GATF

Test Form / Analysis

User Guide
 Version 5.0
This is a multi-version User’s Guide that covers the use and diagnostics of the
common targets found on GATF Test Forms. Not all test forms contain all of
the targets that are described in the manual. Simply locate the target which
you want to use on the test form and then locate the use and explanation page
of the target by locating the target in the index of this guide.

The 25 x 38 Sheetfed Test Form is used to illustrate most of the locations of the
targets in this manual. It is for reference only and not to be confused with the
actual form that you are using.
 GATF
Test Form / Analysis
User Guide

by
Gregory A. Bassinger

Graphic Arts Technical Foundation

PITTSBURGH
 Copyright 1997
Graphic Arts Technical Foundation
All Rights Reserved

Printed in the United States of America

Reproduction in any form by any means without specific written permission is prohibited.

Agreement for Use

All GATF Test Forms are sold for use in a single site by the purchaser. It is unlawful
to distribute this product to other parties without the express written consent of
GATF.

Graphic Arts Technical Foundation

200 Deer Run Road
Sewickley, PA 15147-2600
Phone: 412/741-6860
Fax: 412/741-2311
Email: [email protected]
Internet: http://www.gatf.org
 Table of Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-1
Purposes of the Test Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-3
Planning a Press Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-5
Materials Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-5
Procedure for Press Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-7
Print Analysis Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-10

Targets for Evaluation Section 2

Information Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-2
Line Resolution Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-3
GATF/Systems of Merritt Digital Plate Control Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-4
Digital Ladder Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-6
Star Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-8
Image Fit Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-9
Mottle Patches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-10
Gray Balance Chart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-11
Three-Color Gray Bars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-12
GCA/GATF Digital Proof Comparator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-13
Transfer Grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-17
Color Correction Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-18
Ink Coverage Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-19
Twenty-Step Tone Scales and Vignettes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-20
Dot Size Comparator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-21
Single-Tiered Control Bar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-22
Two-Tiered Control Bar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-23
IT8.7/3 Basic Data Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-24
GATF/RHEM Light Indicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-25

Testing Color Reproduction Section 3

Low Key/Red Couch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-3
High/Key/Wedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-4
Group Portrait/Painting Kids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-5
Female Portrait . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-6
Gray Neutrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-7
Color Gamut/Fruit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-8
Memory Colors/Covered Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-9

Analysis of Print Attributes Section 4

Color Measuring Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-2
Solid Ink Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-4
Dot Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-8
Print Contrast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-11
Ink Trapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-12
Concentric Hexagon Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-13
Color Reproduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-15
 GATF Test Form User Guide 1-1

Introduction
The GATF Test Forms described in this manual are available as film positive or negative as well as a digitally.
They provide a format that is used to diagnose and calibrate the corresponding printing press with up to eight units
depending on the test form that has been selected. The digital versions of the press test forms match the film versions
wherever practical. This is intended to maintain consistency between the press analysis of the digital and film press forms.
There will be differences between the digital and film test forms that result from the nature of digital imaging and
the respective resolutions of the digital and analog modes. Therefore, the GATF Test Forms all share the heritage of the
original GATF Test Forms, including the use of several native PostScript elements to exploit the full potential of the film-
setter or platesetter.
The GATF Digital Test Forms are supplied on CD-ROM written in ISO9660 format, PC and Mac readable. The film
based Test Forms are supplied on .007” film to ensure stability and consistency. The digital files consist of an encapsulated
PostScript (EPS) file with nested TIFF images and native PostScript elements. The test kit also includes a Press Test
Analysis Section to help organize and record the results of a press test.
There are several other GATF products that will enhance the functionality of the GATF Test Forms. Among these
are the ghosting form, the register test grid, and the plate control target, all of which are available as film based targets or
digitally.
To use the digital version of the Test Form, the CD-ROM can be loaded into a CD drive and imaged through most
output system, such as a digital proofer, imagesetter, platesetter, or digital press. There are targets incorporated in the
Test Form that measure the exposure accuracy of an imaging device. It is important to determine that the digital output
device is accurately calibrated and linearized for a quality press test form to be output. The film-based GATF Test Forms
can be used as the standard for comparison with digital output.
1-2 GATF Test Form User Guide

GATF Test Forms consists of a single press test form that provide a wide array of press testing and calibration capa-
bilities. Density, dot gain, print contrast, and ink trapping (the most-common process control elements) are thoroughly
checked. Print problems such as slur, doubling, paper fan-out, and ink/water imbalance can also be detected. The color
reproduction characteristics of the printing system can be calculated once you assure that the press is in good operating con-
dition. When the color separation process is calibrated to the printing conditions of the press with any GATF Test Form,
one has assurance that the printing represents optimum press operating conditions. Conscientious use of a Test Form can
lead to the highest-quality color reproduction attainable from the printing press.
Several factors have contributed to the development of the GATF Test Forms, including increased demand for high-
quality printing, rising costs of operating presses, and cautious buyers who insist on press tests before they purchase new
or used presses. The development of the digital version of these test forms was undertaken in recognition of the increased
use of computer-to-plate and computer-to-press systems.
Producing consistent high-quality printing can be an elusive goal. It requires a combination of good management,
effective communications, properly maintained equipment, craftsmanship, control of raw materials, measurement, and
feedback. High quality does not result from the printing press alone; the press must be viewed as a part of an overall
reproduction system. The diagram on page 1-4 shows the elements of such a system.
These elements should not be considered as individual units, but rather as parts of an integrated whole. The heart of
the system is the printing press. There are optimum results that can be achieved with a press. These results are influenced
by the materials used on the press (e.g., inks, paper, fountain solution, and printing blankets) and by the mechanical condi-
tion of the press. Worn parts and improper adjustments can cause a host of printing problems (e.g., slurring, uneven ink-
ing, misregister, or uneven water distribution). For consistent high-quality color reproduction, a press must be properly
maintained and adjusted.
It is beyond the scope of this user’s guide to present detailed information on the establishment or execution of press
maintenance procedures. Some general guidelines are presented here, but the press manufacturer’s specifications and
recommended procedures should take precedence in setting up a maintenance schedule. It is impossible to overempha-
size the importance of good maintenance practices in achieving high-quality color printing. A properly running press pro-
vides the predictable results on which the entire color reproduction system depends.
Color separation films and, occasionally, original photographs are made to fit the particular reproduction characteris-
tics of a printing system. If the system is stable, uniformly high-quality printing can be expected.
Consistency, teamwork, and effective communications are traits that lead to success. Measurement and record keep-
ing are essential to the process of quality control. Every link in the reproduction chain must be operated within specifica-
tions. Statistical process control can aid in detecting points of variation. Review of printed results and feedback are essen-
tial to keep all systems running at their peak efficiency. Companies producing high-quality printing are characterized by
progressive management. The top executives must create an unthreatening environment conducive to communication.
Likewise, the commitment to quality must start at the top of a company’s hierarchy in order to permeate into all depart-
ments in the printing plant. As new equipment and materials are integrated into the printing system, the GATF Test
Forms provide a valuable tool for maintaining quality and comparing the new printing system with the old one. The con-
scientious use of this test kit allows the printer to establish meaningful plant standards and to uncover sources of error in
the reproduction chain.
 GATF Test Form User Guide 1-3

Purposes of the Test Form

The GATF Test Forms serve three main purposes: diagnostic, calibration, and process control. The diagnostic capabilities
of the test forms have been greatly enhanced in the 5.0 versions by the addition of several native PostScript elements.
These elements are able to, in effect, have a two-way dialogue with a raster image processor (RIP) and address any output
device in terms of its own resolution.

The diagnostic capabilities of the GATF Test Forms include:

• Resolution. The resolution of any output device is measured in terms of pixels and microns in both negative and
positive modes.
• Directional effects. The resolution is measured horizontally and vertically, giving clear indication of directional dif-
ferences in imaging. The star targets are also excellent indicators of directional disturbances.
• Banding. An evaluation of banding is made with full-scale vignettes that are included for all six colors.
• Slur or doubling. Press problems, like slurring or doubling, can be identified with the digital ladder targets incor-
porated into the test form.
• Register. Poor or inconsistent register between any of the six printing units will be displayed in the reproduction of
the transfer grids.

The calibration capabilities of the GATF Test Forms may include the following:
• Contrast and tone reproduction. Twenty-step tone scales are used to measure linearization and tone reproduction
of the printing system. The photographs provide additional visual evidence of the tone and contrast characteristics
of the printing systems.
• Highlight/shadow points. The smallest and largest reproducible dots for each color are measured by the
highlight/shadow scales on page one of the test form.
• Gray balance. The gray balance characteristics of the printing system are measured with the gray balance chart.
The gray balance specified by SWOP for magazine production is incorporated into the three-color gray bars of the
GCA/GATF Digital Proof Comparator 2.0 on the test form.
• Color management. The calibration of color gamut mapping between various input devices, color monitors, and
output devices is facilitated by the ANSI IT 8.7/3 Basic Data Set that is provided on the test form.

