Improving quality consistency and productivity of color separators utilizing the Dupont image manager

RIT Scholar Works

Theses

Thesis/Dissertation Collections

5-18-1990

Improving quality consistency and productivity of

color separators utilizing the Dupont image

manager

Enrique A. Arce B.

Follow this and additional works at:

http://scholarworks.rit.edu/theses

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.

Recommended Citation

OF

COLOR SEPARATIONS

UTILIZING THE

DUPONT

IMAGE

MANAGER

by

Enrique

A. Arce

B.

A

thesis

submitted

in

partial

fulfillment

of

the

requirements

for

the

degree

of

Master

of

Science in

the

School

of

Printing

Management

and

Science in

the

College

of

Graphic

Arts

and

Photography

at the

Rochester Institute

of

Technology.

School of printing Management and Sciences

Rochester Institute of Technology

Rochester, New York

U.S.A.

CERTIFICATE OF APPROVAL

MASTER'S THESIS

This is to certify that the Master's Thesis of

Enrique A. Arce B.

With 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 18, 1990.

Thesis Committee:

Miles

SOllthworth

Thesis Advisor

Joseph L Noga

Graduate Program Coordinator

Utilizing

the

Du Pont Image Manager

I,

Enrique A. Arce

B,

preferto

be

contacted eachtimea request

for

reproduction

is

made.

I

can

be

reached atthe

following

address:

Edison 4125

Santiago

Chile

A

Cecilia

y

Natalia.

..

...mi

hermana

e

hija.

(To Cecilia

and

Natalia...

I

would

like

toexpress

my

thankfulness

to

all

those

who

in

one

way

oranother

helped

me

to get

through

of

RIT

and

this

research.

To

the

instructors,

my classmates,

scanner

operators,

judges,

andthose

I

am not

mentioning, thank

you.

I

would also

like

to

thank

Miles

Southworth

for his

help,

not

only

in

this thesis where

he

was

the

major

advisor,

but

also

in

the

lessons

he

gavetome

in

the

Graphic

Arts field.

A

special mention also

for

Charles Layne for

giving

invaluable

help

not

only in

theresearch

and

statistics,

but

also

editing this

thesis tomake

my English

a readable matter.

I

also wantto thank

very

muchto

my

family,

my

wife

Lizabeth,

my

daughter

Natalia,

my

parents

Enrique

and

Orlanda,

and

my

sister

Cecilia,

for

their

love

and encouragementto continuewhen

being

far from

home

was almost

impossible

toget

through,

and

for

helping

meto

finish

this

degree.

Finally

I

want

to

express

my

gratitudeto thepeoplethatconform

COCHRANE

S.A.,

the

Page

ABSTRACT

1

CHAPTER I

-

INTRODUCTION

3

Endnotes

for

chapter one

6

CHAPTER II

-

THEORETICAL

BASIS

FOR

THE STUDY

7

Tone

reproduction

7

Gray

balance

9

Color

correction

9

Unsharp

masking

9

Endnotes for

chapter

two

11

CHAPTER

III

-

LITERATURE

REVIEW

12

Presenters

13

Presetter/previewers

14

The Du Pont

Image Manager

16

Endnotes

for

chapterthree

18

CHAPTER

IV

-

STATEMENT OF

THE PROBLEM

19

Hypotheses

20

CHAPTER V

-

METHODOLOGY

21

Experimental

design

21

Subjective

evaluation

23

Equipment

or resources

24

Limitations

ofthe

study

25

Endnotes

for

chapter

five

26

CHAPTER VI

-

THE

RESULTS

27

Hypothesis

1

27

Hypothesis

2

28

Judges

preference

for

quality

29

Summary

32

Hypothesis 1

33

Hypothesis 2

33

Separation

preference

33

Implications

34

Recommendations

for further

investigation

34

BIBLIOGRAPHY

35

APPENDIX

A

39

APPENDIX

B

43

APPENDIX C

48

APPENDIX D

50

APPENDIX E

57

APPENDIX

F

63

Page

Table

1

:

Density

for

the

process

inks

9

Table

2:

Productivity

improvement

caused

by

theuse ofthe

Du Pont

Image

Manager

asthemethod

for

thescanner

set-up,

comparedto theoperator's

judgment

method

29

Table 3: Judges

preferredtheuse ofthe

Du Pont Image Manager

asthemethod of scanner

set-up

comparedto the

operator's

judgment

30

Table

C-

1

_:

Calibration

_values

for

_the

Cromalin

_systemused

49

Table

D-

1

:

Gradation

curves used

in

the

Image Manager for

the thesis

51

Table

F-

1

_:

Rejection

_rate

based

_{onthe tonereproduction criteria}

for

_the

1 2 judges

64

Table

F-

1

_:

Rejection

_rate

based

_{onthe tone}reproduction criteria

for

the

12 judges

(cont.)

65

Table F-2: Color

separationpreference

based

ontonereproductioncriteria

for

the

12 judges

66

Table F-3:

Time

required

to

perform

the

color separation

by

thesix scanner

operators

67

Table F-4: An

example ofthe

Scott's

coefficientevaluation

68

Table F-5: Inter-judge

agreement results

for

the

12 judges

68

Table F-6: Rejection

rate

based

on

the

tonereproduction criteria

for

the

six

final

judges

69

Table

F-7: Color

separationpreference

based

ontonereproductioncriteria

for

the

six

final

judges

70

Table

F-8: Comparison

time

for

the

scannersettings

using

eitherthe

operator's

judgment

or

the

instructions

from

the

Du Pont

[image:9.543.66.478.148.662.2]

operators

73

Table G-2: Results

ofthechi-square analysis

for

the

inexperienced

scanner

operators

74

Table G-3: Results

ofthechi-square analysis

for

the

different

originals

(regardless

thescanner

operators'

skill

level)

74

Table G-4: Results

ofthe

ANOVA

analysis

for

thescanner

set-up

time

for

the

experienced scanner operators

75

Table G-5: Results

ofthe

ANOVA

analysis

for

the

scanner

set-up

time

for

the

inexperienced

scanner operators

75

Table G-6: Results

of

the

ANOVA

analysis

for

therejection rate

for

the

experienced scanner operators

75

Table

G-7: Results

of

the

ANOVA

analysis

for

therejection rate

for

the [image:10.543.68.477.68.340.2]

Page

Figure

1

:

Jones Diagram

8

Figure

D-

1:

Image Manager

instructions

for

_thescanner

set-up

for

the

high

key

original

52

Figure

D-2:

Image Manager

instructions

for

thescanner

set-up for

the

low

key

original

53

Figure D-3: Image

Manager

instructions

for

thescanner

set-up

for

the

normal

key

original

54

Figure D-4: Image Manager instructions for

thescanner

set-up

for

theunder

exposed original

55

Figure D-5: Image Manager instructions

for

thescanner

set-up

for

theover

exposed original

56

Figure E-l:

ASPIRIN (HIGH

KEY)

58

Figure

E-2: SUNSET

(LOW KEY)

59

Figure E-3: VOLCAN OSORNO

(NORMAL

KEY)

60

Figure

E-4: VOLCAN OSORNO

(UNDER

EXPOSED)

61

[image:11.543.62.477.150.454.2]

The

magic

key

in any

commercial

transaction

is

to

satisfy

thecustomer's requirements.

