the points. There appeared to

be a preferred angle for both colors black and magenta. In the case of black for conditions

1 and 2, the preferred angle was within the range of 120 -

160 degrees. Magenta for

conditions 1 and 2 had a preferred angle between 60 and 120 degrees. Magenta had a

distinct feature in that the preferred doubling angle was in the orientation of paper travel.

This probably indicates that sideways lateral register variation (sideways lateral paper

movement) does not effect the problem of doubling as much as the vertical or
circumferential register variation on the web. A probable explanation can be the affect of

web tension, that can ultimately affect paper performance thereby exaggerating the

doubling problem. The t test also provided further information that 99% of the variables

involved for the test of correlation did not occur by chance. Based on the graphical

representation and the results of the t test, one can arrive at the conclusion that there exists

a relationship between magnitude of doubling and doubling angle. The results allow us in

turn to reject the null hypothesis.

Third Hypothesis
There was a correlation between dot gain measured on the 50% tint made with a 150
strong
line screen-ruling and dot gain measured on the concentric circles. This was indicated by
the values for the correlation coefficient r for conditions (1) and (2) on the press for both

colors black and magenta (refer to figures (16 and 17). The values were .84 and .79 for

black and magenta in condition (1) and .91 and .88 for conditon (2). Furthermore, the t

indicated that a significant exists between the two values for dot gain in
test relationship

that 99% of the variables involved for the test for the correlation did not occur by chance.
Hence the null hypothesis was rejected.
 38

y =
3.4437 + 0.01 6x R =
0.38 Black 1

a
3
O
Q
O

o
3
?^

C
D)
J
s

90 180

Doubling Angle

y ==
3.339 + 0.0201X R ==
0.43 Black 2

3
o '
i :
Q
*-
. . : 1 "

O 10 -

:
' ; .

V
D
3 ! > i
: . .

'c
(0
'
i ! i
:
|
: , .

: ; .
i .i

0 -
1
90 180

Doubling Angle

Figure 13. Scatter Diagram and Straight Line Approximation between

Magnitude of Doubling and Doubling Angle for Black 1 and 2
 39

y = 1 .9453 + 0.0256x R =
0.46 Magenta 1

3
O
Q
*

o
ffi
o
3

'E
o>
ra

180

Doubling Angle

y = 1.892 + 0.0248x R =
0.49 Magenta 2

3
O
Q

(9
(J
O
Q

c
o

180

Doubling Angle

Approximation between
Figure 14. Scatter Diagram and Straight Line
Magnitude of Doubling and Doubling Angle for Magenta 1 and 2
 40

150
150

180
0 180

OS I.
OS I

Black 1

150 150

180 0 180

OSt OSL

The distance from the center indicates the magnitude of doubling in % dot area.
Direction of paper travel is at 90 degrees.

Figure 15. Polar Graphs of Doubling for Black and Magenta 1 and 2
 41 1

y= -19.8185 + 1.181x R =
0.84 Black 1

o
w
c
o

cs
O
o
D

65 75 85 95

Average Dot Gain on Concentric Circle

y= -

12.6855 + 1.0999X R = 0.91 Black 2

5?

s
c
o

ca
U
o
a

65 75 85 95

Average Dot Gain of Concentric Circle

between
Figure 16. Scatter Diagram and Straight Line Approximation
Dot Gain on 50% Tint and Average Dot Gain on

Concentric Circles for Black 1 and 2

 42

y=
-30.9446+ 1.31 05x R =
0.79 Magenta 1

o
lO
c
o

(3
o
Q

65 75 85 95

Average Dot Gain of Concentric Circle

y= -8.7546 + 1.0568x R = 0.88 Magenta 2

ss
o
IT)
C
o
c
'<5
(3
+*

o
Q

65 75

Average Dot Gain on Concentric Circle

Figure 17. Scatter Diagram and Straight Line Approximation between

Dot Gain on 50% Tint and Average Dot Gain on
Concentric Circles for Magenta 1 and 2
 43

Table 3. Table of r and t values for Dot Gain on 50% tint

versus Average Dot Gain on Concentric Circles

Dot Gain on 50% Tint

Black 1 Magenta 1 Black 2 Magenta 2

r Ave. Dot Gain on Concentric .84

.79 .91 .88

Circles
t Ave. Dot Gain on Concentric 34.28 28.83 47.48 40.78
Circles
significance at alpha level of .99 .99 .99 .99

Further Observations

The Black unit had a higher value for doubling than Magenta under both

conditions (1) and (2). Black being the first unit is expected to double more than Magenta.

Magenta being the last unit printed on had a lower value for doubling. This was observed

from the mean values for Magenta and Black in the analysis.

Table 4. Mean Value and Standard Deviation for Magjnitude

of Doubling and 50% Tint

Doubling 50% Tint

Mean St.Dev Mean St.Dev. Min. Max. Dot Gain

Black 1 4.93 2.68 77.27 2.82 61.7 92.9 27.3

Magenta 1 3.86 1.70 81.08 2.13 76.9 90.1 31.1

Black 2 5.13 2.95 76.39 2.39 71.5 84.4 26.4

Magenta 2 3.63 1.62 77.91 1.72 73.5 82.2 27.9

 44

The test run under both conditions where the web alignment and automatic register

system was activated and deactivated did result in a statistically significant difference in

magnitude of doubling and the mean value for dot gain on the 50% tint for both conditions.

The mean value of the magnitude of doubling for black was slightly higher under condition

2 than it was for condition 1, while the reverse was true for the mean value of dot gain for

the 50% tint. In the case of magenta, the mean values obtained for magnitude of doubling
and dot gain were slightly less under condition 2 than under condition 1. Even though

these differences are statistically significant, a 0.2% area difference due to doubling has
simply no practical significance.

All tests showed average dot gain within the SWOP specifications of 24 + 4%

except Magenta 2 which had an average dot gain of 31%.