The process control functions that are performed with the GATF Digital Test Forms may include the following:
• Electronic processing data. Native PostScript elements in several locations on the test form ascertain and display
electronic processing data like the name of the RIP, its PostScript level, horizontal and vertical resolutions, direc-
tion of imaging, screen ruling, dot shape, and screen angle.
• Exposure. The plate or film exposure are monitored by positive and negative microline elements on the
GATF/Systems of Merritt Plate Control Target included on the test form. This target is available as a stand-alone
process control device from GATF. It can be used to measure several important attributes during the imaging of
computer-to-plate (CTP) printing plates or photographic films.
• Color proofing calibration. The test form includes the GCA/GATF Digital Proof Comparator 2.0. This multi-ele-
ment quality-control device is available as a stand-alone product from GATF. After the test form is used to diag-
nose and calibrate the press system, the Digital Proof Comparator is used to monitor the exposure and consistency
of subsequent proofs.
1-4 GATF Test Form User Guide

Planning a Press Test

Specific objectives should be set when planning a test for
Planning Stage
any press. Is it to be a test of printing and imaging equip-
ment only, or will it be a test of color reproduction char- Set Objectives for Testing
acteristics as well? When specific objectives are clearly
expressed and key personnel are alerted, chances for a Identify Equipment for Test
successful press test are maximized.
The materials to be used for the test (paper, ink, Identify Materials for Test
and fountain solution) should be selected and pretested.
If the properties of the materials for printing show that a Identify Personnel to Carry out Testing
problem is likely, that problem must be resolved before
the press test is started. Press maintenance and adjust- Schedule Events and Notify
ments are current should be up to date before the press Key Personnel
test is conducted. GATF technical consultants can be
brought in to perform a technical assessment in conjunc- Pretest Printing Materials
tion with a press test.
The GATF Press Test Analysis Form should be Perform Equipment Maintenance
photocopied and partially completed in advance of the
press test. Record pertinent press test information and GATF Technical Plant Audit
press measurements in the analysis form. Attach any
technical lab reports generated from materials testing to
the analysis form. Testing Stage
The flow diagram to the right shows a possible
sequence of events surrounding the running of a GATF Make Color Separations
Test Form. Good communication is crucial to the success (If Transparencies are Purchased)
of any testing program. During the planning stage, objec-
tives and procedures for testing should be clearly com-
municated to all key personnel. After the test, review Assemble Films
results and recommendations with all departments that
participate in the test effort. Make Printing Plates

Perform the Press Test

Results & Conclusions

Analyze the Printed Results

Correct Problems that Are Discovered

Enter the Results into a Data Base

Repeat the Process at

Scheduled Intervals
 GATF Test Form User Guide 1-5

Materials Testing
It is a good idea to pretest the materials that will be used for a press test. Laboratory testing can identify properties of ink,
paper, and fountain solutions that could lead to problems during a pressrun. This can provide a significant savings of
money, time, and effort when compared to discovery of the problem during a press test. It is not uncommon for the
results of a press test to be invalidated due to incompatible printing materials. Samples of the materials you intend to use
for your pressrun can be sent to GATF’s Materials Testing Lab for analysis. For instance, ink should be tested for tack,
viscosity, and relative strength.
The pH and conductivity of the fountain solution can be measured in addition to the extent to which it emulsifies in
the printing ink. Ink/water take-up curves show the amount of fountain solution emulsified in the ink over time. These
curves provide information indicating the likelihood that a given ink and water combination will present problems on
press.
The paper for the press test can be pretested for absorptivity, surface strength, printability, basis weight, brightness,
opacity, and smoothness. Below is a partial list of the standard tests offered by GATF. The Materials Testing Lab at GATF
can be contacted directly for more information on testing procedures.

Paper Tests Ink Tests

1. Basis Weight 1. Tack
2. IGT Surface Strength 2. Stability (Break and Slope)
3. Elmendorf Tear Strength 3. Misting
4. Dennison Wax 4. Viscosity (Brookfield and Laray)
5. K&N Absorptivity 5. Tinctorial Strength
6. Croda Red Absorptivity 6. Fineness of Grind
7. Gloss 7. Surland Emulsification
8. Smoothness 8. Bleed in Fountain Solution
9. Porosity 9. pH Change in Fountain Solution
10. Tensile Breaking Strength 10. Percent Nonvolatile Material
11. Elongation to Break 11. Wash-out through 325-Mesh Screen
12. Mullen Bursting Strength 12. Color Comparison
13. MIT Folding Endurance 13. Surface Tension
14. Paper Surface Efficiency 14. Specific Gravity
15. Surface pH
16. Paper pH Print Tests
17. Caliper 1. Gavarti Compression/Abrasion Test
18. Brightness 2. Drying Time
19. Opacity 3. Sutherland Rub Resistance
20. Groundwood Content 4. Print Gloss
21. Moisture Content 5. Color Diagrams
22. Ash Content 6. Photomicrographs
23. Water Absorbency 7. Evaluation of Quality Control Test Images
24. Tri-Pick Coating Test 8. Evaluation of Debris (Tape Pulls)
25. Carlson Curl Test 9. Spectrophotometric Curves
26. Stiffness 10. Efficiency, Hue, and Grayness
27. Fluorescence 11. Blockling
28. Spectral Reflectance (Color) 12. Lightfastness
29. Oil Absorption
Fountain Solution Tests Blanket and Roller Tests
1. pH 1. Swelling
2. Surface Tension 2. Smoothness
3. Emulsification in Ink 3. Durometer
4. Conductivity 4. Resistency
5. Specific Gravity
1-6 GATF Test Form User Guide

Procedure for Press Testing

Any GATF Test Form can be used to test a printing press without involving the color scanner. The press testing procedure is
described in this section. A variety of different color tests can also be performed with the test form. These are described in
section 3 of this user guide— “Testing Color Reproduction.”
When the objective is solely to test a printing press, follow this procedure:
1. Complete the Press Test Information section of the Analysis Form
2. Open the digital file of the test form, fill in the requested information, and image the file to film or directly to print-
ing plates
3. Assemble the films for platemaking
4. Image the printing plates
5. Mount the plates on the press
6. Complete press makeready
7. Print the test form
8. Analyze the results
9. Communicate the findings

Most GATF Test Forms include a test information Company Test Run #
block that automatically records information to identify the
test run. In addition to the information recorded automatical- Press Sequence
ly, there is important information that should be completed in Date Stock
the analysis form and stored with printed samples from the
press run. The press run information should include the fol- GATF Consultant Inks
lowing items:
• Company. The name of the company and location of the facility where the test was conducted.
• Press. The name, model, serial number and any code that uniquely identifies the press being tested.
• Date. The date of the press test.
• GATF Consultant. The name of the person(s) responsible for supervising and coordinating the press test.
• Test run number. An identification number assigned to the particular press test being conducted.
• Sequence. The printing sequence or ink rotation.
• Stock. The substrate on which the press test will be printed.
• Inks. The ink set that will be used during the press test.

The films output from any GATF Test Form are taped onto clear polyester carrier sheets that fit the plant’s register
system. Care should be taken to properly register the films during this process.
A collimating magnifier is needed to prevent misregister due to parallax error. The addition of the carrier sheet in the
light path during exposure will cause exposure times to increase compared to the standard exposure times for a single thick-
ness of film. The exposure of the plates should be measured with the GATF Plate Control Target (available separately from
GATF).
When attaching the Plate Control Target to the assembled flats, tape the scale with a single piece of tape at each end of
the film strip. Be sure that the target is positioned for emulsion-to-emulsion contact with the printing plate. The same target
is used for all plate exposures, which ensures uniformity in monitoring the exposures. Between successive plate exposures,
move the target from flat to flat. Always handle the target by its edges to avoid fingerprints or scratches.
For a successful press test, it is essential that good quality be obtained in the platemaking process. The maintenance on
the plate exposure unit should be up to date. It should also pass two tests: a register accuracy test and a vacuum drawdown
test. The register accuracy test ensures that register will be maintained between the film flats and the printing plates. The
GATF Register Test Grid can be used to facilitate this test. It is designed to measure the register accuracy of vacuum frames
with a precisely-ruled grid pattern and inset sections of highlight halftone dots. The procedure for testing a vacuum frame for
register accuracy is described in the enclosed user’s guide for the Register Test Grid.
The printing plates should be exposed in the center of the vacuum frame. This provides the most uniform exposure
across the surface of the plates. The plate exposure levels are monitored with the GATF Plate Control Target. After the
plates are exposed and processed, inspect them carefully. Each of the plates is labeled by color in the plate bend area. Check
 GATF Test Form User Guide 1-7