In

the

color separation marketthecustomer's expectationsare

consistency

in

the

product,

and

films

thatwhen printed will reproduce a clean visual matchto theoriginalcopy.

The

variablesthataffectthe

quality, consistency,

and

productivity

for

thecolor separation

process,

werereviewed.

The

resultofthisreviewwasthat theoperator's scanner

set-up

for

tonereproduction

is

themost critical variable

in

theprocess.

An

estimateofthetotalcost

in

the

United

States

caused

by

rescans

due

to

inconsistency

ofthe scanner

set-up

gave the

result of

$462,852,000

peryear.

One

solutionto theproblem of

inconsistency

ofthe scanner

set-up

seemsto

be

theuse of

pre-scan systems.

The

pre-scan system

helps

thescanneroperator

determinate

theoptimum

tone

reproduction

for

eachcolorseparation.

A

reviewoftherelated

literature described

seven presetterand sevenpresetter/previewer

systems.

To

testpre-scaneffectiveness the

DuPont

Image Manager

was used toproduce the suggested scanner

set-up using

a computerized method

for

the tone reproduction

determination

with

very

little

operator assistance.

This study

answered

the

research

question,

does

the

DuPont Image Manager improve

the

quality, consistency,

and

productivity

of color separations made on a

Hell DC 399

Scanner?

The

Hell DC

399

scanner was used.

Two

hypotheses,

and the experimental

design

arepresented.

The

experiment produced color separations

using

the

judgment

ofthreeexperienced and

three

inexperienced

scanner

operators,

and

by

using

thepre-scansystem.

With

the

films

obtained,

Cromalin

color proofs were made.

Judges

evaluated the proofs

using

pair comparison

to

study

the

quality consistency

improvement.

To

study

the

productivity,

the

averagetimerequired

for

thescanner

set-up using

both

methods was

measured,

using

the

scanner operator's

judgment

and

the

DuPont Image Manager.

The

resultsofthe

study

yieldedthe

following

conclusions:

- the

scanner operators.

- the

Du Pont Image Manager

produces an

improvement in

the

productivity

of

setting

up

a

scanner

in

aratio of

45%

for

the

inexperienced

scanneroperators.

-the

use ofthe

Du Pont

Image Manager

did

not

significantly increase

the color separation

acceptance either

for

experienced or

inexperienced

scanner

operators,

or

for different kind

of originals.

-

for

low

key

originals,

the operator's

judgment

produced a

better

overall visual color

separation,

compared

to

whenthe

Image

Manager

was used.

-

for high

key

originals there

is

no

difference

in

the visual preference

for

the color separations made

using

either oneofthe twomethods analyzed.

-

for

normal

key,

over

exposed,

and under exposed originals thevisual preference was

higher

for

those

color separationsthatwere made

following

the

Du Pont Image Manager

instructions

for

thescanner set-up.

The

recommendation

for further

study

is

that thesameresearchcan

be

performed

in

areal

working

situation,

at

five

or six

shops,

using

theirnormal

jobs in

the

test,

and

asking

their

INTRODUCTION

In any

commercial

transaction

customers wantto receivethe

best

quality

product

for

the

price

they

are askedtopay.

The

same

is

true

in

thecolorseparation and

printing

market.

But

thenpeople

find

themselveswiththeproblemof

how

to

define:

what

is

good

quality

in

thegraphic arts?.

Before going further it is better

topoint outthat the

only

stage

in

thecolor

reproduction process where

it is

possibleto control

the

nature ofthe

image

that will

be

produced

in

the

printing

stage

is

the color separation processQ).

Now,

going back

to the

original question

it

can

be

saidthat

quality

is

to

fulfill

thecustomerrequirements.

But

one

will

find

the

scanner

operators'

complainthat the customers

do

not

know

what

they

want,

or thatone customer wants

something

different

thananother customer(2).

One

will

find

some peoplethatwill

say

"Clean

and

Bright is

always

right"(3),

and others will

say

that

it is

relative to the subject and to the observer.

But

the onetrue

quality

attribute

everybody

agreeswith

is

thatcustomers want

consistency

from

scanto

scan,

scannerto

scanner,

and

from

operatortooperator(4).

How

can consistent colorseparations

be

produced?.

The

answer

to

thisquestionseems to

be

with aconsistent

input,

separation

process,

and output.

The

analysisoftheanswergiven

leads

oneto the

quality

ofthe

input

ororiginal.

Here

all

people

know

thata

large

originalwill give more

detail

and

less

graininess.

However

most photographersprefertouse

35

mm

film. If

theoriginal

is

seen

by

itself,

one can

say

ideally

it

should posses correct

exposure,

even

light,

color

balance,

good

contrast,

sharp

focus,

and

fine

grain(5).

If it is

asked,

how

many

transparencies that are used

for

the color

separation

satisfy

all those

requirements,

themostprobable answer will

be

very

few.

So

the

scanner operator

has

to

deal

withthesevariations

in

theoriginal.

The

separation process and output can

be

considered as an unique

point,

where the

equipment and materials should

be

standardized.

The

materials used

in making

a color

separation are

the

result ofan

industrial

process,

that

includes very

well controlled raw

variation.

When

theequipment

is

considered one sees some variation

from

scannerto

scanner,

but it

can

be

taken

into

consideration when each scanner

is

initially

calibrated.

The

scanner,

because

of

its

solid state

electronics,

can perform the

functions

of

analyzing

the

picture,

processing

the

job,

and

exposing

the

film

in

a

very

precise way(6).

Scanners

contain computers made

from

physical components which are more repeatable than chemical

components,

which are affected

by

temperature

and

humidity(6).

In

ordertocontrol the

relationship

between

thescanner's

exposing light

intensity

andtheresults obtained onthe

film,

it is

suggestedtocheckthescanner

linearization

at

least

once a

day(7).

By

periodically

doing

a

linearization

test thisguarantees consistent and predictable output(8).

This

means

thatwhen

the

scanner operator regulates the scannerto produce a

50%

dot,

he

will get a

50% dot

on

the

film.

So if everything

seemsto

be

consistent or nottoo

far away

from

what

is

expected,

why is

there

so much

difference from

scan to scan?.

The

answer

is

that there

is

some

inconsistency

in

the scanner set-up.

Now,

taking

a

look

at the

decisions

the scanner

operator

has

tomake

in setting

up

thescanner

many

tasksare

found. For

every

single

job

the scanner operator must

know

the

printing

characteristics such as the

dot

gain, gray

balance,

trapping,

printing sequence,

ink

hue,

ink

strength,

and papertype.