Table 5. Table of r and t values between Magnitude of Doubling and

Lateral-Circumferential Register

Magnitude of Doubling

Black 1 Magenta 1 Black 2 Magenta 2

-.03
Lateral Register .04 .19
.04

Circumferential Register -.02 .16

-.14 .17

Lateral Register .89

4.54 .88 -.73

Circumferential Register -.45 3.71 -3.10 3.95

no no no
significance at alpha level of .99

no .99
no .99
 45

CHAPTER EIGHT

SUMMARY, CONCLUSIONS AND OBSERVATIONS

Two main hypotheses were tested in this experiment. The first hypothesis was if

doubling occurs, does it necessarily occur in a preferred direction. After analyzing the data
to try and indicate whether a
relationship exists between the two variables, the results

obtained from the correlation coefficient r showed that

doubling angle showed no

correlation to the magnitude of doubling. This means that generally, smaller values of

doubling are not associated with different angles than larger values of doubling. However
the scatter and polar plots indicated that for both black and magenta, there was a preferred

range of doubling angles. Reading the doubling angle off the target was not an accurate
procedure and an element of
uncertainty was evident in the readings.

The second hypothesis was a test of the target's usefulness and to further

determine whether the target is capable of representing a 50% tint in terms of dot gain. The
original intent in the design of the concentric circles was that it should indicate

non-directional dot gain similar to what happens on a 50% dot pattern. The target permits

separate measurement of directional and non-directional dot gain. The reading is important

in that it throws light on what total portion of total dot gain is due to the directional part of

the overall dot gain value. One of the drawbacks of this method and of the target in

general, is that one can only take measurements manually and not with a scanning
densitometer. One has to visually determine the light and dark segment on die concentric

circles before attempting to take actual readings with the densitometer. A smart

densitometer could overcome this problem if it was programmed to take readings at

specified locations on the concentric circle. The resulting values can be calculated to

provide the numeric values for directional and non-directional dot gain. Targets similar in

design to the Kodak Customized Color target make it easier to use a scanning densitometer

for readings are required.

long press runs where many
 46

The concentric circles were made with the intent of simulating a 50% tint of 150
line screen-ruling as far as non-directional dot gain is concerned. The values for dot gain

on the concentric circles and the 50% tint did not however match. A strong and significant

correlation was indicated in the values obtained for the correlation coefficient r. However
the value for dot gain on the concentric circles was higher than that of the tint. Further
studies are still needed to account for this discrepancy especially if the error cannot be
attributed to screen-ruling.
*&

Other observations drawn from this experiment:

The Experimental RIT Doubling Target was found to be a sensitive target in its

configuration and design to directional dot gain. The components of the target lend

themselves to visual and numerical evaluations of dot gain. Primarily, it was designed to

test for all the aspects of mechanical dot gain, directional and non-directional. Numerical

values for the latter and the former are possible through the use of densitometric readings

from the dark and light segments of the concentric circles. The original intent of having a
3% tint was for visual identification of the particular kind of directional dot gain, be it slur

or doubling. The 3% highlight dot was probably not the best size to visually identify
directional dot gain. This is due, in part, because these small size dots tend to get lost in

the roughness of the paper. It takes more than a loop of lOx magnification to observe it.

During the press run of the second experiment, it was evident on some of the press sheets

that a double dot exists. An improvement on the target that could aid in visuallizing

directional dot gain more distinctively would be having several highlight dots made

within a range of 2% to 6% in order to visually distinguish a double dot when it

preferably
occurs.

An observation was noted in the value obtained for magnitude of doubling on the

press sheets. As mentioned earlier the magnitude of doubling was calculated as a

difference in dot area of the two segments on the concentric

circles. This was carried out
 47 7

with the use of the Murray-Davies Equation. The value for doubling on every second

revolution of the plate for the impression of the blanket, was found to be significantly
higher than the value obtained on the first revolution of the plate. This variation is possibly
due to the slack that paper undergoes at the second impression of the plate and the blanket
because of the plate-gap blanket nip. The blanket could experience a slackness as it goes

through the area of the plate, which might result in

gap a momentary decrease in web

tension. This slackness in paper

tensioning potentially could cause register variation, or

doubling.

The RIT Doubling Target can be used as a visual and quantitative measure of the

two aspects of Mechanical Dot Gain namely Directional and Non-Directional Dot Gain.
Applications of the target can be directed towards press checks before they are placed in
operation or routine checks on older presses. This applies to production sheet-fed presses

and web-presses. It can also be used to detect any slur or doubling on proof sheets, before
the actual production sheets are run.

In conclusion, even though there are some weaknesses to the target, it is a useful

tool to quantitatively and qualitatively determine directional and non-directional dot gain.
 48

CHAPTER NINE

RECOMMENDATIONS FOR FURTHER STUDY

The following is a list of recommendations that would prove useful in

understanding the target's capabilities, that were not covered in this study.

1 . The test runs that were carried out in this study were done on presses that did not

necessarily represent the conditions of presses out in industry. Doubling was

studied in relation to paper and register effect. It needs to be added that equipment

wear is also a major contributor to the problem of doubling. The presses that were

used are in good condition and therefore tended to minimize the problem.

Implementing the slur and doubling experiment on industry presses to test whether
equipment wear is a contributor could prove to be a useful study.

2. The data for Cyan (being the second unit on the press) could also be studied in
order to test its relevance to the problem of doubling.

3. The web press run actually included two other conditions in addition to the two

outlined in this study. The third and fourth conditions were situations where the

web aligner was on manual while the register control system was on automatic and

the fourth condition was the opposite. A study of these conditions and their effect

on doubling could throw additional light on the problem of doubling.

4. The second experiment that was carried out in this study should be carried out on

another web press in order to determine whether the preferred angle for black

(being the first unit) and magenta (being the third unit ) would match the results
 49

for the preferred angle found in in this study. Furthermore, it would be interesting
to find out why the preferred doubling angle differs from one unit to another.

The relationship between magnitude of misregister and misregister angle was not

analyzed in this study. The magnitude of misregister and misregister angle are

two values that can be derived from the values of vertical and horizontal register

variation. This relationship could be studied and compared with that relationship

of magnitude of doubling and doubling angle. (Please refer to chapter 5.)