that all plates contain the correct elements, and then carefully analyze the plate for imperfections. Perform the plate analysis
called for in the Press Test Analysis Form and complete that section of the form.
If a plate-reading device is to be used to preset the ink fountain keys, read the plates at this point.
The plates are then bent for mounting on the press. The plate bending device should not be taken for granted. Check
the bends for signs of unevenness. The bends need to be parallel to the plate edge and uniform from plate to plate.
The plates are mounted on the press. The plate packing has to be selected to provide the proper squeeze pressure dur-
ing printing. The manufacturer’s specifications should be consulted to determine both the recommended squeeze pressure
and the proper height of the plate plus packing compared to bearer height.
Continue with the press makeready by bringing the plates into register. There are several register marks to assist in reg-
istering the test form. There are four register marks located on centers along all four edges. In addition, there are four GATF
transfer grids around the central quality control target area.
After the press is in register, the inking levels are adjusted. In daily production, the color proof or color OK sheet is
used as a visual standard for process control during a pressrun. Ideally, the density, dot gain, print contrast, and ink trapping
are measured on the OK sheet. During the press run, these attributes are monitored on sample sheets to aid process control.
Unlike the press control in production printing, the running of a GATF Test Form should follow the instrument read-
ings in preference to visual appraisal of color match. This is because the test form is intended to show how color will appear
under ideal press operating conditions. There is no advantage to altering press conditions to match proofs in this instance.
Instead, the test form is used to calibrate proofing systems to the press. Determination of numerical values in establishing
process control aim points and tolerance limits can be an elusive activity. Industry norms, or specifications such as SNAP,
SWOP or GRACoL, can serve as a useful starting point, but the most meaningful aimpoints will result from careful in-house
testing to find the highest quality level achievable with a given printing system.
Adjust the ink densities across the test form to conform with the aimpoints for density. Alternatively, density aimpoints
can be obtained from the SWOP High/Low Ink References, available from the International Prepress Association (IPA). The
SWOP density levels are intended for sheetfed proofing or web offset production. Often commercial sheetfed printers will
set ink density aimpoints higher than SWOP densities; for example, SWOP high reference value plus 0.10 density units. If
optimum densities for a given printing system are to be measured, refer to the procedure on page 4-4 of this user guide.
When making adjustments to ink keys, allow sufficient time for the press to react before evaluating the effect of that
adjustment. It is best not to make more ink key adjustments than necessary to balance out the ink levels. It is common to
print scrap stock during this process to preserve the good printing stock. Ten sheets of good printing paper are followed by
100 sheets of scrap stock, then ten more sheets of good paper, and so on. The density levels on the good sheets are measured.
Corrections are made to the ink key settings based on these readings. Scrap stock is printed while the press slowly changes to
reflect the effects of the inking adjustments.
The water should be set at the minimum level needed to maintain clean nonprinting areas. To adjust this, the water
feed is lowered until the point is reached where scumming begins to occur. Then, it is increased by one increment. This “just
above scumming” dampening condition avoids printing problems that occur when too much water is used. When excess
water is run, the rheological properties of the inks might be adversely affected. The ink will tend to emulsify more water,
which will lead to lower ink tack and less-efficient ink transfer. Ink piling on the rollers can be the result of excessive water.
Furthermore, excess water can cause defects known as “wash marks.” Wash marks are sections of lighter density that start at
the lead edge of a solid and progress for a short distance in the direction of travel before they fade into the darker solid ink.
The GATF Test Forms contain color bars at the leading and trailing edges, when room permits, to monitor ink density
levels during makeready. The bar of repeating solids on the trailing edge offers maximum control over the density levels of
adjacent inking keys. It can be read by many scanning densitometers, and it is the one most accessible during makeready.
Neither color bar extends above the ladder targets. Instead, the ladder targets are flanked by solid ink patches of the
appropriate colors at the leading and trailing edges. Since there is no ink takeoff for the other colors in these ink zones, it is
not necessary to achieve density aimpoints for any color other than the color of the ladder target in that zone.
Once the press has stabilized in conformance to the ink density and dotain aimpoints, collect samples for further
analysis. The number of samples and the frequency with which they are collected will depend on the experimental pur-
poses of the press test being conducted. For instance, if the purpose of the testing is to troubleshoot the mechanical oper-
ations of a press, 200 consecutive samples taken after the press is balanced out may be sufficient for analysis. If, however,
the purpose of the press test is to evaluate the variability of a particular printing system over time, five consecutive sheets
every 500 impressions over 20,000 total impressions may be a more appropriate sampling plan.
1-8 GATF Test Form User Guide

Print Analysis Procedure

The analysis procedures for a single sample sheet are contained in this guide. Suggested procedures for performing
process variability studies are also included. To analyze a single sample sheet, the user needs a hand magnifier, a reflection
densitometer, and a color viewing booth. The types of analysis performed on the sample can be divided into visual and
instrumental categories. The visual analysis include assessment of slur, doubling, toning, mottle, image fit, vignette repro-
duction and highlight and shadow reproduction. Many print defects can also be identified visually, such as the presence of
hickeys, missing spots in screen tint, broken lines, ink fill-in, wrinkles in the sheet, or wash marks. The instrument-assisted
category of print analysis includes densitometric determination of ink density, dot gain, print contrast, ink trapping, hue
error, and grayness.
All evaluations of color are made under standard lighting conditions as specified in ANSI PH 2.32-1972. Concise
information about this standard is contained in GATF Technical Services Report No. 7233, A Review of Color Viewing
Conditions for the Graphic Arts.
The procedure for analyzing a printed sample varies with the goals of the press testing. There is no rigid sequence
for performing the analysis steps, but there are some conditions that must be satisfied before it is logical to proceed with
the analysis. For example, if density aimpoints are not being met, an analysis of gray balance will only yield misleading
information. A possible sequence for the analysis of a single printed sample follows:
1. Check the sample for wrinkles.
2. Examine the gripper marks at the lead edge of the sheet for evenness and pressure.
3. Scan the overall sheet for signs of toning, fill-in, or tinting.
4. Examine register and fit of the printed images.
5. Scan the sheet for defects such as picking, hickeys, or dissolved coating debris.
6. Measure the solid densities across the test form.
7. Measure the dot gain in several locations.
8. Measure the print contrast.
9. Measure the ink trapping for blue, green, and red.
10. Evaluate the digital ladder targets.
11. Examine the star targets to evaluate slur and doubling.
12. Examine the vignettes for smoothness of gradation.

When sample sheets are taken, they should first be scanned in the viewing booth for gross print defects. Wrinkles,
setoff, background toning, hickeys, and white spots in printed areas are identified, among other defects. An evaluation is
made of the external register of the form to the sheet. Are the margins even? Is the form square to the sheet? An evaluation
of image fit is also made. Are the images in good register? Are any colors protruding outside of the common windows?
After this initial look at the sample, a more thorough evaluation based on the quality control targets is undertaken. In
practice, the use of the densitometer to measure print attributes, and the succession of visual evaluations made of the sam-
ple do not follow an absolute order. In this user’s guide, the targets designed for visual evaluation are presented first, fol-
lowed by a description of the targets designed for instrument measurement.
 Targets
for Evaluation
Descriptions of the quality control targets from the GATF Test Forms that are primarily analyzed by visual
inspection follow. A brief explanation of how each target is evaluated accompanies the physical description of the target.
Where possible, likely causes of common printing problems that the targets might uncover are listed. For more informa-
tion on sheetfed press troubleshooting based on problems found in printing see the Lithographic Press Operator’s
Handbook and Solving Sheetfed Offset Press Problems. Both books are available from GATF.
2-2 GATF Test Form User Guide

Company Press

Date Sequence

Imagesetter Stock

DPI/LPI Inks

Information Block
The information block is where information about an imaging system and printing system is to be recorded. Most
of the GATF digital test forms also include an information block that is filled in automatically during the RIPping stage of
the print function. This block includes information gotten from the imaging device itself and can help diagnose problems
with the imaging device before the press test is done. This information includes the output device name, PostScript ver-
sion, screen ruling, and dot shape. The user also has the ability to enter data about a specific press test, including the press
identification, ink sequence, printing stock, and ink identification.
A quick examination of this information gives the user a good basis for evaluating the images on the rest of the page.
 GATF Test Form User Guide 2-3

0.01 point 0.40 point

0.05 point 0.60 point

0.10 point 0.80 point

0.20 point 1.00 point

Line Resolution Target

The line resolution target and the GATF/Systems of Merritt Digital Plate Control Target are examined on the films or
plates prior to printing the Digital Test Form. The line resolution target is used to evaluate the ability of the imaging sys-
tem to handle positive and reversed (negative) line elements in a variety of orientations. The line dimensions are specified
in points (72 points per inch) in a range from 1 point to 0.01 points. The lines are oriented in four directions: vertical, hori-
zontal, +45°, and –45°. Additionally, there are semicircles in both positive and negative that meet. The horizontal lines
also meet between positive and negative.
The line resolution on the output from a digital system is evaluated for accuracy of rendering between positive and
negative, consistency of line width in different angular orientations, smoothness of rendered curved lines, and finest lines
that can be resolved. Some imaging systems default all smaller dimensions up to their minimum imaging dimension.
When this is the case, the smallest groups of lines (0.01, 0.05, and 0.10, for instance) would be imaged at the same size.
When film output is used, the smallest elements of the line resolution target are good indicators of proper exposure.
If negative line elements are unduly filled in, overexposure of the film is the probable cause.
This evaluation is the same for printing plates imaged directly from the digital files.
2-4 GATF Test Form User Guide