And for

the original

he

will

have

to

know

the emulsion

type,

original

size, exposure,

development,

density

distribution,

highlight

and shadow

densities,

color

balance,

graininess,

and reproduction size.

These

are

just

some ofthevariables

the

scanner operator must

handle.

The

scanner

set-up is

a

very

complex matrix of

decisions

based

on

many

variables.

Thus

one can

easily

understand

how

a scanneroperatorcan makethe

wrong set-up

decisions(9).

By

taking

a

look

at the color separation market one can get a good

idea

ofthe real

dimension

of

this

problem.

Usually

when a customer

is

unsatisfied withacolor

separation,

it is necessary

to

rescan the original.

This

not

only

will

be

an extra

cost,

but usually

the scanneroperator

has

no record of what

he

has done

before,

so

he has

to restartthewhole

process of

making

the

decisions

for

thenewscanner set-up.

Looking

at

the

extra cost caused

by

a

wrong

color

separation,

it

can

be

assumed

that

a good color separation can cost

$59,

anda not so good one can cost

$149(3).

It has been

reported

that the

averagerescan rate

is

as

large

as

30%(10).

One

can estimate

the

financial loss due

to poor scans.

In

1987

therewere

1 1,000

scanners worldwideand

3,000

of

those

were

in

the

assume the average production

for

a scanner

is 100

color separations per scanner per

week(ll),

with a rate of

2

operators perscanner* 12).

In

the

United States

there

is

atotal estimated

6,900

scanner operators.

This

means

there

are

6,900 different

criteria

for

setting

up

a scanner

that

can also

vary

from

day

to

day.

Then

it is

logical

thatonecould

find

some

inconsistency

in

the

color separation process.

Now,

evaluating

the totalcost caused

in

the

United States

by

rescans

due

to the

inconsistency

in

the scanner

set-up

one gets

$462,852,000

per year.

Even

with

these

variations

the

color separation market

is

stillprofitable.

But

one can see

that

it

could

be

a

lot better.

Looking

at

the

scanner

efficiency,

measured as the numberof

good scans made

during

each

shift,

one will

find

that

scanners spin

only

for

15

to

20

minutes per

hour

(13).

So

ascanner

working

rate of

25

to

33%;

and what

does

thescanner operator

do

the rest ofthe

time?.

He has

to

mount,

dismount,

develop

the

film,

and analyzethe

transparency,

which

is

themosttime

consuming

operation,

plus

set-up

the

100

or more controlsonthecolor scanner.

One

could assumethat

by having

an automatic

set-up

onthescannerthe rescan rate could

be

reduced.

This

is

the

track

the

scanner manufacturers are

trying

to

follow

nowadays(4).

But

there

is

still

the

problem of

having

over

10,000

scannersworldwidewithoutautomatic

set-up.

It is necessary

to

find

a practical solution

for

thosescanners.

A

pre-scan

device

can

help

thescanner operatortoevaluate the

original,

and give

him

the

directions for

setting

the

scanner's controls(14).

Some

authors report a rescan rate of

2.6% only

by

using

the

pre-scansystem(15).

Bruno is

moreconservative.

He

reports a rescan rate reduction

from

30%

to

20% using

a presetter

device,

and

from

30%

to

10%

for

a presetter/previewer

device(10).

In

this thesis one ofthese pre-scan

systems,

the

DuPont Image Manager

wasused.

The

final

results obtained

using

thissystem

for

colorseparationswere compared withthe result

of color separations made

by

experienced and

inexperienced

scanner operators without

any

assistance.

The

system performsan automatic analysisofthe

transparency

andsuggeststhe

optimum scanner

set-up

for

thecustomer's requirements.

This

research

has

answeredthe

question,

does

the

DuPont Image Manager

help

to

improve

the

productivity

and

quality

1.-

Field

Gary,

Color

and

its

Reproduction. Graphic Arts

Technical

Foundation,

Pittsburgh,

1988.

2.-

Southworth

Miles,

Quality

Control

Scanner,

Prepress

can

be

a controlled process,

Volume

9,

Number

10.

3.-

Southworth

Miles,

Quality

Control

Scanner,

Scanner

quality takes team work.

Volume

2,

Number

1 1.

4.-

Southworth

Miles,

Quality

Control

Scanner,

Color

reproduction past, present, and

future.

Volume

7,

Number

7.

5.-

Southworth

Miles,

Quality

Control

Scanner,

Original

transparencies:what's

best

for

reproduction?.

Volume

7,

Number 3.

6

.

-

KODAK

publication

No

Q

-

7 8

,

The Color Separation Scanner.

7.-

Color

Separation Class

Notes.

Summer

1988,

Prof. Joseph Noga.

8.-

Molla

Rafiqul,

Electronic Color

Separation,

R. K.

Printing

&

Publishing Co.,

West

Virginia,

1988.

9.-

Southworth

Miles,

Quality

Control

Scanner,

Color

scanner set-up:

kev

to control

quality .

Volume

9,

Number 6.

10.-

Bruno

Mike.

Status

_of

Printing

in

_the

U.S.A..

1987.

11.-

Bruno

Mike,

Status

_of

Printing

in

_the

U.S.A..

_up-dated

1988.

12.-

Southworth

Miles, Quality

Control

Scanner,

Color

control

for

cost and quality.

Volume

4,

Number 3.

13.-

Southworth

Miles,

Color Separation

Techniques,

Third

edition,

Graphic Arts

Publishing

Co., Livonia,

New

York,

1988.

14.-

Graphic Arts

Monthly,

October

1983,

From Camera

to

Scanner

and

Beyond,

p.

45-52,

Peter Johnston.

THEORETICAL

BASIS FOR THE STUDY

Good

color separationsarethose

that

contain

dot

values on the

film

which when

printed,

will yield

the

desired

colored

).

To

produce thisresult

the

scanner operatormust understand

the

basics

of color reproduction: tone

reproduction, gray

balance,

color

correction,

and

unsharp

masking.

Tone

Reproduction

is

the

relationship

ofthe

density

in

theoriginal to the

density

ofthe

reproduction and

the

percent

dot

printedonthe

final

product.

For

a

better

understanding

of

thisone

first has

to realizethat thephotographic original

has

agreater

capacity

oftones than

the

printing

process.

So

for

thereproduction one needs tocompressthe tonal rangeofthe

original(2).

The

most

important

point

here is

toreproduce

the

range withtonal

detail in

the

important

areas,

sacrificing

the tonal

detail in

areas

less important

to thepicture.

To

control thechanges

in

the

density

levels

ofanoriginal

image

when

it is

convertedto the

halftone film

andthenprinted one

has

to

know

certaincharacteristics.

These

characteristics

arethe original

image

andthe

printing conditions,

suchas

kind

of

ink,

solid

ink

density,

screen

ruling,

dot

shape,

paper

characteristics,

dot

gain,

ink

trapping,

press,

andso on(l).