 50

BIBLIOGRAPHY

Private Communication : Franz Sigg, Technical and Education Center,

Rochester Institute of Technology

G. P. Cooke, CA. Hill, Some factors affecting halftone dot growth in Lithography
(PT 17ii) Pira Report No. 161/1069

De Paoli, Alan, The Effect of Printing Conditions on Dot-Gain -

An Offset Newspaper
Study TAGA Annual Proceedings 1981, pp. 17-42

Sigg, Franz, Doubling, an unpublished report.

Anon., GATF Scales Detect Dot Gain. Canadian Printer and Publisher,
April 1986, pp. 36-37.

Saleh, Abdel Ghany, The Analysis

of the Dot Gain problems and its Effect on Colour
Reproduction TAGA Proceedings 1984, pp. 497-517

Louis S. Tyma, In go Koebler and Herbert Stoeckle, Bearers A Necessary Evil ?

TAGA Proceedings 1982, pp. 402-416

Printing Industries of America Web Offset Section Meeting.

Proceedings: 1987 Web Offset Section Annual Meeting

and the Control of Dot Gain. Australasian Printer Magazine,

Leslie, Geoffrey, Blankets
Vol.38 No. 10 November/December 1987, pp. 8,10 and 11

Tritton, K.T., Modifying The Dot Gain Characteristics of Sheetfed offset Presses.
(PR/163 (c)) Pira Report PR4 (R) 1981

B..Web Offset Press Operating. Second edition Graphic Arts and

Crouse, David
Technical Foundation 1984

know Paper. Graphic Arts Technical

Bureau, William H., What the Printer should about

Foundation, pp. 190, 215

Annn The. Paper. Press or Blanket mav cause Cut Sheets to wrinkle. American Printer,
Vol. 201, No.5 September 1988, pp. 102-3
 51

Anon' Pressroom. Curing Double Vision: Precautions can help prevent Dot Doubling
American Printer, Vol.201 No.4,
July 1988, pp.78-9
Sigg, Franz, A Few things about Microlines that most neonle. do not know TAGA
Proceedings 1988, pp. 428-449.

Hull, Harry H., The GATF Ladder Target- A New Test Tmage Graphic Arts Technical
Foundation Research Progress Report #99, October 1973.

Maurer, R.E., Customized-Color Computer Printing Analysis TAGA Proceedings 1982,

pp. 518-550.

Jorgensen, George W., The GATF Star Target for Ink Spread and Resolution
Measurements. Research Progress Report No. 52,
February 1961,
Graphic Arts Technical Foundation, Pittsburgh, PA.

Arnamo, A., Dot Density Analysis in Web Offset. PIRA/ARIGAI International Conference
on Applied Lithographic Technology 1970. London, October 7-9, Paper 14
Milton Pearson, Irving Pobboravsky, Chester Daniels, Instrumentation for the
Measurement of Slur and Fill-in on a Lithographic Web Press. TAGA Proceedings
1979, pp. 162-175.

Elyjiw, Zenon, GATF Standard Offset Color Control Bars. Research Progress Report
No.76, February 1968, Graphic Arts Technical Foundation, Pittsburgh, PA.

Frank Preucil, Zenon Elyjiw and Robert F. Reed, The GATF Dot Gain Scale. Research
Progress Report No.l 19, 1983, Graphic Arts Technical Foundation,
Pittsburgh, PA.

Jorgensen, George W., Control of Color on Press: Halftones. Research Progress

Report No.l 19, 1983, Graphic Arts Technical Foundation, Pittsburgh, PA.

Treff, Ernie H., An Engineer's View of Dot Gain on a Web Offset Press. A presentation
made at the PIA Annual Meeting in San Francisco on May 10, 1988, pp. 1-42.

Anon., Make the Halftone fit the Paper. Research Progress Report No. 15,
September-October 1949, Graphic Arts Technical Foundation, Pittsburgh PA.

Rhodes, Warren L., Study of Objective Methods for Evaluating Sharpness in Lithography
TAGA Proceedings 1955, pp. 109-122
 52

Jorgensen, George W., Sharpness of Halftone Images on Paper. Research Progress

Report No. 47, July 1960, Graphic Arts Technical Foundation, Pittsburgh PA.

J. A. C. Yule and W. J. Nielsen, The Penetration of Light into Paper and its Effect on
Halftone Reproduction. TAGA Proceedings 1951, pp. 65-74.

Soufhworth, Miles F., Dot Gain Causes and Cures. Quality Control Scanner, Vol.2
No.9 1982.

Porter, A.S., Lithographic Presswork. First Edition, Graphic Arts and Technical
Foundation, 1980.

David R. Anderson, Dennis J. Sweeney, Thomas A. Williams. Statistics Concepts and

Applications. Second Edition. West Publishing Company: St. Paul, 1986.
 53

APPENDIX A

CALCULATION OF THE CHANGE IN PRINT LENGTH

CAUSED BY A CHANGE IN PACKING

Notations:

1 micron =
10~6m =
.001mm

120 microns is the side length of a 50% square dot of 150 line screen-ruling
120 microns =
.12 mm

Diameter of plate and blanket =

220 mm
(Heidelberg specifications)
Change in packing = =
.3mm

Sheet size =
17 1/2 x 22 inches

Circumference of a Cylinder =
re x Diameter
=
re x 220 mm

=
691.15 mm

Sheet dimension around the cylinder =

17 1/2 inches x 25.4mm/inch =
444.5 mm

% of circumference occupied by image =

Ratio of length of paper x 100

circumference of cylinder

Percentage =
444.5 x 100 = 64 %

691.2

in formula D= 2PrcC / 100

To calculate the change print length, the

was used where D denotes change in print length; C represents the percentage of

the circumference occupied by the image and P is for packing change.

D= 2 x .3mm x 3. 1415 x 64% = 1.206 mm

 54

The percent of change in print length is then calculated as

1.206/444.5 x 100 =
.27%

A dot would change by the same % length. An example is that of a 50% dot made

with a 150 lines/ inch screenruling having a side length of 120 microns.

=
.12mm x .0027 .0003 mm.

.1203 . 12 =
.0003
meaning the change in size of a 120 micron dot is

approximately .3 microns.