A B C D E F G H

GATF/Systems of Merritt Digital Plate Control Target

The GATF/Systems of Merritt Digital Plate Control Target is imaged for each color along the outside edges of the test
form. This native PostScript target is a multi-element diagnostic and process control device that monitors exposure of digi-
tal output systems such as platesetters, filmsetters, or proofs. The stand-alone version of the target is available from GATF
and can be purchased separately.
The plate control target queries the RIP and reports information in the left-hand section of the target. This includes
the identification of the imaging device, PostScript level of the interpreter, horizontal and vertical resolutions, direction of
imaging, screen ruling, dot shape, and screen angle (section A).
The next section of the digital plate target measures the resolution of the system with pixel-based test patterns. The
first of these are positive and negative patterns of horizontal and vertical microlines (section B). The microline patterns
should be cleanly rendered because the imaging system is addressed in terms of pixels (not in physical dimensions that
interact with the bitmap pattern of the ripped data). The physical dimensions in microns of each line element are calculat-
ed and printed below the target elements. The negative microlines are very sensitive to fill-in if a negative-acting system is
overexposed. Similarly, the smallest positive lines will disappear if a positive-acting system is overexposed. Not all imaging
systems can resolve a positive and negative one-pixel line pattern at high-resolution (small pixel size) settings.
The second pixel-based target group consists of checkerboard images (section C). These patterns are constructed of
regularly sized squares with spaces of the same dimension between each square. The one-pixel checkerboard pattern has
image elements one pixel high by one pixel wide. These are the smallest marks at a given resolution followed by the small-
est obtainable spaces. This element cannot be adequately imaged with all output systems, thus there are 2-, 3-, and 4-pixel
patterns to aid in the evaluation.
The positive and negative microline arcs (section D) challenge the imaging system with curved line elements with
the smallest addressable dimensions. The microline arc patterns present the most difficult challenge to the imaging sys-
tem. If an output device is simultaneously holding good detail in the 1-pixel positive and negative line arcs, then good
exposure conditions are indicated.
The star target (section E) is composed of a series of wedge-shaped elements 5° wide and spaced at 5° intervals. The
wedge elements join at a point in the center of the target. The higher the resolution of an imaging system, the more accu-
rately it will image the center of the star target. Star targets with the clear open centers going down to a fine point indicate
 GATF Test Form User Guide 2-5

A B C D E F G H

high resolution. Conversely, star targets with excessive fill-in in their centers are a sign of low-resolution systems or
improperly exposed media.
If the center of the star target is not round, there are directional differences in the imaging system. Some imaging
systems have higher resolution in the horizontal direction. This would cause a star target center that had an elliptical cen-
ter where the long axis of the ellipse is oriented vertically.
Below the star target is a solid coverage patch used to measure the density associated with 100% coverage. This solid
patch is needed for measurements of apparent dot area that will be made from the tone patches that accompany it. The
first two of these tone patches are the 50/150 reference patches (section E).
Each of these two patches has 50% dot area coverage. One patch is at a screen ruling of 150 lpi, the other is at 200
lpi. These patches are used to assess tone value increases in an imaging system when higher spatial frequency screens are
used. These differences are objectively measured with a reflection densitometer.
The remainder of the digital plate control target consists of two sets of matched tone scales (including highlight and
shadow patches) as seen in section F.
The difference between the two tone scales is that the top scale is sent through the RIP bypassing any compensation
program that is applied to other files. The bottom scale does not bypass compensation settings. Comparison of the two
scales clearly shows the effects of any such compensation programs. If the two scales are identical, no compensation was
being applied at the RIP.
To use the tone scales, first visually select the highlight and shadow limits of the imaging system with the aid of a
hand magnifier. Next use a reflection densitometer to measure the tone steps from 10 to 90% and construct a dot gain
curve.
2-6 GATF Test Form User Guide

Digital Ladder Targets

The GATF digital ladder targets are very sensitive to a number of print problems including slur, doubling, wash marks,
and paper fan-out. The targets are imaged for each color along the outside edges of the test form. The targets are large
(0.9423.78 in. or 24607 mm) extending almost completely around the printing cylinders. The digital ladder target is
divided into two or three vertical bands, with each band being comprised of 50%, 150-lpi (60-line/cm) straight-line screen
tints. The tints are of equal value, but their orientations (90°, 0°, and 90°) differ, as seen in the above callout.
The analysis of the digital ladder target is usually performed visually. However, at times, densitometry can be used
to quantify visually-observed density differences. The ideally reproduced target is one where the three different screen
orientations cannot be distinguished by the observer at a normal viewing distance. The uniformity of the digital ladder tar-
get bands should be confirmed on the films or plates prior to printing. If the targets are not uniform in tone on the film,
then a problem is indicated with the filmsetter. The targets should be even in density from gripper to trailing edge of the
printed sheet. If any of the three strips is noticeably different in density, a printing problem is indicated. The direction of
line orientation is influencing the density that results on the printed sheet. This can be due to slurring or doubling of the
lines. Closer examination with a magnifier of both the target and printed halftone dots will allow the user to distinguish a
slur from a double. Slurred dots are round with a comet-like tail; while doubled dots appear as two partially superimposed
dots where one is darker than the other. The direction of the slur is perpendicular to the screen orientation that had the
darkest printed density. If the darkened condition is uniform from gripper to tail, the directional gain is consistent around
the entire cylinder for that color. For example, if the center section (90° orientation) of the digital ladder target is darker
than the 0° orientation section, slur in the direction of travel is indicated.
This condition may be caused by a loose printing blanket. If the trailing section of the target shows cross-directional
density gain that the leading edge does not show, the cause might be paper movement (fan-out) during printing. When
the left-side digital ladder target and the right-side digital ladder target do not show the same conditions, uneven cylinder
pressures or paper slippage may be the cause.
If the central section of the target is consistently lighter from gripper to tail than the two outside sections, cross-cylin-
der movement and excess water may each be at fault. If there is excess water, the lines perpendicular to printing (central, 90°
orientation section) are lower in density than the lines parallel to travel. This is because the excess water will have the effect
 GATF Test Form User Guide 2-7

Figure 2-7b. Test variability with the fanout method.

Figure 2-7a. Analysis of digital ladder targets (simulated).

of washing across the cross-direction line elements. When a hand magnifier is used to examine the target, signs of small wash
marks may be seen.
An effective short-term remedy for many of the printing defects that the digital ladder target diagnoses is to reduce
the ink density in the affected area. Experienced press personnel will instinctively lower the ink feed to even the tonal
appearance of the printing. However, this solution is not ideal because the midtone range is saved at the expense of the
shadows and saturated colors. A better approach is to identify and cure the source of the excess directional gain so that
density aimpoints can be achieved without sacrificing midtones. The digital ladder targets on the Digital Sheetfed Test
Form 4.2 have solid ink targets before and after each target to monitor inking level in-line with the printed target. The
density aimpoints should be met to prevent misleading evaluation of the printed target.
In studies of process variability, there are many samples taken during the run according to the sampling plan appro-
priate for the experimental purposes. To analyze process variability with digital ladder targets, the sample sheets are
stacked in order and fanned out so that consecutive targets from one side of the sheet can be seen together. When suffi-
cient samples are available, this viewing is facilitated by cutting the samples perpendicular to the digital ladder targets.
The cut targets are then fanned out as shown in Figure 2-7b.
If the pattern of digital ladder targets produced by the fanned-out display shows noticeable variation, periodic image
transfer problems are occurring. Other periodic disturbances such as gear streaking, cylinder bounce, or inker variation,
are more easily seen when multiple samples are evaluated together.
The digital ladder target is a very sensitive device. If a sample is found in conformance for density, if all digital lad-
der targets are imaged such that none of the individual sections on any target are visibly distinct from the others, and if all
digital ladder targets are uniform in density from gripper to trailing edge, there is good assurance that many serious print
problems are absent from the printing system. This is a very good indication of overall reproduction quality.
2-8 GATF Test Form User Guide

Figure 2-8a. GATF star targets.

2-8b. Star targets indicating (from left) normal reproduction, dot gain, slur, and doubling.