Therefore,

one

may

conclude that the tone reproduction

is

specific

to

each press under

given

printing

conditions.

The

optimum

tone

reproductioncan

be

determined

by

constructing

a graphthatrelatesthe

densities

of anaverageoriginalto thoseone canprint.

Then,

to

determine how

toproduce

it

one relates thesize of

the

dots

onatest

film

to

theirrespective printed

densities using

a

plate/press graph.

Finally,

to

find

the required color separation curves the two graphs

mentioned above are puttogetheranda45straight

line

(transfer

curve).

Following

these

steps one can

build

the

Jones Diagram

(shown

in

figure

1),

and so

determine

the required

quadrant

I: desired

density

relationship

between

the

original and

its

reproduction

quadrant

II:

desired relationship

between

the

dot

size onthe

film

andthe

density

printed

by

those

dots,

plate/press graph

quadrant

TQ:

transfer

line

(45)

quadrant

IV: determine

the

dot

sizerequired

for

the

different

original

densities

toreach

the

[image:19.543.104.427.159.436.2]

measuredwiththevisual

filter

ofa

densitometer.

This gray

has

noapparentcolor_cast, even

though

it is

produced with

cyan,

magenta,

and yellow

dots

of

different

sizes(3).

The

gray

balance

correction

is necessary

because

of

the

hue

error the process

inks have. The

densities for

aset oftypicalprocess

inks

are:

filter

B

G

R

yellow

1.00

0.05

0.02

magenta

0.65

1.30

0.10

cyan

0.12

0.34

1.20

total

1.77

1.69

1.32

Table

1.-

Densities for

_process

inks

(source

Field, G.,1988,

p

229)

Here

oneseesthere

is

morered

light

reflected

(i.e.

lower

density)

than

blue

and green

light,

so

it

will produce a

brownish

cast(l).

It

is

necessary

toadjustthe

different dot

sizes

in

the

gray

areasto reproducethe same

density

for

the three

filters

for

given

ink

densities. Once

the correct

dot

sizestoproduce a neutral

gray

are

determined,

they

should not

vary

unless

the

printing

conditions change<2).

Color Correction is

needed

because

withtheavailable pigments

it is impossible

tomake

a

perfectly

pure

ink.

As

it

can

be

seen

in

table

1

, each

ink

absorbs some ofthe

light it

should transmit.

Color

correction

is

the compensation

for

unwanted absorptionof colors

caused

by

ink deficiencies(2).

It

is important

topoint out

that

color correction

does

not affectthe

gray

areas(3).

Color

correction

helps in

reproducing

the

memory colors,

like

green

grass,

blue

sky,

and redapples.

Scanners

have

twovariationsof colorcorrectionthatcan

be

accomplished atthe control

panel.

There

are adjustments

for

overallcolorcorrection usedtocompensate

for

any

overall

color change.

These

adjustments affect

every

color

in

whichthechanged color appears

(i.e.

a magenta adjustment will affect

every

pink,

red,

orange and purple).

There

are also

selective color controls thatpermit changes

in

a smaller

band

ofthecolor

spectrum,

and

make a change

in just

one

hue

similarto

the

color

desired

to

correct(4).

Unsharp Masking

accentuates thecontrast

between

adjacent

tones,

and

increases

the edge effects where tones change.

This

accentuation gives the appearance ofa sharper

picture with more

detail(4).

It increases

thesharpnessofthereproduction relative to the [image:20.543.175.368.162.260.2]

cylinderrotation.

Unsharp

masking is

possible

in

both directions

whenthe

Unsharp

Mask

is

a

digital

function like

onthe

CEPS(6).

Proper

adjustment ofthese

four

characteristics

is

the tool scanner operators

have

attheir

disposal

toproduce agood color separation.

It is

necessary

thatoptimumtonereproduction

and

gray balance be

achieved

before

the

color correction can

be

evaluated(5).

The

unsharp

masking is independent

ofthe other

three

characteristics.

Finally,

one can

say

that tone reproduction and

gray

balance

are related

but

different, being

thatcorrecttonereproduction

is

themost

important

ofall

these

four(3).

Therefore,

since all color attributes areaffected

by

the

tone reproduction,

over

90%

of good separations can

be

achieved

by

making

correct

tone reproduction

adjustments,

taking

into

account the original _attributes,

printing

conditionsand customer's

desires(6).

When

onetries toproduce a color reproduction withso

many

restrictionssuch as

different

tonal

ranges, only

three

inks,

dots

torepresent continuous

tones,

impure

inks,

and so

on,

one

has

to

keep

in

mindthat thecolor reproductionprocess

is

a compromise.

While

one

ENDNOTES FOR CHAPTER II

1.-

Field

Gary,

Color

and

its

Reproduction.

Graphic Arts

Technical

Foundation,

Pittsburgh,

1988.

2.-

KODAK

publication

No

0-78.

The Color Separation Scanner.

3.-

Southworth

Miles,

Southworth

Donna,

Quality

and

Productivity in

the

Graphic

Arts,

Graphic

Arts

Publishing

Co., Livonia,

New

York,

1989.

4.-

Color Separation Class

Notes,

Summer

1988,

Prof. Joseph Noga.

5.-

Southworth

Miles,

Color Separation

Techniques,

Third

edition,

Graphic

Arts

Publishing Co., Livonia,

New

York,

1988.

6.-

Miles Southworth

and

Enrique

Arce

meeting

at the

School

of

Printing

at

RIT,

CHAPTER

III

LITERATURE

REVIEW

In

the

introduction it

was seenthat

getting

consistent color reproduction

is

highly

dependent

on thescanner set-up.

Southworth

said that one can expect as

many

opinions about the

scanner

set-up

asthereare scanner operators(l).

Probably

themostcritical

step in

the

setting

up

of a scanner

is

toevaluatetheoriginal

for

optimum results

in

theseparation

for

a given

printing

condition.

Even

an experienced operator

may

make

wrong

decisions,

and the adjustments

in

a scanner

may

notalwaysproducetheexpected results(2).

It is

obvious that

some

device

is

needed to assist the scanner operator to make these

decisions

more scientifically.

A

pre-scan analysis system seemsto

be just

such a

device.

But,

what

is

a pre-scan analysis system?.

A

pre-scan analysis system

is

a generic term

being

usedto

describe

any

systemthat

has been introduced

to

help

scanner operators make

quicker and more accurate scanner set-ups(3).

The

pre-scan system

basically

helps

the

operatortoevaluatetheoriginal

copy in

termsof

emulsion,

detail,

under or over

exposure,

tonal

range,

andother

factors.

Once

thecharacteristics oftheoriginalare

entered,

the

device

will selectanoptimaltone reproductioncurve

for

that

original,

and agiven setof

printing

conditions.