Since Slur is usually in the cylinder rotation direction, the overall increase in dot size is going

to be in one direction only and not all the way around the dot. Therefore dot area is affected

by a very insignificant amount.

 55

APPENDIX B

DERIVATION OF THE MURRAY-DAVIES EQUATION

The Murray-Davies Equation is used to represent the relationship of dot area to the

reflection or trasmittance characteristics (density) of a halftone dot.

50% Dot 100% Dot

1
.5
It =100%

I^= Incident Light

Intensity of

n=Intensity of Transmitted Light

Transmittance Opacity = 1 =
Ij
=
It
T L

Density =
D =
log10 Opacity
=
log^o I
T

a = dot area ; a = 1 T

Transmittance of Solid Black = 0% :

Transmittance of a Clear Film = 100%=1:

50% Dot = 1/2 black and 1/2 clear = .5

Transmittance
 56

Question: What is the density of a 50% dot ?

D 1/.5
log10 log10 2
= = =
.301

for a 75% dot T=25%; Density =

log 10 1/.25 =
log 4 =.602

for a 87.5% dot T=12.5%; Density 1/. 125

=

log10 =log8 =.9

for a 90% dot T=10%; Density l/.l 10

=
log10 =
log =1

Murray Davies equation permits calculation of dot area when the density of the dots is
not infinite.
Transmittance of a 90% dot =
10% =
. 1

TtInt =
(axTdot) + ((l a)xTclJ

^tint =
aTfot +
Tclear aTclear
-

Ttint= + l
a(Tdofl)

1 T^t =
a (1 -Tdot)

a = 1 -
T
tim
Since the transmittance of the dot equals the transmittance of a solid,

10"Dt
1 Tdot the equation becomes a = 1 -
T t[nt
= 1 -

10"Ds
1
1 -
T 11 -
1U
'oM

Example:
90% dot where transmittance of dot area =10% =
. 1

T90% ((1_a) x T10%clear) + (a x T90%tint)

(.1x100%) + (.9x10%)
=
.1+.09 .19

1/.19 log 5.263 =

Densitytint =
log10 = .72
 57

APPENDIX C

Table of Regression Equations and Analysis of Variance

for the variables in Data of Black (1)

The data that was accumulated was stored on a floppy disc and was not printed with this
thesis. Access to the data is possible by contacting Mr. Franz Sigg.

Computer Analysis Using Minitab:

CI C2 C3 C4 C5 C6 C7 C8

1 0.18 0.08 110 79.4 83.4 6.9 0.004107

2 0.18 0.06 120 75.4 82.5 0.0 0.015952
3 0.20 0.08 90 78.7 83.0 6.3 0.019801
4 0.18 0.07 120 75.2 81.5 2.6 0.005973
5 0.18 0.08 95 80.4 83.3 9.3 0.004107
6 0.18 0.06 100 74.7 80.6 1.6 0.015952
7 0.18 0.07 130 76.9 81.4 4.1 0.005973
8 0.18 0.08 120 80.9 83.4 9.4 0.004107
9 0.19 0.06 125 76.8 81.2 6.3 0.018495
10 0.20 0.08 120 80.6 83.6 11.8 0.019801

C2: Circumferential Register

C3: Lateral Register
C4: Doubling Angle
C5: Dot Gain on 50% Tint
C6: Average Dot Gain on Concentric Circles
C7: Magnitude of Doubling
C8: Magnitude of Misregister

Correlation C2-C8

C2 C3 C4 C5 C6 C7
C3 0.096
C4 0.010 -0.086

C5 0.113 0.213 0.175

0.113 0.175 0.126 0.838
C6
0.38C) 0.595 0.552
C7 0.040 -0.020

0.02C -0.092 -0.118 0.061

C8 -0.116 -0.497
 58

Analysis of Variance and Estimation of the Population Regression Line with the
Data from Black 1

The Regression Equatiorlis: C7 =

3.31 + 9.0 c:

Predictor Coef Stdev t-ratio P

Constant 3.307 1.838 1.80 0.073
C2 9.02 10.15 0.89 0.375

s =
2.680 R-sq =
0.2% R-sq(adj) =
0.0%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 5.672 5.672 0.79 0.375
Error 498 3577.543 7.184
Total 499 3583.215

C2: Horizontal Register

C7: Magnitude of Doubling

The Regression Equation is: C7 =

5.05 -

1.44 C3

Predictor Coef Stdev t-ratio p

Constant 5.0459 0.2708 18.63 0.000
C3 -1.445 3.197 -0.45 0.652

s = 2.682 R-sq = 0.0% R-sq(adj) = 0.0%

Analysis of Variance

SOURCE DF SS MS F P
1.468 1.468 0.20 0.652
Regression 1
Error 498 3581.746 7.192
Total 499 3583.215

C3: Vertical Register

C7: Magnitude of Doubling
 59

The Regression Equation is: C7 = 3.44 + 0.0160 C4

Predictor Coef Stdev t-ratio P

Constant 3.4437 0.1970 17.48 0.000
C4 0.015968 0.001742 9.17 0.000

s = 2.481 R-sq =
14.4% R-sq(adj) =
14.3%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 517.35 517.35 84.04 0.000
Error 498 3065.86 6.16
Total 499 3583.21

C4: Doubling Angle

C7: Magnitude of Doubling

The Regression Equation is: C7 = -

38.8 + 0.566 C5

Predictor Coef Stdev t-ratio P

Constant -38.761 2.646 -14.65 0.000
C5 0.56555 0.03422 16.53 0.000

2.156 35.4% R-sq(adj) = 35.3%

s =
R-sq =

Analysis of Variance

SOURCE DF SS MS F P
1269.1 1269.1 273.12 0.000
Regression 1
Error 498 2314.1 4.6
Total 499 3583.2

C5: Dot Gain on 50% Tint

C7: Magnitude of Doubling
 60

The Regression Equation is: C7 = -

55.8 + 0.739 C6

Predictor Coef Stdev t-ratio P

Constant -55.828 4.115 -13.57 0.000
C6 0.73917 0.05004 14.77 0.000

s =
2.237 R-sq =
30.5% R-sq(adj) =
30.3%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 1091.7 1091.7 218.19 0.000
Error 498 2491.6 5.0
Total 499 3583.2