Star Targets
The GATF star targets are also extremely sensitive to critical dot transfer problems that occur on press. The digital star
targets slightly differ from the familiar film-based version in that the digital targets do not have a consistant open area in
the center. The sharpness of the star target isinfluenced by the resolution of the imagesetter. Lower-resolution systems
will cause star targets to have larger solid filled-in areas in their centers. The closer an imaging system comes to presenting
the center of the star target as a single point, the higher its resolution is. Also, any directional irregularities in portraying
elements that are oriented at different angles will be seen as a non-circular star target center. If the star targets have ellip-
tical centers, then the resolution of the imaging system is higher in the direction of the short axis of ellipse than the resolu-
tion parallel to the long axis.
A star target is comprised of a circular pattern of alternating solid and clear wedges tapering to a very fine central
point. A quick visual analysis of a star target, aided by a hand magnifier, easily reveals when too much dot gain is occurring
or when a directional gain, such as slurring or doubling, is present. Figure 2-8b shows a variety of print problems dis-
played by GATF star targets.
When the central section of a star target is filled in to a greater extent than on the OK sheet, higher-than-expected
tint densities are the result. If the filled-in star target section is round, a nondirectional gain is indicated. This may also be
displayed as higher dot gain for that color. This condition may be caused by ink that has emulsified too much water. This
causes a decrease in tack, reducing the efficiency of ink transfer. Frequently, heavier ink films are run in an attempt to
meet solid density aimpoints.
If the enlarged central section of the star target is asymmetrical, the gain has a directional association. If the central
section is an ellipse, ink slurring is occurring. The direction of the slur is perpendicular to the long axis of the ellipse. If,
for example, a target had an elliptical center oriented across the test form, slurring around the cylinder would be the diag-
nosis. The cause might be paper slipping in the grippers or a loose blanket.
When the central filled-in area looks as if it has two centers, image doubling is the diagnosis. The causes might be
that a given ink is not setting quickly enough and it is re-imaging from a following printing blanket or from the same print-
ing blanket but slightly out of register.
 GATF Test Form User Guide 2-9

Image Fit Target

The image fit target is positioned below the woman’s portrait on the test form. This target is made up of four horizontal
bands (CMYK) that each have a series of colored triangles and circles positioned across them.
The pattern of triangles and circles is arranged to give all combinations of the other three process colors against each
process color background. Furthermore, the triangle and circle were chosen because these shapes presented horizontal,
vertical, diagonal, and curved lines forming the boundaries between colors.
The image fit target is used to evaluate the register accuracy of any output device (e.g., a digital press or proofing
system).
All the elements in the image fit target are placed with no trapping between colors. Butt fits like these are not used
in production work because the slightest amount of misregister in a printing system will result in objectionable white lines
between portions of the elements that are supposed to be touching. These white lines stand out against the colored back-
ground and are easily seen. The direction and magnitude of the misregister are determined from the location and thick-
ness of the white lines.
The image fit target can also be used to evaluate the effectiveness of image trapping programs that are applied to the
test form files. After image trapping has been applied, there should be no white lines in the image fit target even when
slight misregister is present. Close examination of some of the geometric element clusters shows the magnitude of the
trapping manipulation. As a rule, trapping is accomplished by spreading the lighter of the two colors and leaving the dark-
er color at actual size.
Close examination of the edges of the geometric elements will reveal thin lines at the boundaries that are combina-
tions of the two primary colors. For example, when yellow is spread under cyan, a thin green outline will show the magni-
tude of the yellow spread. Also, if a trapping program improperly spread the cyan instead of the yellow, then the shapes of
the geometric elements at the cyan/yellow interface would be distorted.
2-10 GATF Test Form User Guide

Mottle Patches
Ink mottle ranks high among the print problems that plague sheetfed lithography. Ink mottle, a visual defect, is
the nonuniformity of color or gloss in large areas of ink coverage. Smaller areas of coverage are sometimes insufficient to
detect mottling when it occurs. The condition is particularly troublesome in the overprint colors and solids, but it is some-
times found with single colors and tints as well. Causes of ink mottle can be paper- or ink-related.
The GATF Test Forms include large solid patches and 50% tint patches of each ink color, as shown above. These tar-
gets are of a sufficient size to evaluate ink mottle.
The evaluation of the printed mottle patches should be performed under controlled viewing conditions. Examine
the patches closely from many viewing angles. Patterns in the appearance of uniform color patches are undesirable. Begin
the investigation with the solids where the problem is most prevalent. When a new set of inks or a new paper is introduced
into a printing system, a comparative study of ink mottling at the outset may prevent expensive problems from occurring
later. When successive tests are run using a GATF Test Form, compare the results against samples from earlier trials.
 GATF Test Form User Guide 2-11

Cyan Yellow C-30

C- Y-
Y-28 Y-26 Y-24 Y-22 Y-20 Y-18 Y-16

C-7
C- Y-6
Y- Y-5 Y-4 Y-3 Y-2 Y-1 M-
M-28

M-
M-6 M-26

M-5 M-24

Magenta
M-4 M-22

M-3 M-20

M-2 M-18

M-1 M-16

C-80 Y-78 Y-76 Y-74 Y-72 Y-70 Y-68 Y-66 C-60 Y-58 Y-56 Y-54 Y-52 Y-50 Y-48 Y-46

M-78 M-58

M-76 M-56

M-74 M-54

M-72 M-52

M-70 M-50

M-68 M-48

M-66 M-46

Gray Balance Chart

The gray balance chart is a process standardization target used to determine the three-color halftone dot require-
ments for the cyan, magenta, and yellow films that will reproduce a neutral tonal scale at four different tone values. The
gray balance chart will only yield reliable information if the printing process is free of any serious problems and if the
process is operating within control limits. Each tone level of the gray balance chart consists of a field of tint patches with
gradually changing magenta values along the vertical axis and gradually changing yellow values in the horizontal direction.
The cyan dot size is constant for all squares in a given matrix, and is identified by the number in the upper left corner of
each matrix of squares.
Visually evaluate the printed gray balance chart to locate the most neutral patch in each matrix of squares.
Instrument measurement is also possible with a densitometer to establish process control aimpoints for a three-color
gray run control bar. A colorimeter can also be used for measuring the gray balance chart to investigate the colorimetric val-
ues of the squares selected as most neutral or to measure various squares to find the most neutral square (where the a* and
b* values are closest to zero). For visual evaluation of the gray balance chart, a gray standard of appropriate value is used as
a comparator. The gray standard most frequently used is a reflection gray scale. Sometimes 1⁄4-in. (6-mm) holes are punched
in the gray scale to isolate one printed gray patch at a time for evaluation. In some cases, the black tone scale is cut from a
sample printed sheet of the Test Form and used as the standard for comparison. Whether a photographic gray scale or a
printed black tone scale is used, a gray step of appropriate value serves as an aid in picking the most neutral patch from a
given field of gray squares. The square that comes closest to matching the gray standard under controlled lighting condi-
tions is identified as the most neutral square. The film dot values that printed the square are identified by the numbers that
head the rows and columns containing the selected square. When the most neutral squares are identified, the appropriate
dot values are recorded in the gray balance data table in the Sheetfed Press Test Analysis Form.
Use the gray balance values to construct a series of gray balance curves. The curves show the plots of film dot area
against the reflection densities (visual filter) of the selected most-neutral patches. Separate curves are produced for cyan,
magenta, and yellow. The four data points for each primary ink color are used to generate a best-fitting curve for that
color. The gray balance curves are used to decipher the dot area requirements for magenta and yellow when given any
cyan dot size.
Reproduction of near neutrals is particularly important because human tolerance for deviations in these hues is low.
Even very slight hue shifts are noticeable to most observers. A color separator identifies neutral elements in the original
photograph and assigns cyan, magenta, and yellow dot values taken from the gray balance curve representing those areas.
2-12 GATF Test Form User Guide

50C, 39M, 39Y

50K

Three-Color Gray Bars

The hue of the three-color gray bars is influenced by shifts in cyan, magenta, or yellow ink densities (50C, 39M,
39Y) or image transfer characteristics. A black tint of equal value (50K) is printed adjacent to the three-color target to serve
as a basis for making visual comparisons. Any changes in the hue of the three-color gray will be quickly noticed by the
unaided observer.
The ideal use of a three-color gray bar for process control involves providing customized dot sizes to fit the three-
color gray requirements of a printing system. The GATF Test Forms can be used to establish the appropriate dot size
relationships. The test form is printed to conform with aimpoints for density, dot gain, and trapping. If the evaluation of
the samples shows no significant printing problems, a most-neutral square is chosen from the gray balance chart at the
desired lightness level using an appropriate comparator. The dot values from the selected patch are used in producing
three-color run control bars. The tint value from the black ink tone scale found to be the closest match is used for the
comparator element to run adjacent to the three-color overprint.
A color scanner or color electronic prepress system can be used to image control bars with screen tints at the correct
percentages and screen angles. Furthermore, the size and shape of the bar can be customized to fit available trim space.
The three-color gray bars incorporated into the Test Form are not custom-fitted to a specific printing system;
instead, the SWOP 50% gray balance values are used. Under correct press operating conditions the three color tints may
not be neutral. However, this does not destroy their usefulness as process control targets. When the OK sheet is saved
from a pressrun, the specific condition of the three-color target becomes the standard for the run. The targets from sam-
ple prints are compared with the OK sheet target. Any shifts in hue, even toward more neutral grays, indicate changing
conditions on the press.
It is advisable to use a densitometer for analyzing the three-color gray bars. Take density readings along the bar
using the instrument’s primary filters (blue, green, and red). The reflection values from the OK sheet serve as the aim-
point for measured densities on samples taken during a pressrun. The direction of a density shift can be interpreted from
the changes in reflection density readings. For instance, if the green-filter reading increases in relation to the blue- and
red-filter readings, an increase in the magenta inking level is a likely cause.
Three-color gray bars printed across a form are very sensitive to slight variations between neighboring ink keys. If
there is a band of slight density shifts, the results will often be visually noticeable.
 GATF Test Form User Guide 2-13