Then

thesystemprintsoutthe

directions

for

setting

the

scanner's

controls,

or

in

some cases

it

programs the scanner

itself(4).

The

system generates

the

optimum tone

reproduction

data

after

considering

the characteristics ofthe

original,

the paper to

be

printed

on,

andthepress variables.

From

the optimumtone reproduction

data

a computer

cangenerate

density

or

dot

percentvalues

for

each oftheprocess colors.

Then

the

desired

scannercontrol settings

for

thegradationcan

be

set(2).

Basically

therearetwo types of pre-scan systems: off-line scanner

presetters,

and on-line

The

presetter/previeweras

its

name

implys is

a

presetting device

with a video

display

to

showthe effects of

the

settings and corrections

before

producing

the

film

and the

hard

proof,

all

done

using

a quick

low

resolution scan.

Even

in

thiscase one needs tomake a trial scanner

set-up,

and

according

to the

_{results shown on a video}

display,

_the scanner

operator

decides

the next

step in

the color separation process.

Most

ofthese systems are

dedicated

to

proprietary

scanners,

and

they

reduce makeovers and proofs

by

about

two-

thirds.

Scans

_{made with presetter/previewer average}

about

10%

makeovers and

between

1.25

to

1.5

proofs per scan(5).

There is

also athird

type

which

is

avideo

display

usedon thescannerafter a

preliminary

scan

is

madeto

determine

the

accuracy

ofthe

set-up before

producing

the

film

and proof(3).

In

thiscase one uses anormal

scanning

speed.

One only

savesthe

film

andproof

material,

because

thescanner operator

has

to

decide based

on

his

own experiencethe changes

in

the

scanner

set-up

if

needed,

if

not

he

goesto the

film

andthe

hard

proofwith the

first

scan

made.

Presetters

are the systems used off-line that

help

to

determine

a scanner

set-up

by

analyzing

the original characteristics.

Following

are a

brief description

ofthe most

important

features for

seven ofthesesystems:

-Carlson

ScanCal

Prescan System:

the operator makes ascanofthe original

copy

with a

black

and white video camera(6).

With

the

density

values obtained the system makes a

density frequency

histogram. It

calculatesthe

tonal

contentof

the original,

then

determines

thescannergradationcurve

according

to the tonereproduction and

kind

oforiginal copy(7).

This

manufacturer guarantees

his

system will

increase productivity

by

at

least

20%(8).

-Chemco/CAPS

(Computer

Aided Prescan System):

theoperator uses a visual comparison

ofthe original

transparency

with aset ofreference

transparencies

to

classify it first into

high,

normal,

or

low

key,

then

into

over, under,

or normal

exposed,

and

then

into

high,

normal,

or

low

contrast.

This information is

then

keyed into

the

CAPS

subprogram^).

Afterwards

thecomputer convertsthis

information

to tonereproductioncurves andscanner

settings(5).

-Colortune

Viewer: is

a small portable computer controlled viewer.

The

scanner

set-up is

colorcorrected

for

the

transparency

color

imbalance

by

adding CC

filters

while

it is

being

viewed(6).

After

reaching the

pointthatproduces a good visual

appearance, those

values are translated

into

the

corresponding

changes

in

dot

sizes

in

the

final

color separations(2).

readings(6).

The

system analyzes

the

result

by

comparing it

with a

database

ofover

20,000

color reproductions analyzed

in

the

Brunner

and

DuPont

color reproduction studies(5).

According

to thecustomer preferences

(clean

and

bright,

facsimile,

lighter,

or

darker),

and

the

characteristics of

the

customer's press and stocktheaimpoint

for

thescanner

set-up

is

produced(2).

-Colorcomp:

This

system uses a spectrophotometer to measure the color of the

transparency

ratherthan a

densitometer.

This

data

is

then

fed into

the computer which

automatically

generates the

data

for

setting

up

the scanner

for

the optimum color

separations(2).

-Spectronics

Scantronics: The densitometer

or the scanner

density

readings

from

a

transparency

are

input into

a computer.

Then it

calculatesthe tonereproductioncurves and

theproper settings

for

thescanner set-up.(5).

-Numakura

G.I.S. Equations:

this system allows the scanner operator to

determine

the

ideal

tonereproduction curves

needed,

using

the

density

readingstaken

from

the original

copy,

and the relation

between

the

density

ranges

for

the original andthe

reproduction,

without operator

judgment

1,11).

Presetter/Previewers

areon-linesystemsthatallowthescanner operatortouseatrial

scanner

set-up,

then theresult ofthecolor separation

is

viewed on a video

display. The

scanner operator

decides

then

if it is

necessary

tochangethescanner

set-up

or acceptthe

result obtained.

Following

area

brief description

ofthemost

important

features for

seven

ofthesesystems.

-Crosfield

Scan View: is

asystem thatuses a

high

resolution

CRT

(1024x1024)

toviewthe

expected

reproduction,

after the original

copy is

scanned at twice the normal

scanning

speed withsettings estimatedto

be

correct

by

thescanner operator.

When

theresults are

viewed onthescreen

the

settings are

left

aloneor changed

before

the

final

high

resolution

scan

is

made(2,5).

-Scanalizer

200

1:

from

Royal Zenith

also uses a

low

resolution

fast

scantoproduce avideo

display

using

the trial

setting

onthescanner.

Then

theresults are

visually

analyzed.

If it is

necessary

the scannersettings are

changed,

and a normal speed scan

is

made

to

film

and

proof.

The

system

has

a

memory

to

keep

the

different

settings and

helps

the

operator make

visualcomparisons and critical evaluations(2,5).

-Scan

Proof: is

made

by

a

Japanese

company

(Toppan).

It

uses a

20

second

fast

scan.

The

reproduced.

Necessary

set-up

changes can

be

displayed

immediately.

It

can also store

images

from

previous settings

for

visual comparisons(2,5).

-Scanvision:

is

a

Swiss company

(Gramag

Ag)

product and

it

works

using

the same

principlesas the

Scan

Proof

system(2).

-Skanskop:

is

part of the

Hell

Chromacom

prepress system.

It

can

only

make

high

resolution scans.

The

scanned

image is

shown on a color monitor

display(2,5).

-Royal

Zenit 210-CM:

is

an

interactive

system thatcan produce a

fast

low

resolution scan

(in

40

seconds).

The

image is

viewed onamonitor.

The desired

changes are made andthen

thescan

is

produced at normal speed(2,5).

-Scitex

Pre-sponse Console:

a standard scanner

set-up

can

be

used to make a

high

resolution scan.

The

result

is

thenviewed on a color video monitoreither as a

full

coloror

as

individual

separation

for

each color.

If

the

scan requires morethanslight changes the

transparency

can

be

rescanedwhile

it is

still on the

drum,

and theevaluation process

is

repeated.

It

can also

be

corrected

digitally(2,3,5).