C6: Average Dot Gain on Concentric Circles

C7: Magnitude of Doubling

The Regression Equation is: C7 =

4.76 + 6.19 C8

Predictor Coef Stdev t-ratio p

Constant 4.7560 0.1780 26.72 0.000
C8 6.192 4.526 1.37 0.172

s =
2.677 R-sq = 0.4% R-sq(adj) =
0.2%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 13.417 13.417 1.87 0.172
Error 498 3569.798 7.168
Total 499 3583.215

C7: Magnitude of Doubling

C8: Magnitude of Misregister
 61

The Regres sion Equatitan is: C5 = -

19.8 + 1.18 C6

Predictor Coef Stdev t-ratio P

Constant -19.818 2.833 -6.99 0.000
C6 1.18099 0.03446 34.28 0.000

s =
1.540 R-sq =
70.2% R-sq(adj) = 70.2%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 2786.7 2786.7 1174.82 0.000
Error 498 1181.3 2.4
Total 499 3967.9

C5: Dot Gain on 50% Tint

C6: Average Dot Gain on Concentric Circles

The Regression Equation is: C4 =

91.6 + 63 C8

Predictor Coef Stdev t-ratio P

Constant 91.627 4.242 21.60 0.000
C8 63.3 107.9 0.59 0.557

s =
63.81 R-sq = 0.1% R-sq(adj) = 0.0%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 1404 1404 0.34 0.557
Error 498 2027651 4072
Total 499 2029055

C4: Doubling Angle

C8: Magnitude of Misregister
 62

Table of Regression Equations and Analysis of Variance

for the variables in Data of Magenta (1)

Computer Analysis Using Minitab:

CI C2 C3 C4 C5 C6 C7 C8

1 0.08 -0.04 70 78.5 85.1 3.9 0.041445

2 0.08 -0.02 120 80.6 84.7 4.5 0.022111
3 0.08 -0.03 60 79.4 84.9 4.0 0.031675
4 0.08 0.00 85 83.1 84.7 6.5 0.007754
5 0.08 -0.02 50 79.8 85.0 1.3 0.022111
6 0.08 0.00 80 80.9 84.8 5.5 0.007754
7 0.08 -0.01 90 79.1 85.2 3.0 0.013211
8 0.08 -0.02 75 79.9 84.9 5.5 0.022111
9 0.10 0.00 80 78.0 83.8 4.2 0.027729
10 0.08 -0.01 80 81.1 84.9 6.4 0.013211

C2: Circumferential Register

C3: Lateral Register
C4: Doubling Angle
C5: Dot Gain on 50% Tint
C6: Average Dot Gain on Concentric Circles
C7: Magnitude of Doubling
C8: Magnitude of Misregister

Correlation C2-C8

C3 C4 C5 C6 C7
C2
C3 0.357
C4 0.251 0.012
C5 -0.099 0.450 0.115
0.369 0.791
C6 -0.330
-0.149

0.164 0.463 0.463 0.150

C7 0.199
0.295 0.277 0.041
C8 -0.018 0.501 -0.041
 63

Analysis of Variance and Estimation of the Population Regression Line with the
Data from Maaenta 1

The Regression Equation is: C7 = 2.76+ 15.4 C2

Predictor Coef Stdev t-ratio P

Constant 2.7557 0.2560 10.77 0.000
C2 15.375 3.388 4.54 0.000

s =
1.666 R-sq =
4.0% R-sq(adj) =
3.8%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 57.157 57.157 20.59 0.000
Error 498 1382.408 2.776
Total 499 1439.565

C2: Horizontal Register

C7: Magnitude of Doubling

The Regression Equation is: C7 =

3.86 + 5.26 C3

Predictor Coef Stdev t-ratio P

Constant 3.86321 0.07501 51.50 0.000
C3 5.264 1.420 3.71 0.000

s = 1.677 R-sq = 2.7% R-sq(adj) = 2.5%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 38.677 38.677 13.75 0.000
Error 498 1400.888 2.813
Total 499 1439.565

C3: Vertical Register

C7: Magnitude of Doubling
 64

The Regression Equation is: C7 =

1.95 + 0.0256 C4

Predictor Coef Stdev t-ratio P

Constant 1.9453 0.1781 10.92 0.000
C4 0.025631 0.002199 11.65 0.000

s =
1.507 R-sq =
21.4% R-sq(adj) =
21.3%

Analysis of Variance

SOURCE DF SS MS
Regression 1 308.51 308.51 135.84 0.000
Error 498 1131.05 2.27
Total 499 1439.57

C4: Doubling Angle

C7: Magnitude of Misregister

The Regression Equation is: C7 = -26.1+ 0.370 C5

Predictor Coef Stdev t-ratio P

Constant -26.141 2.574 -10.15 0.000
C5 0.37009 0.03174 11.66 0.000

s = 1.507 R-sq = 21.4% R-sq(adj) =

21.3%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 308.75 308.75 135.97 0.000
Error 498 1130.82 2.27
Total 499 1439.57

C5: Dot Gain on 50% Tint

C7: Magnitude of Doubling
 65

The Regression Equation is: C7 = -

13.1 + 0.198 C6

Predictor Coef Stdev t-ratio P

Constant -13.061 5.017 -2.60 0.010
C6 0.19802 0.05868 3.37 0.001

s = 1.681 R-sq =
2.2% R-sq(adj) = 2.0%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 32.184 32.184 11.39 0.001
Error 498 1407.381 2.826
Total 499 1439.565

C6: Average Dot Gain on Concentric Circles

C7: Magnitude of Doubling

The Regression Equation is: C7 =

3.77 + 2.09 C8

Predictor Coef Stdev t-ratio p

Constant 3.7692 0.1315 28.66 0.000
C8 2.094 2.298 0.91 0.363

s = 1 .699
R-sq = 0.2% R-sq(adj) =
0.0%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 2.396 2.396 0.83 0.363
Error 498 1437.169 2.886
Total 499 1439.566