Digital Proof Comparator

The GATF test forms contain a GCA/GATF Digital Proof Comparator 2.0, a native PostScript multi-element tar-
get that is used as a stand-alone target with digital proofing systems for process control and troubleshooting. The digital
proof comparator is included on the GATF Test Forms to provide a thorough analysis of imaging system resolution, and to
establish a reference for using the Digital Proof Comparator as a process control device.
The GCA/GATF Digital Proof Comparator 2.0 is the first native PostScript version of this device, which has been
used to monitor analog and digital proofing systems since 1984. (The first digital version appeared in 1990.)
The digital proof comparator reports information about the RIP. The information displayed includes the registered
user, PostScript version, horizontal resolution, vertical resolution, direction of travel, screen ruling, dot shape, and screen
angle.
Several pixel-based resolution elements are clustered together in the lower left portion of the target. In each of the
resolution patterns the imaging system is addressed in its basic units (pixels). The size in microns of the target elements
are computed and displayed as well. The first series of patterns are horizontal and vertical microlines (Figure 2-14a) in
positive (20/80) and negative (80/20) fields. Such patterns are displayed for 1-, 2-, 3-, and 4-pixel line widths. For high-res-
olution imaging systems these pixel line patterns are very sensitive exposure monitoring targets.
Adjacent to the pixel line patterns are the pixel checkerboard patterns (Figure 2-14b) in 1-, 2-, 3, and 4-pixel sizes.
The one-pixel checkerboard, for example, is made up of a pattern created by one-pixel-on, one-pixel-off exposure. This is
the first pattern of squares that the system can write. Some imaging systems cannot successfully hold a one-pixel checker-
board pattern. Each of the checkerboard patterns should appear as a medium shade of gray with equally sized black-and-
white elements.
The last group of resolution targets are the positive and negative semi-circles (Figure 2-14c) at 1, 2, 3, and 4 pixels.
Drawing a curved line is more difficult for an imaging system than a horizontal or vertical line. Thus, the one-pixel curved
lines are the most difficult challenge given to the imaging system.
All of the resolution targets are sensitive to exposure inaccuracies during imaging. The straight line and semi-circular
elements are also sensitive to directional bias in the imaging system.
The highlight/shadow section (Figure 2-15a) of the digital proof comparator is used to find the smallest and largest
2-14 GATF Test Form User Guide

Figure 2-14b. Figure 2-14c. Positive

Pixel checkerboard and
patterns. negative
Figure 2-14a. Positive and negative semi-circles.
microline targets.

reproducible dots of an imaging system. The dot values vary from 1 to 5% and from 95 to 90% for each process color. A
hand magnifier is used to find the first and last imaged dots for each color. These values form the end points for the tone
values that should be used.
Figure 2-15b shows a section of the GCA/GATF Digital Proof Comparator 2.0 that contains the black and three-
color tone patches and the total coverage patches. The black and three-color patches are imaged in the user-defined
screen ruling and dot shape on the right. They are repeated on the left at 133 lpi. The values of the three-color patches are
as follows:

25% = 25C 16M 16Y

50% = 50C 39M 39Y
75% = 75C 63M 63Y
100% = 100C 90M 90Y

The 25%, 50%, and 75% gray values are recommended by SWOP for reproducing neutrals with typical magazine
production techniques and materials. Having the black tint and three-color gray butt each other facilitates visual evalua-
tion of the neutrality of the three-color squares.
The two total coverage patches are at 400% (100C, 100M, 100Y, 100K) and 300% (80C, 70M, 70Y, 80K) values. The
400% is the darkest obtainable tone for the imaging system. The 300% patch is the darkest tone allowed by SWOP for
magazine production. These patches, as well as all the other tone value patches, are measured with a reflection densitome-
ter to quantify the findings. The values should be consistent between samples from the same imaging system, and they can
be used to compare the results of different imaging systems.
The GCA/GATF Digital Proof Comparator 2.0 contains 25%, 50%, 75%, and 100% tone values (Figure 2-15c) to
characterize the rendition of the process colors, plus blue, green, and red. There are two sections of this element that have
the same arrangement of patches but different dot shape and screen ruling.
The left side scales are imaged at the screen ruling and dot shape selected by the user at output. The right side
scales are imaged at 133 lpi with square dots, as specified by SWOP. The tone patches are measured with a densitometer.
Differences that are due to screen ruling and dot shape will be evident by the different readings from the two sides of the
target. The tone patches in these sections can be used to construct multiple hexagons at the 25%, 50%, 75%, and 100%
 GATF Test Form User Guide 2-15

Figure 2-15a. Highlight/shadow

section. Figure 2-15c. 25, 50, 75, and 100% tone values. Left-side
scales are imaged at the screen ruling and dot shape selected
by the user at output. Right-side scales are imaged at 133 lpi
Figure 2-15b. Black and three- with
color tone patches and total cov- square dots.
erage
patches.

levels. The use of concentric color hexagons as an analysis tool was described previously in this users guide.
The GCA/GATF Digital Proof Comparator 2.0 contains vignettes for each process color that graduate from 0 to
100% dot size (Figure 2-16a). The vignettes are composed of the halftone dots selected by the user. The vignettes are use-
ful for evaluating whether an imaging system is exhibiting “banding.” Banding is an objectionably abrupt tone jump in
what should be a smooth transition of tones. When the number of gray levels sup-ported by a color system is too low
banding will occur, but it can also be exhibited by high-resolution imaging systems due to a variety of mechanical and pho-
tochemical causes.
The transition of tones in vignettes of different process colors should be the same. For one instance, one color
should not change value more abruptly than the others.
The star targets (Figure 2-16a) that are located next to the vignette elements are used to indicate the resolution of
the imaging system and to display any directional bias that a system exhibits. When the filled-in center of the star target is
enlarged, lower resolution is indicated. If the central portion of the target is distorted (i.e., not round), a directional bias in
the imaging device is present. If the central portion of the star target is elliptical, for instance, the imaging system has
higher resolution parallel to the short axis of the ellipse and lower resolution parallel to the long axis.
The pictorial montage that occupies the center of the Digital Proof Comparator (Figure 2 16b) provides a carefully
composed photographic image to visually assess differences between output systems and successive samples from the
same output device.
The photographic image is used to assess tone reproduction, color correction, and contrast of imaging systems. The
woman’s face in the photograph provides familiar flesh tones for the human evaluation. Subtle differences in the contrast
of the output system will be seen as changes in the reproduction of the flesh tones. The yarns and other objects in the
photograph contain a variety of saturated and pastel colors from different areas of the color space.
Shifts in the color rendition of an output device would result in color differences in these objects. The slate surface
that the yarn and other objects sit on is a difficult shade of gray. Subtle differences in color rendition will cause noticeable
hue changes in this near-neutral surface. When an output system is calibrated with a GATF Test Form, the photographic
image in the Digital Proof Comparator is used as a visual reference against which later prints are compared.
Visual three-color gray bars (Figure 2 -16c) are also contained on the GCA/GATF Digital Proof Comparator 2.0.
The 25%, 50%, and 75% gray values on these bars are from SWOP:
2-16 GATF Test Form User Guide

Figure 2-16c. Three-color gray bars.

Figure 2-16a. Vignettes and star targets for each process color.

Figure 2-16b. Pictorial mon-

tage.

25% = 25C 16M 16Y

50% = 50C 39M 39Y
75% = 75C 63M 63Y

Each of the three-color gray patches is adjacent to a black tone patch of 25%, 50%, or 75% value. The black tone
patches provide visual references for accessing the neutrality of the three-color patches.
Each of the gray patches is divided vertically into two halves of matching cyan, magenta, and yellow combinations.
The left halves of the targets are reproduced with the user-defined screen ruling and dot shape, while the right halves are
composed of 133-lpi square dots. Since different screen rulings will exhibit different tone value changes in an imaging sys-
tem, the two halves may not match visually. Such differences are readily apparent with the arrangement of patches on the
Digital Proof Comparator. It is possible to measure these target areas with a hand-held spectrophotometer to quantify dif-
ferences or access neutrality.
The stand-alone version of the GCA/GATF Digital Proof Comparator 2.0 should be used to monitor digital
output devices such as proofing systems. The combination of the photographic image with the precise native PostScript tar-
gets provides a highly analytical measuring device that is also easily used and visually accessible.
 GATF Test Form User Guide 2-17