One has

to

keep

in

mind that althoughthe presetter/previewer seems to

be

more

useful,

viewing

a soft proof requires a skilled operator

because

the

viewed

image

has

to

be

transformed

in

one's mind

from

what

is

seentowhat willprint.

The

reason

for

thisvisual

discrepancy

is

the

differences in

hue, luminance,

andresolution

from

thevideoscreen to

the

printed

imaged

0).

Also

all video systems requireanestimatedtrial tone reproduction

setting

toproducethe

first

scan.

There

are some other

devices like

the

Scanner Data

Terminal,

Data

setter

DS-700,

and so

on.thatcan

improve

scanner

set-up,

and scanner

productivity,

but

they

arenota pre-scan

device

as

it

was

defined before(2).

All

thesesystems and others not mentionedallowpeople

to

improve

thescanner

set-up

by

either

producing

a print-out with the tone reproduction

setting

specifications that the

scanner operatorcan

use,

or

by directly

programming

them

into

thescanner.

Some

ofthese

systems are able

to

workwith

any

scanner andothers are name scanner

dependent.

Another

characteristicofthepre-scan systems

is

that

they

can

be

usedasa

teaching

tool.

The

systemtells theoperator

how

tosetthescanner

for

a particular

original,

theoperator

does

so,

looks

atthe

original,

looks

at

the separation,

and

learns

a

little bit

more about the

scanner controls(8).

Using

the pre-scan analysis system one would expect

less

experienced operators could

producemore

high

quality

scansthanwould

normally

be

expected

for

their

skill

level(9).

critical

decisions

abouttone

reproduction, allowing

_{theoperator moretime}

for

specialized

correctionsofspecificcolorsand

final film

evaluation,

_thescanneroperator acceptance of

the

pre-scansystems

is

still a problem(9).

This may be

because

thescanner operators see the

pre-scan equipmentas a replacement

for

their

skill andathreat to their

job.

This

author

believes

the pre-scan system

is

a

help

to the scanner operator to

improve

quality

consistency

and

productivity

ofthecolor separation process.

For

themanagerswho

have

to

make

the

decision

to

buy

this

equipment

there

is

a

lack

of valid

information

proving

the

advantages and

disadvantages

of

using

these

systems.

The

color scanner used with a pre-scan analyzer can work almostcontinuously.

The drums

can

be

preloaded so thescanner stops

spinning only

for

reloading

the

drum

and

film. The

final

cost per scan

is

less

because

theremakes are

fewer,

less film is

used,

and more

high

quality

color separationscan

be

made

in

ashorter amountoftime(10).

The Du Pont Image Manager:

although the objective of pre-scan systems

is

to

improve quality

and

consistency,

the

fact

they

have been

available

in

themarket

for

a

long

time,

and

have

not

been widely

accepted could

lead

one to the question of their

effectiveness.

There

are estimated to

be

200

worldwide,

including

all the

brands

that

exist(12).

This

thesis work was

done

on the

Du Pont Image Manager

system.

The Du Pont Image

Manager

system

is

a presetter

device

thatanalyzestheoriginal

transparency

with

very

little

or no operator assistance.

This

researcher

decided

to use the

Du Pont Image Manager

system

because

as

90%

ofthe efforttoproduce good color separations

is

achieved

by

the

correct tone

reproduction,

and this system appeared

to

the author as

having

the

more

reliable method

for

determining

the tone reproduction adjustment.

This

system

is

ableto

determine automatically

the optimum tone reproduction

for

each original

copy using

a

computerized,

and objectiveevaluation.

The

system

is

programed

for

aspecificclientand

press/ stock combination

by

entering

thegradationcurvesthatproducethe

best

results

for

thecolor separation

in

that case.

Then,

thesystem generates

10

more variations

from

the

gradations

already

entered,

four

for lighter

original and six

for

darker

originals.

In

the

analysis

the

system

determines

the

highlight

density

and

its

location,

andthe

tonal

range

for

the

transparency.

With

this

information

and

the

data base for

the

20,000

samples analyzed

in

the

Brunner

and

Du Pont

colorreproduction

studies,

thesystem generatesthecurve that

predicts

the

best

relationship

between

theoriginal and

its

reproduction.

Finally,

depending

generatesthe

halftone

gradationcurvestoproducethe

desired

reproduction.

This data

are

printed out on a scanner

set-up instruction

sheet.

The

system can also printa

black

and

white,

low

resolution reproduction ofthe

transparency

pointing

outthe

location

for

the

highlight

and shadow.

This is

useful

information

if it is

wantedtoeliminate thecolorcast

ofthe

transparency

andtocheckthe

data

given

by

thesystem.

The Image Manager

analyzes a

transparency

in

the

following

manner(5,13):

-

The

transparency

is

mounted

in

a

holder

andthen

introduced into

the system.

There

the

system

is

adjusted

automatically

for

the size of the

transparency

and a

fast

scan

is

performedto

find

theminimum

density

or

highlight.

-

When

the

highlight is found

the

image is

displayed

on a monochrome monitor with a

flashing

pixel

showing

the

highlight location.

Then,

theoperatorcan adjust the

highlight

level

while

he is viewing

theresults onthemonitor.

-

A

second scan with

the

highlight

zeroed-out

is

performed,

andthesystem

determines

the tonalrange

for

the transparency.

This data is

sentto thecomputer.

-

The

_{computer analyzes} the

data,

and

according

to the customer's preferences such as

Clean

and

Bright, Facsimile, Lighter,

or

Darker,

and the

data base it has

available the

computer

determines

the optimum scannersettings

using

the

basic

gradation

previously

enteredto thesystem whenthe

customer,

press/stockcombination was

determined.

-

The

_results

for

the _{scannersettings areprinted as}

five

_{pointgradations}

along

with other relevant

information

for

the

job.

ENDNOTES

FOR CHAPTER

HI

1

.-

Southworth

Miles,

Southworth

Donna,

Quality

and

Productivity in

the

Graphic

Arts,

Graphic

Arts

Publishing Co., Livonia,

New

York,

1989.

2.-

Molla

Rafiqul,

Electronic Color

Separation.

R. K.

Printing

&

Publishing Co.,

West

Virginia,

1988.

3.-

Southworth

Miles, Quality

Control

Scanner,

Aids

_{toscannerefficiency, quality}and

profit.

Volume

4,

Number 1.

4.-

Graphic Arts

Monthly,

October

1983,

From Camera

_to

Scanner

and

Beyond,

p.

45-52,

Peter Johnston.

5.-

Bruno

Mike,

Status

_{ofPrinting in the}

U.S.A..

1987.

6.-

Southworth

Miles, Quality

Control

Scanner,

Print

85,

Volume

5,

Number

5.

7.-

Laser in

Graphics,

Vol

1,

1984,

p.

612-617,

The Application

of

Digital Picture

Processing

to

Prescan

Evaluation.

Archer Brent.

8.-

Editor &

Publisher,

July

13, 1985,

Prescan Equipment Growing in Popularity,

Fitzgerald

Mark.