C7: Magnitude of Doubling

C8: Magnitude of Misregister
 66

The Regression Equation is: C5 =

-30.9+ 1.31 C6

Predictor Coef Stdev t-ratio p

Constant -30.945 3.886 -7.96 0.000
C6 1.31050 0.04545 28.83 0.000

s =
1.302 R-sq =
62.5% R-sq(adj) =
62.5%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 1409.7 1409.7 831.23 0.000
Error 498 844.5 1.7
Total 499 2254.2

C5: Dot Gain on 50% Tint

C6: Average Dot Gain on Concentric Circles

The Regression Equation is: C4 =

76.8 38.3 C8

Predictor Coef Stdev t-ratto P

Constant 76.763 2.375 32.32 0.000
C8 -38.26 41.50 -0.92 0.357

s = 30.68 R-sq = 0.2% R-sq(adj) = 0.0%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 800.3 800.3 0.85 0.357
Error 498 468823.4 941.4
Total 499 469623.7

C4: Doubling Angle

C8: Magnitude of Misregister
 67

Table of Regression Equations and Analysis of Variance

for the variables in Data of Black (2)

Computer Analysis Using Minitab:

CI C2 C3 C4 C5 C6 C7 C8

1 0.21 -0.01 60 77.3 82.5 9.1 0.018170

2 0.21 -0.03 20 73.9 80.6 3.1 0.008565
3 0.21 -0.02 60 76.1 81.8 6.5 0.010087
4 0.21 -0.02 65 74.0 79.0 0.0 0.010087
5 0.21 -0.01 80 75.7 81.3 11.1 0.018170
6 0.21 -0.03 40 73.9 80.0 4.3 0.008565
7 0.20 -0.03 70 76.0 80.1 7.0 0.004212
8 0.21 -0.02 20 72.6 79.3 4.0 0.010087
9 0.20 -0.02 140 79.6 84.6 5.6 0.006793
10 0.21 -0.03 155 78.7 82.9 7.2 0.008565

C2: Circumferential Register

C3: Lateral Register
C4: Doubling Angle
C5: Dot Gain on 50% Tint
C6: Average Dot Gain on Concentric Circles
C7: Magnitude of Doubling
C8: Magnitude of Misregister

Correlation C2-C8

C3 C4 C5 C6 C7
C2
C3 0.272
C4 -0.018 -0.002

C5 -0.087 -0.330 0.310

0.194 0.905
C6 -0.048 -0.273

0.429 0.748 0.613

C7 0.039 -0.138

0.157 0.004 -0.041 0.172

C8 -0.067 0.091
 68

Analysis of Variance and Estimation of the Population Regression Line with the
Data from Black 2

The Regressioni Equation is C7 = 1.92+ 15.8 C2

Predictor Coef Stdev t-ratio P

Constant 1.922 3.644 0.53 0.598
C2 15.84 18.01 0.88 0.379

2.950 R- 0.2%
s =
sq
=
R-sq(adj) = 0.0%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 6.735 6.735 0.77 0.379
Error 498 4333.347 8.701
Total 499 4340.082

C2: Horizontal Register

C7: Magnitude of Doubling

The Regression Equation is: C7 = 4.06 -

40.3 C3

Predictor Coef Stdev t-ratio p

Constant 4.0604 0.3678 11.04 0.000
C3 -40.33 13.01 -3.10 0.002

s = 2.924 R-sq = 1.9% R-sq(adj) = 1.7%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 82.142 82.142 9.61 0.002
Error 498 4257.940 8.550
Total 499 4340.082

C3: Vertical Register

C7: Magnitude of Doubling
 69

The Regression Equation is: C7 =

3.34 + 0.0201 C4

Predictor Coef Stdev t-ratio P

Constant 3.3390 0.2067 16.15 0.000
C4 0.020121 0.001901 10.59 0.000

s =
2.667 R-sq =18.4% R-sq(adj) =
18.2%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 797.23 797.23 112.06 0.000
Error 498 3542.85 7.11
Total 499 4340.08

C4: Doubling Angle

C7: Magnitude of Doubling

The Regression Equation is: C7 = -

65.4 + 0.924 C5

Predictor Coef Stdev t-ratio P

Constant -65.428 2.804 -23.33 0.000
C5 0.92361 0.03669 25.17 0.000

s =
1.958 R-sq =
56.0% R-sq(adj) =
55.9%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 2430.0 2430.0 633.53 0.000
Error 498 1910.1 3.8
Total 499 4340.1

C5: Dot Gain on 50% Tint

C7: Average Dot Gain on Concentric Circles
 70

The Regression Equation is: C7 = -

69.4 + 0.920 C6

Predictor Coef Stdev t-ratio P

Constant -69.352 4.302 -16.12 0.000
C6 0.91965 0.05311 17.32 0.000

s =
2.332 R-sq = 37.6% R-sq(adj) =
37.5%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 1631.0 1631.0 299.83 0.000
Error 498 2709.0 5.4
Total 499 4340.1

C6: Average Dot Gain on Concentric Circles

C7: Magnitude of Doubling

The Regression Equation is: C7 =

4.39 + 72. 1 C8

Predictor Coef Stdev t-ratio p

Constant 4.3858 0.2303 19.04 0.000
C8 72.12 18.52 3.89 0.000

2.908 3.0% R-sq(adj) = 2.8%

s =
R-sq =

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 128.24 128.24 15.16 0.000
Error 498 4211.85 8.46
Total 499 4340.08

C7: Magnitude of Doubling

C8: Magnitude of Misregister
 71

The Regression Equation is: C5 = -

12.7+ 1.10 C6

Predictor Coef Stdev t-ratio p

Constant -12.685 1.877 -6.76 0.000
C6 1.09990 0.02317 47.48 0.000

s =
1.017 R-sq =
81.9% R-sq(adj) =
81.9%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 2333.1 2333.1 2253.95 0.000
Error 498 515.5 1.0
Total 499 2848.6