Transfer Grids
It is important to evaluate two types of registration during a pressrun: external register of the printed elements to
the substrate and internal register, or fit, between color separation halftones. Maintaining tight registration between the
four colors is essential for good color reproduction. The photographic originals are perfectly registered, since the three
dye images (cyan, magenta, and yellow) have not left the substrate on which they were formed. In color separation, the
information from each of these dye layers is recorded on a separate film to be recombined on the printing press with pri-
mary inks. This separation process leads inevitably to some degree of misregister in the recombined image. A GATF study
found that about 0.004 in. (0.1 mm) of misregister is the limit for many subjects beyond which appearance of the color
reproduction noticeably suffers (GATF Research Project Report No. 6114, Control of Color Register, by George W.
Jorgensen). Many printing plants have a rule-of-thumb limit for misregister tolerance of half a row of dots.
Where once the simultaneous four-color fit among many images on a large multi-page press form was the respon-
sibility of a film assembly department, the task is now commonly performed with color electronic prepress systems.
Analysis of register on the printed Test Form is best carried out by evaluating the eight GATF transfer grids. The printed
image of a single transfer grid is shown above. On the 2538-in. Digital Sheetfed Test Form 4.2, the transfer grids are
positioned along the inside edges of the digital ladder targets and consist of a single vertical row 15 crosses in length.
The transfer grid is a matrix of crosses, with each color cross printed by two units of the press. Each cross is accom-
panied by two numbers to identify the press units that imaged it. When the two crosses are indistinguishable from each
other due to being superimposed, good register has been achieved between those two cylinders. By displaying all of the
two-cylinder printing combinations in a single matrix, the transfer grid eases the task of determining which of the press
cylinders to move in relation to the others to most efficiently achieve register during makeready. The amount of misregis-
ter across and around the cylinders an be measured on the transfer grids with a magnifier that is equipped with a reticle.
Enter this information in the image register data table.
If warranted, short- and long-term variability studies of misregister can be performed to reveal how much register
variation to expect from a printing system.
2-18 GATF Test Form User Guide

% of Cyan Cyan Cyan Yel. Yel. Yel. Mag. Mag. Mag.

M-C-Y 95% 90% 85% 95% 90% 85% 95% 90% 85%

95%

92%

89%

86%

83%

80%

77%

74%

71%

68%

65%

62%

Color Correction Target

The purpose of the color correction target is to provide a method for determining the color correction specifica-
tions required to accurately reproduce optimal hues of red, green, and blue in a printing system.
When viewed under controlled lighting conditions, the rows of patches for blue, green, and red show slight changes
in hue based on gradual adjustments to the dot percentage of the overprint colors.
For example, if cyan is held constant at 95% while magenta is decreased from 95% to 62%, the result is a range of
blues with subtle hue changes. At the bottom of the first blue column, the magenta component of blue is reduced to 62%
while cyan is held at 95%. This combination of dot percentages yields an overprint blue that appears too cyan. At the top
of the column, both the magenta and cyan components have 95% dot sizes. This causes the magenta component in the
overprint blue to appear too dominant.
The best rendition of pure blue is likely to have dot percentages close to 95% cyan and 80% magenta for this combi-
nation of ink, paper, and press. Further evaluation will reveal that approximately 95% yellow and 85% cyan will yield an
optimal hue of green and 95% magenta and 80% yellow will produce the best hue of red.
Determining the optimal dot percentages for blue, green, and red overprints with a specific printing system is criti-
cal to decisions regarding color correction in the color separation process. For example, if you scan a photograph featuring
a clear blue sky with white clouds, and the reading for the sky is 80C, 50M, 10Y, and 0K, the blue in the sky will appear
too cyan on the press sheet. If your goal is to render a realistic-looking sky, it will be necessary to add about 15% more
magenta to the separations. In this manner, the pure blue, green, and red hues established with the color correction target
provide a reference point for color correction adjustments.
 GATF Test Form User Guide 2-19

74 77 80 83 86 89 92 95
64 68 72 75 79 83 87 90
64 68 72 75 79 83 87 90
302 313 324 333 344 355 366 375

100

298 309 320 329 340 351 362 371

96

294 305 316 325 336 347 358

92

290 301 312 321 332 343

88

286 297 308 317 328

84
SWOP 80 300
282 293 304 313 70
70
80 80

278 289 300 400% 100 400

100
76 100
100

Ink Coverage Target

The ink coverage target is used to determine the maximum dot area coverage that will be used for a printing sys-
tem. Maximum dot area is the sum of the dot areas of the four-color separation films. The target consists of a matrix of
printed patches at varying amounts and proportions of dot area coverage, as seen in.
The ink coverage target is laid out with the black dot areas varying by rows from 76% to 100% coverage. The three-
color coverage is varied by columns from 202% to 275% total dot area coverage. The specific cyan, magenta, and yellow
dot values for the target are displayed above.
The printed sample of the ink coverage target is read with a densitometer, and the black density values are entered
into the ink coverage density table in the Sheetfed Press Test Analysis Form.
The analysis of the density readings in the Ink Coverage Density Table will lead to a process specification for the
maximum dot area coverage to be contained in the color separation films. The determination of maximum dot area is also
based on considerations of ink transfer and drying. With the widespread use of programmed color reduction such as gray
component replacement (GCR), a need exists to clearly set the limit of how much undercolor (or cyan, magenta, and yel-
low process inks) can be removed before the maximum black of the system is affected. When a great deal of GCR is
desired in the midtones, the shadow undercolor often becomes too low. In this case, the undercolor addition (UCA) func-
tion puts back cyan, magenta, and yellow into shadow areas until the maximum film dot areas have been met. The maxi-
mum dot coverage aimpoint is based on in-house conditions. As a starting point, many sheetfed plants allow 340-360%
total coverage. The SWOP specifications for color separations prepared for heatset magazine publication was raised in the
1986 revision of SWOP from 280% to 300% total dot coverage.
2-20 GATF Test Form User Guide

Twenty-step Tone Scales and Vignettes

There are tone scales for each of the printing units near the center of the test form. The tone scales are indexed
in 5% steps from 5% to 100%. Each patch is 55 mm, which is sufficient to be conveniently measured with most hand-
held densitometers.
The screen ruling, dot shape, and screen angles of the tone squares are not predetermined; they will be applied
by the user at the time of output.
The tone scales also contain highlight/shadow patches. The tone scales are used to measure thedot gain of the print-
ing system. This data can be used to construct dot gain curves.
Each tone scale is accompanied by a vignette, where the tone is gradually changed from 0% to 100%. The printed
vignettes are evaluated for any signs ofbanding, which is an objectionable, abrupt change in the tonal gradation.
 GATF Test Form User Guide 2-21

Dot Size Comparator

The test forms include a native PostScript Dot Size Comparator in black only to determine the tone value
changes that result from using different halftone screen rulings.
The numbers to the right of each row show the screen ruling in lines per inch, while the numbers that head the
columns give the percentage of dot coverage. Halftones with higher screen rulings and also stochastic screens experience
higher levels of dot gain than do coarse line screens. When two unlike systems are being calibrated together, like digital
proofs and printing presses, the value increase must be accounted for. Poor tone reproduction is acknowledged to be a
more critical defect in a picture than is poor color reproduction. The tone structure in a photograph establishes the appar-
ent lighting conditions for the scene.
The dot size comparator is measured with a densitometer to calibrate the dot values of the various patches. A series
of four-point dot gain curves can be constructed at the five different screen rulings contained in the target.
These curves clearly show differences and abnormalities in the dot value increases associated with different imaging
systems.
2-22 GATF Test Form User Guide

Single-Tier Control Bar

There are two color control bars incorporated into the GATF Test Forms, the first is a single-tier control bar which runs
along the trailing edge. It includes the following elements: repetitions of solid patches for KCMY plus spot colors, 25%
screen tints for KCMY plus spots, slur/highlight targets for KCMY plus spots, repetitions of 50% screen tints for KCMY
plus spots, repetitions of RGB overprints, and 75% screen tints for KCMY plus spots. Overall dimensions for the color
control bar vary from form to form. The size of an individual patches also depending on the form.
The single-tier control bar is convenient for use in balancing out ink key settings during make ready. It contains a
wide enough variety of elements that, if tight trim space permits the use of only one bar, there are sufficient patches for
measurement of dot gain, print contrast, and ink film trapping.
 GATF Test Form User Guide 2-23

Six-Color Two-Tiered Control Bar

The two-tiered control bar, imaged across the leading edge of the test form is an ideal process control bar to use
after a printing system has been calibrated with the GATF Test Form. Master films for the GATF Test Forms were pre-
pared to provide a match between the double-tiered control bar on the form and the corresponding stand-alone target.The
unique advantage to using it on a press previously tested with the GATF Test Form is that values for all attributes measured
by the bar can be obtained from the printed test form. This confirms that the press is in good operating condition and that
process aimpoints were met when the sample was selected to act as a comparator for later pressruns.
The six-color two-tiered control bar is two patches in height by 114 patches in length with dimensions of 0.3922.47
in. (10.057.1 mm). Individual patches measure 0.200.20 in. (5.05.0 mm) and ink zones are 1.18 in. (30 mm) long.
The top tier of the bar is made up of solid ink patches that include 19 repetitions of KCMY plus 5th and 6th colors while
the bottom tier consists of an initial paper white patch followed by 3 repetitions of 25%, 50%, and 75% tint patches for
KCMY plus 5th and 6th, two repetitions of star targets for KCMY plus 5th and 6th, two repetitions of Dot Gain Scale II©s
for KCMY plus 5th and 6th next to blue, green, and red overprints, two repetitions of slur/highlight targets for KCMY plus
5th and 6th, and blue, green, and red overprints in adjacent solids and 50% tints.
This control bar can be used for the following diagnostic and process control procedures:
• Solid ink density can be adjusted across the press with solid targets reading every 1.18 in (30 mm).
• Ink trapping levels are readily measured for blue, green, and red.
• 50% dot gain can be measured for all six colors.
• Print contrast for KCMY plus 5th and 6th colors can be measured with 75% tint patches and adjacent solid ink patches.
• Quarter-tone dot gain can be measured for KCMY plus 5th and 6th from 25% tint patches and adjacent solid ink patches.
• Hue and saturation for overprint midtone tints can be measured from the 50% overprint patches of blue, green, and red.
• The three-color gray reproduction of the printing system can be compared to a standard or OK sheet using the three-
color gray 25% tint patches.