9.-

Graphic Arts

Monthly,

October

1984,

Prescan Evaluation Systems Increase Scanner

Productivity,

p.

47-52.

10.-

Southworth

Miles,

Color Separation

Techniques,

Third

edition,

Graphic Arts

Publishing Co., Livonia,

New

York,

1988.

11.-

Takashi Numakura

and

Enrique Arce

meeting

at the

School

of

Printing

at

RIT,

October

26,

1989.

12.-

Miles Southworth

and

Enrique

Arce

meeting

at

the

School

of

Printing

at

RIT,

December

12,

1989.

CHAPTER

IV

STATEMENT

OF THE

PROBLEM

From

the evidence presented

in

theprevious chaptersone can

say

thatone ofthe

most,

if

not the

most,

serious problems

in

the color separation process

is

the scanner

set-up

inconsistency. This

problem causes

many

other problems such as

rescans,

film

waste,

delays

in

the

delivery,

proofing

material

waste,

and so on.

As it

was seen

in

the

literature

review

there

are

many

pre-scan

devices

that can

help

the

scanner operator with the

set-up

task.

Some

systems are

presetters;

same are

presetter/previewers;

while others serve

only

as a previewer.

But

all ofthem

have

one

final

purpose,

which

is

to reduce rescans

by

producing

a good

quality

colorseparation at the

first

scan.

If

one

looks

at

the

color separation market requirements

it

can

be found

that the more

important

points

for

thepeoplethatare aware ofthesuccess oftheir

business

are:

The

overall visual acceptance

by

theircustomerto

the

proofs produced with the

films

obtainedastheresultof aconsistentcolor separation process.

To

get a

high

production

level,

which

is

directly

related

to the

ability

to

produce more

color separations

in

the same period oftime at a

low

rescan rate.

It

means toreduce not

only

the

labor

costs

but

alsothe

film

and

proofing

material costs.

These

pointsmentionedabove show

that the

Du Pont

Image

Manager

can

be

a

very

good

alternative to

fulfill

the color separation requirements.

Thinking

in

this

way

the

author

developed

the

following

research question

for

this thesis:

Does

the use of the

Du Pont Image Manager improve

the

quality

consistency,

and

In

ordertoanswerthisquestionthe

following

hypotheses

weretested:

HYPOTHESIS 1

"Color

separations produced onthe

Hell DC 399 Scanner

using input set-up

from

the

Du Pont Image

Manager

will

be

more

visually

acceptedthanwhenthecolor separations are

produced

using

scanneroperators'

judgment

for

thescanner set-up. "

HYPOTHESIS

2

"The

color separation process on

the

Hell DC 399 Scanner

will

be

more productive

(considered

asthenumber of color separationsproduced

in

a periodof

time)

when the

scanner

set-up input is

determined

using

the

Du Pont Image Manager

than when

it

is

determined

using

scanner

operators'

CHAPTER

V

METHODOLOGY

1.-

Experimental

Design

In

orderto test

the

hypotheses

formulated,

it

was

necessary

tocomparethe overall visual

acceptance of color proofs made withthe

films from

thecolor separationsproduced

using

input

from

the

Du Pont Image Manager

againsttheunassistedtraditionalmethod of scanner

set-up.

To

produce the

scanning

operation,

three experienced andthree

inexperienced

scanner

operators were askedto

help

in

the

development

ofthisproject.

For

defining

these two

categories an experienced scanneroperator was considered as one who

has

produced at

least 1

,000color separations.

Five

transparencieswere used as representative ofthe

kind

of originals a color separation

shop

couldreceive.

They

were

high, low,

andmedium

key,

and

for

thesame subjectas the

medium

key

an

over,

and anunder exposed

transparencies

were produced.

This

researcher

decided

tousethe

Hell DC 399 ER

scanner

because

it

wasthought

it

was

going

to

be

connected

directly by

an

interface

withthe

Du Pont Image Manager.

Finally

the

interface

was not

available,

but

as the gradations and all the previous process were

developed

for

thatscanner

it

was

decided

to

continuethis research

there,

but any

scanner

could

be

used.

As

theobjective

for

thisthesiswasto comparetheresults

from

the

Image Manager

to the

traditional method of scanner

set-up,

only

changes

in

tone reproduction were produced.

The

scanner operatorswere also

instructed

tonot removethecolor

cast,

to

keep

the color

correction settings at

the

normal conditions

for

the

scanner,

and toproduce aclean and

bright

colorseparation.

A

copy

of

the

instructions

for

the

scanner operators

is

presented

in

Appendix

A.

The

tonereproduction characteristics

for

theprocess were

defined

by

theuse

of

the

Cromalin

color

proofing system,

which was

previously

calibrated to

the

SWOP

standards.

The

proofing

calibration

data

arepresented

in

the

Appendix C. The

Cromalin

methods used

for

thescannerset-up.

The five

originals were analyzed

first

on the

Du Pont Image Manager

to

determine

the

system's recommended scanner set-up.

The

gradation curves enteredto the system were

determined

by

trial and error

for different

gradations.

Finally,

the set of gradations

recommended

in

the

Image

Manager Owner's Manual

for Cromalin

proofing

on coated

stock were used.

The

gradation curves andtheprint out

from

the

Du Pont Image Manager

for

each

transparency

are presented

in

theappendix

D.

Each

original was analyzed

through this

system six

times.

An

average value

for

the time

required

for

each analysis was

then

evaluated.

The

system was operated

in

theautomatic

mode,

so no operator

judgment

was required.

In

the color separation process each scanner operator was asked to produce a clean and

bright

reproduction

for

each original.

After

the scanner operator

finished

with the

separation made

based

on

his judgment he

was askedtoproduce anewcolorseparation

for

the same original

but

now

following

the

Du Pont Image Manager instructions. All

the

scanner operator's settings were

recorded,

then

if it

were required to repeat a color

separation,

only

theoriginal settings should

be

used.

This

was

done because only

the

first

scans were considered

in

thisstudy.

No

rescans wereallowed.

The

time required

for

all operations

using

both

methods was recorded.

This

time period

included

from

when the transparencieswere

handed

to the operatorstowhenthescanner

was

ready

to start scanning.

For

the analysis

using

the

Du Pont Image Manager

the

complete process was timed.

When

theabove procedures were completed

60

sets of

four

color separation

films

were available.

The

color separations were then color proofed.

To

minimize

the

variability in

the color

proofing

system

(although

it

was

calibrated)

each pairof color separationsthatwere

going

to

be

compared were proofed simultaneously.

Thus if

for

any

reason a minorvariation

happen in

one color proof

it

also

happened

to the another.

All

the color proofs were

numbered

in

a random

way

to

eliminate names and

any bias in

the

distribution for

the

evaluationthatcould

influence

the

results.

The

60

color proofs were

then

evaluated

by

12

judges.