C5: Dot Gain on 50% Tint

C6: Average Dot Gain on Concentric Circles

The Regression Equation is: C4 = 74.5 + 1399 C8

Predictor Coef Stdev t-ratio P

Constant 74.454 4.919 15.14 0.000
C8 1399.1 395.5 3.54 0.000

s = 62.11 R-sq = 2.5% R-sq(adj) = 2.3%

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 48261 48261 12.51 0.000

Error 498 1920954 3857

Total 499 1969216

C4: Doubling Angle

C8: Magnitude of Misregister
 72

Table of Regression Equations and Analysis of Variance

for the variables in Data ofMagenta (2)

Computer Analysis Using Minitab:

CI C2 C3 C4 C5 C6 C7 C8

1 0.12 -0.03 100 75.3 78.8 1.3 0.0207019

2 0.11 -0.03 65 75.5 79.9 4.5 0.0123194
3 0.11 -0.03 80 74.7 78.7 2.0 0.0123194
4 0.11 -0.03 85 76.1 80.9 6.8 0.0123194
5 0.11 -0.04 75 74.7 79.0 1.2 0.0089534
6 0.11 -0.04 80 76.1 79.4 5.1 0.0089534
7 0.11 -0.04 90 74.1 79.9 0.7 0.0089534
8 0.10 -0.02 90 77.7 80.6 4.4 0.0186163
9 0.11 -0.03 0 75.3 80.4 3.0 0.0123194
10 0.10 -0.03 110 77.8 81.6 3.6 0.0086582

C2: Circumferential Register

C3: Lateral Register
C4: Doubling Angle
C5: Dot Gain on 50% Tint
C6: Average Dot Gain on Concentric Circles
C7: Magnitude of Doubling
C8: Magnitude of Misregister

Correlation C2-C8

C2 C3 C4 C5 C6 C7
C3 0.039
C4 0.025 0.194
C5 -0.179 -0.365 0.097
C6 -0.165 -0.414 -0.125 0.877
C7 -0.033 0.174 0.491 0.315 0.165
C8 0.111 0.015 0.053 -0.131 -0.186 -0.081
 73

Analysis of Variance and Estimation of the Population Regression Line with the
Data from Magenta 2

The Regressioni Equation is:: C7 ==

4.19 5.56 C2

Predictor Coef Stdev t-ratio P

Constant 4.1899 0.7784 5.38 0.000
C2 -5.562 7.661 -0.73 0.468

1.620 R- 0.1% 0.0%

s =
sq =
R-sq(adj) =

Analysis of Variance

SOURCE DF SS MS F p
Regression 1 1.384 1.384 0.53 0.468
Error 498 1307.306 2.625
Total 499 1308.690

C2: Horizontal Register

C7: Magnitude of Doubling

The Regression Equation is: C7 = 4.26 + 16.3 C3

Predictor Coef Stdev t-ratio p

4.2557 0.1743 2441 0.000
Constant
16.290 4.123 3.95 0.000
C3

3.0% R-sq(adj) = 2.8%

s = 1.596 R-sq =

Analysis of Variance

DF SS MS F p
SOURCE
39.774 39.774 15.61 0.000
Re^sSon
Regression 1
^ ^^ ^
499 1308.690
Total

C3: Vertical Register Variation

C7: Magnitude of Doubling
 74

The Regression Equation is: C7 =

1.89 + 0.0248 C4

Predictor Coef Stdev t-ratio p

Constant 1.8920 0.1517 12.48 0.000
C4 0.024810 0.001971 12.58 0.000

s = 1.412 R-sq =
24.1% R-sq(adj) =
24.0%

Analysis of Variance

SOURCE DF SS MS
Regression 1 315.76 315.76 158.37 0.000
Error 498 992.93 1.99
Total 499 1308.69

C4: Doubling Angle

C7: Magnitude of Misregister

The Regression Equation is:

C7=- 19.5 + 0.296 C5

Predictor Coef Stdev t-rauo P

Constant -19.468 3.118 -6.24 0.000
C5 0.29645 0.04002 7.41 0.000

1.539 9.9% R-sq(adj) = 9.7%

s =
R-sq =

Analysis of Variance

SOURCE DF SS MS
1 129.90 129.90 54.88 0.000
Regression
Error 498 1178.79 2.37
Total 499 1308.69

C5: Dot Gain of 50% Tint

C7: Magnitude of Misregister
 75

The Regression Equation is: C7=- 11.7 + 0.187 C6

Predictor Coef Stdev t-ratio P

Constant -11.677 4.109 -2.84 0.005
C6 0.18663 0.05010 3.72 0.000

s =
1.599 R-sq =
2.7% R-sq(adj) =
2.5%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 35.475 35.475 13.88 0.000
Error 498 1273.216 2.557
Total 499 1308.690

C6: Average Dot Gain on Concentric Circles

C7: Magnitude Of Misregister

The Regression Equation is: C7 = 3.86- 13.5 C8

Predictor Coef Stdev t-ratio P

Constant 3.8594 0.1476 26.15 0.000
C8 -13.497 7.480 -1.80 0.072

s = 1.616 R-sq = 0.6% R-sq(adj) = 0.5%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 8.500 8.500 3.26 0.072
Error 498 1300.190 2.611
Total 499 1308.690

C7: Magnitude of Doubling

C8: Magnitude of Misregister
 76

The Regression Equation is: C5 =

-8.75+ 1.06 C6

Predictor Coef Stdev t-ratio p

Constant -8.755 2.126 -4.12 0.000
C6 1.05678 0.02592 40.78 0.000

s =
0.8271 R-sq =
77.0% R-sq(adj) =
76.9%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 1137.4 1137.4 1662.64 0.000
Error 498 340.7 0.7
Total 499 1478.1

C5: Dot Gain on 50% Tint

C6: Average Dot Gain on Concentric Circles

The Regression Equation is: C4 =

66.9 + 176 C8

Predictor Coef Stdev t-ratio P

Constant 66.911 2.927 22.86 0.000
C8 176.1 148.4 1.19 0.236

s = 32.05 R-sq =
0.3% R-sq(adj) ==
0.1%

Analysis of Variance

SOURCE DF SS MS F P
Regression 1 1447 1447 1.41 0.236
Error 498 511551 1027
Total 499 512998