In

order

to

analyze

the two

hypotheses

stated

it

was

necessary

to perform a subjective

evaluation ofthe

different

colorseparations producedto

determine

which method produced

the

highest

number of accepted scans.

To determine

which method

had

the

highest

productivity, the

time

it

took theoperators

to

set-up

the

scanner

using

both

set-up

methods

2.-

Subjective Evaluation

A

colorproof

using

the

Cromalin

system was made

for

all the scanner

operators' color

separations,

which was proofed side-by-side withthe respective

Image Manager's

color

separation produced

by

thesame scanner operator

for

thesame originaltransparency.

So

30

pairs of color proofs were obtained.

They

were

judged

by

a panel of

1

2

judges

selected

from

thestudents

taking

the

Electronic

Color Image

and

Control

courseofferedthe

Spring

quarter,

1990 in

the

School

of

Printing

at

RIT. This

class was selected

because it

involves

students

interested in

the

color pre-press

area,

and

because

they

have

asolid

background in

the

area of

tone

and color reproduction.

All

the

color proofs were numbered

from

one

to

60,

andtherewasno

identification

ofwho

made which color separation.

The

order

for

thepresentation ofthecolorproofs was chosen

randomly.

All

thenon-image areas were cuttoavoid

any

visual

interference,

andtheproofs

weremounted on a neutral

gray

cardboard.

Before

the

evaluation,

detailed

instructions

were

handed

to the

judges

along

with the

answer sheetsand a

copy

ofthe scanner

operators'

instructions.

The

judges

were asked toperform an evaluation ofthe colorproofs

in

two

different

ways.

First,

showing

the

judges

one color proof atthe

time,

they

wereasked:

Based only

ontone

reproduction

criteria,

is

thecolor separation acceptable?

Second,

showing

the

judges

each

pair of color

proofs,

they

were asked:

Based

only

ontonereproduction

criteria,

whichone

of

the

twocolorseparations

do

youprefer?

Once

the paired comparison was performed

by

all the

12

judges,

an assessment of

inter-judge

agreement was made

using

the

resultsobtained

from

thesubjective evaluation.

The

Scott's

coefficient was evaluated

for

each

judged).

To have

a

judge

considered

consistentwiththeothers a

Scott's

coefficientof

76%

was required.

With

the results

for

the evaluationsthe

12 judges

were

ranked,

andsix out ofthe

12

were chosen.

Those

six

judges

had

a

Scott's

coefficient of at

least

67%,

which was

less

than preferred

but

still

considered acceptable

for

thepurposesofthestudy.

The

resultsofthe

Scotf

s coefficient

3.

-Equipmentor

Resources

The

list

of

materials, equipment,

_{and resources used}

in

this

study is

as

follows:

-

Hell Chromagraph

399 ER

electronic color scanner

-

Dupont

Image

Manager

_{prescan system with software version}

4.0

-

Kodalith film

processor

-GTI,

Graphic

Technology

Inc.,

Viewing

Booth

with

illuminant

at

5000 K

-3M,

Electronic

Dot

Generated Scanner

Film,

52.5

x

64.6

cm,

or

20.6

x

25.4 in.

-

Du

Pont Cromalin Laminator

(positive)

-

Du Pont Cromalin

processor unit

ATM II

(positive)

-

Frame U.V.

light

Teaneck Graphic

Corp.,

L 1250

lamp,

Kokomo

Filter,

light integrator

Expo

19,

GTI Graphic

Technology

Inc.

-

Cromalin Commercial Receptor CR/1

-

Du Pont

toner

powderyellow

SOP/Y

pos-6,

magenta

SOP/M

pos-4,

cyan

SOP/C

pos-4,

and

black

SOP/K

pos-4.

-

Positive

working proofing

film

C4/CP,

25 in

x

300

ft,

batch

roll

1630-0010.

-Transparency

2

1/4

x

2

1/4

in,

named:

ASPIRIN

(HIGH

KEY).

-Transparency

35

mm,

named:

SUNSET

(LOW

KEY).

-Transparency

35

mm,

named:

VOLCAN OSORNO

(NORMAL

KEY).

-Transparency

35

mm,

named:

VOLCAN OSORNO

(UNDER

EXPOSED).

LIMITATIONS

OF

THE

STUDY

Although

the experiment performed answered the research _{question, there} are some

limitations

thatrestrictedtheresults obtained.

These

arethenatural

limitations in

any study,

such as

limitations

in

time,

resources,

and

money

available.

Even

though the

study

answeredthe

hypotheses

with

the

restrictionsthere

are,

it is

still

interesting

to

know

what

limitations

thisauthor saw.

The

experiment was performed

by

only

sixscanner

operators,

threeexperienced andthree

inexperienced.

It

would

be

better

to

have

a

larger

number ofscanner operators to

better

account

for

theoperators'

variability.

Also

allthe scanner operators

had

thesame

basis in

their

education,

it is

because

all of

them

are

from

the

School

of

Printing

atthe

RIT.

Five

different

originals were used

assuming

that thosearerepresentative ofthemarket.

It

might

be

better

to

analyzereal

jobs

and compare the results of

using

or not

using

the

Du Pont

Image

Manager.

The

scanner chosen was

the

Hell DC

399,

so the results are

directly

applicable

for

that

scanner,

andthere

is

not

any

evidence about what should

happen in

a

different

scanner.

The

pre-scan system studied

only

involved

tone reproduction.

The

color correction and

unsharp masking

functions

were

kept

constant.

But it is known

thata change

in

the tone

reproductionaffectsthecolor_correction,so

it

could

be

interesting

tocheck

if it is

easier

for

the

scanneroperatortoproducethecolor correction once

he is

sure

the

tonereproduction

is

correct.

The

comparisonwas

done

by

judges

thatwere

only

analyzed

for

inter-judge

agreement,

and not

for

visualcolor

deficiency

orcolor preferences.

The

number of

judges

usedwas

limited

due

to the

availability

ofqualified

judges.

The judges

were studentsthat

had

asolid

theoretical

background but may

not

have had

the practical experienced that customers

ENDNOTES FOR THE CHAPTER V

1.-

Public Opinion

Quarterly, Vol.19,

1955.

p.

321-325. Reliability

ofcontent analysis:

CHAPTER

VI

THE RESULTS

The

subjective evaluation

by

the

12 judges

was performed under standard

viewing

conditions.

The data

collected

for

the rejection

rate,

which

is

ameasurement ofthecolor

separation

acceptance,

and the color separation preference

based

on tone reproduction

criteria are presented

in

appendix

F,

tables

F-l

and table

F-2

respectively.

The

time

recorded

for

all the operations performed

by

the sixscanner operators are presented

in

appendix

F,

table

F-3.

With

the

data from

the table

F-2

the inter-

judge

_agreement

study

was performed as

it is

shown

in

table

F-4. To

compare allthe

judges

toeach

other,

66

evaluations

for

the

Scott's

coefficient were