C4: Doubling Angle

C8: Magnitude of Misregister
 77

APPENDIX D

PRESS RUN DOCUMENTATION

Specifications of Web Test Run:

Set-up: Makeready plus Actual Run

Press: Harris M-1000 -

Blanket to Blanket, Duotrol Dampening

System, TEC 2 Zone Dryer
Paper: Great Northern Paper

Color: Blue White

Basis Weight: 38.8

35"
Size:

Finish: Gloss

Roll#: 7U40504

Goundwood

Plates: 3M Viking
Fountain Solution: ROSOS KSP 500 M4 AS 6 1/2 ounces to a

gallon of H20 pH 3.8, Conductivity 2450

Blanket: Reeves 718's 3 ply Compressible Blanket

Ink: GPI ,
Black-Magenta-Cyan

Final trim size of signature: 8 1/2x11

Densities: Tolerance of + or -
.05

M: 1.35

C 1.35

K 1.6
system turned off at this unit only
Y Blank plate with the dampening
 78

Press Conditions: 1200 feet / min (Equivalent to 37613 impressions per hour)

Infeed Tension: 133 pounds

Conditions:
1- Web Aligner and Register Control were on automatic

2- Web Aligner and Register Control were turned off

 79

Sheetfed Test for SLUR

Press: Heidelberg MO 19 x 25 m Single Color Sheetfed, Bearer to Bearer

Ink: Black Offset Sheetfed Super Ultraset / 1/D Dense Black manufactured
by
Morrison Printing Ink

Lithographic Fountain Solution: 2 ounces Seamist Fountain mixed with one gallon

water with additional Fungus Arrester. Manufactured by Anchor /

Lithkemko

Paper: Dull-coated

Size: 17 1/2 x 22 1/2 Weight: 70 -

58M
Color: White Grain: Long Caliper: or. 10 mm

-
Plate: 3M Viking Gl exposed for 40 units and machine processed

Blanket: Compressible

Speed: 5500 impressions / hr

Normal Set-up: Pressure between the plate and blanket cylinder is .07 to .10 mm

(.0028-.004")
Best printing result as recommended by Heidelberg is achieved with the rubber

blanket packed to .05 mm (.002") under bearer height.

Impression Pressure between blanket and impression cylinder should be .10 mm

(0.004"). Normally this is taken care of with the thickness of the paper.

Plate + Underlay = .65mm

(.026") which brings it to .15mm
(.006") above

bearers. The squeeze pressure would be .10mm (.004").

 80

Phase I: The above normal

set-up was used, however the scale setting on the impression
cylinder was increased from 0 to (.004")
.1 mm and gradually increased to .12 mm

(.0048")
Solid Ink Density was in the range of 1.3 1.4
Blanket +
Underlay =
.03 mm (.0012") under bearer height instead of .05 mm

(.002")

Phase II: (.008") of plate was removed and added to the blanket
.2mm

packing packing.

This meant that the plate was now .05mm

(.002") below bearers.
Blanket is now packed above bearers by .17mm
(.0065")
No adjustment of impression pressure even though it was excessive due to the

increased blanket packing.

Phase III: An additional .075mm

(.003") was removed from the plate and added to the

blanket.

Plate =
.13mm
(.005") under bearer.
Blanket =
.24mm (.0095").
Impression pressure was decreased at one point While moving or
adjusting the

impression pressure, directional dot gain was noticed on the sheets because of the

movement of cylinders. However once the impression pressure was stabilised,

there was no sign of directional gain on the targets.

Mottle was noticed on the printed sheets.

Phase IV: Decreased impression pressure to .08 mm (.0035") then increased back to

.1 mm.

Mottle was present. A most probable cause would have been a lack of

sufficient squeeze pressure.

A noticeable change or increase in image length due to decreasing plate packing.

 81

APPENDIX E

FILMS FOR LAYOUT

1- GATF Star Target 36 wedge

RIT Star Target 72 wedge Size: .75 x

2- GATF Dot Gain Scale 2.5"

and Slur Gauge Size: .5 x

3- Kodak Customized Color Analysis

1.5x4.75"

Size:

1.25"
4- GATF Ladder Target Size: width of

1.4"
5- RTT Experimental Size: 1.4
Doubling Target x

1/2" 1/8"
6- Image Size: 5 x 4

1/2"
7- RTT Visual Registration Target Size: 3 x

1.75x1.75"
8- RTT Registration Target Size:

9- Parallel Line Tints Size: 8 x 24 mm

1 2 3 4 5 e 7 8 9 lO 11 12 13 20u 25u 30u 40u 55u 70i 90'
90% 100% D 995% 99% 9fi% 97% 96%

o
RIT COLOR TEST STRIP

v^PGE"

HIT DOUBLING

""
141
160

^^^fl
9 20

40
I.M

120 f
V

100 ,

'
BO

6C z
r
...
40 "^^B
20 160
0
Rochester -
olTecnnc
1

o CvJ

0) lO
c6S

D FUT DOUBLING TARGET

Hi "C_S_'.G
~i=G=~ .-

|
160
0
160 20 20
01
140 R^^40 *^
140 ^^^^40
01

12c ^M ^^L e
C 2: ^^k 6C

100 ^H o 100 i

LO

80 '
ao-l
d
60"^H
z
6C
^^r 'z

w.--
=*- 40^

20 15
20 16 0
6 Rochester irv mute o' Technology 1
Rochester in;titute of
Technology 1

Q
LU

2 H
^
o
OO CE
D <

::_:_:
'--IE" *
=C o
140 ^^^H
160 ?
kL
^^^-:
I Q -I

!
oo
120 ^^k & ^ o

100-j

80 1

60 ^^r ;

^,.
40 *^^H ^^^140
20 160

3456789 10
 4%
8 8 8 S 20b 30% 40% 5" fe 1% 2% 3%

'
!
12 3 4 ~7 8 9 lO 11 12 13 20u Zbu 30u 40u 55u 70u 60% 70% B0% 90% 100% 90 0 99 5% 99% 98% 97%

0
RIT COLOR TEST STRIP

5 6 7 8 9 10