For construction of concentric hexagons, solids and 50% tints of single colors and two-color overprints are needed.
The six-color two-tiered control bar provides a number of measuring options in a minimal amount of space. It combines
the capacity for frequent monitoring of ink zones on one tier with a variety of diagnostic elements on the second tier. It
also combines quickly-assessed, visually sensitive elements with precise tint patches intended for instrument evaluation.
2-24 GATF Test Form User Guide

A B C D E F G H I J K L M N

10

11

12

13

IT 8.7/3 Basic Data Set

A standard color field, the IT8.7/3 Basic Data Set is included on most of the GATF Test Forms. This data set is
described in ANSI IT8.7/3-1994. The basic data set is used to test the reproduction of a variety of tertiary colors, or colors
that are made up of components of all three process-color inks plus black. The colors in the IT8 color set were chosen to
provide good coverage of the color gamut and to facilitate color space transformations and gamut mapping.
The patches in the basic data set are specified by a column letter and a row number (e.g., B13 refers to the blank
patch in the 13th row of column B).
The basic data set is the best target on the test form to use for color management purposes. Color management
involves the calibration of color imaging systems through three stages: input (scanners), processing (monitors), and output
(dye sub, ink jet, electrostatic).
A hand-held spectrophotometer is used to measure the transmission and reflectance characteristics of the color
monitor and the printed output. Selected patches from the basic data set are projected on the monitor and enlarged as
necessary to be measured. The file is then output and the corresponding patches are measured on the color print. The
spectrophotometric measurements from the monitor and the print are used to adjust for color appearances for a better
match.
There are various software programs available that ease the task of calibration. Inevitably, they require proprietary
color field targets to be used. If these programs are in use, the basic data set on the test form can serve to test the effec-
tiveness of the calibration.
 GATF Test Form User Guide 2-25

The GATF/RHEM Light Indicator

(simulated), as shown under a
5000K standard light source.
Stripes do not appear.

Stripes do appear when the Light

Indicator (simulated) is viewed
under non-standard light, such
as incandescent.

Stripes will also appear when the

Light Indicator (simulated) is
viewed under other non-
standard lighting, such as cool or
warm fluorescent.

GATF/RHEM Light Indicator

A target area is identified in the lower portion of the test form for a GATF/RHEM Light Indicator. The light
indicator (simulated above) is a process control device that aids in identifying viewing conditions that do not have the rec-
ommended (according to ANSI PH 2.30-1994) standard color temperature of 5000K.
The GATF/RHEM Light Indicator is a pressure-sensitive sticker (available in rolls of 50 from GATF) that is coated
with a stripe pattern made from two carefully formulated inks. The inks are a deliberate metameric pair, being optimized
to match under 5000K and not to match under lights of other color temperatures. The light indicator acts as a warning
device alerting the evaluator when it is not wise to make color judgments.
Since subjective quality control depends on matching color appearances between unlike colorants (e.g., between
sublimation dyes and printing inks) it is imperative that all judgments be made under identical (ANSI standard) viewing
conditions.
 Testing
Color Reproduction
The GATF Color Printing Test Form is designed to challenge the printing performance of a press and the color
reproduction characteristics of a printing system. Printing performance should be evaluated first to eliminate any mechan-
ical deficiencies in the printing system.
The GATF Test Forms include several photographic images. Transparencies of these images are available from
GATF for a single package price as product 7125PT. GATF’s separations are made with tone and color characteristics
suited for sheetfed printing. If a printing plant receives supplied separations in its normal work flow, users will find the
GATF separations similar to the films received from trade houses. As a whole, the set of seven GATF Quality Control
photographs provides the opportunity for comprehensive analysis of a printing system in terms of how closely it can match
standards for photographic realism.
If the color scanner is being tested along with the sheetfed printing press, then a special procedure is requierd to
substitute the user-generated color separations for the GATF separations. The GATF photographs must be obtained in
the appropriate format (transparency or reflection print) for the scanner under the test.
 Unless special arrangements are made with GATF, tradi-
tional film assembly techniques must be used to substitute
the user-made color separations for the GATF color separa-
tions. The following procedure can be used to perform this
application.
• Scan the replacement images to the characterizations that you have determined to be correct
• Place the GATF Test Form .eps file into a page layout program
• Draw a new picture box over the existing images
• Place the replacement image into the new picture box
• Output the films or plates as you normally would
 GATF Test Form User Guide 3-3

Low-Key/Red Couch

The red couch photograph is referred to as a low-key image because of the predominance of dark hues in the
midtone-to-shadow portion of the tone reproduction curve. In order for this image to be reproduced with enough tone
compression to facilitate printing with offset lithography or a direct digital press, it is necessary for the midtone placement
to be shifted toward the shadow end of the tonal scale so that increased contrast can provide sufficient separation between
the dark hues in the image.
These dark hues include reds, greens, and blues plus a variety of finished wood tones that provide reproduction
challenges in most parts of the color gamut. If separations are produced using a CCD (charge-coupled device) scanner,
the red couch photograph may be particularly useful since CCDs typically have difficulty in capturing subtle hue differ-
ences in the midtone-to-shadow region of a photograph.
Another aspect of the reproduction difficulties presented by the red couch is the high level of resolved detail con-
tained in the original. The effort to reproduce this degree of detail requires high-resolution scans and large file sizes that
place greater demands on color separators and electronic prepress facilities.
3-4 GATF Test Form User Guide

High Key/Wedding

The wedding photograph is referred to as high key because the dominant tones are light pastels and whites, which fall in
the highlight-to-midtone portion of the tone reproduction curve. For this image to be reproduced accurately with enough
tone compression for offset lithography or a direct digital press, the midtone placement must be shifted toward the high-
light end of the tonal scale so that there is sufficient contrast to differentiate one light hue from another.
White and off-white hues were chosen for the wedding image because of the difficulty in maintaining color purity
without introducing unwanted color casts. Small shifts in color balance resulting from a printer’s effort to match the more
saturated midtone hues are easily detected in the neutral whites and light grays, such as the wall behind the largest flower
arrangement.
In evaluating tone reproduction in the wedding photograph, placement of the smallest printable halftone dot (typical-
ly 3% on coated stock) is especially critical in order to hold nuances of detail in the window shears. If the dot placement is
too low, the window appears overly bright with no detail. Conversely, if the halftone dot is placed too high, such as in a
specular highlight, the photograph will appear dull due to insufficient brightness in the highlights. Note that there are spec-
ular highlights in the window shears that are intended to drop out of a halftone rendition and reproduce as paper white.
 GATF Test Form User Guide 3-5

Group Portrait/Painting Kids

In the painting kids photograph the challenge is for color separators and printers to reproduce all five skin tones equally
well. Races represented in the image from left to right are Caucasian, Latin American, Japanese, African American, and
Indian.
Light and medium gray T-shirts worn by the models in painting kids are useful for evaluating neutral gray balance.
As press operators make adjustments in ink density to fine-tune the various flesh tones, slight hue shifts will become
apparent in the gray tones. Because the image also contains patches of saturated color in the form of smeared paint on the
T-shirts and the background, selective color correction must be properly adjusted to maintain these hues without skewing
gray balance.
3-6 GATF Test Form User Guide

Female Portrait
Because the most frequently reproduced flesh tones are those of female Caucasian models, the female portrait is
a photograph that emphasizes this challenge to process color printing. Close attention is required to capture the differ-
ences in tonality between face, chest, arms, and hands and the more subtle transitions between forehead, cheeks, nose,
and chin.
Secondly, the model’s brunette hair and the black cloth in her outfit provide an additional challenge to color separa-
tors and printers in attempting to maintain highlight detail in the hair while holding shadow detail in the hair and cloth.
The brown canvas background used for the female portrait contains light tones that are sensitive to subtle hue shifts.
If gray balance is skewed to finesse flesh tone reproduction that was not properly rendered in the separations, a color cast
will be visible in the highlight-to-midtone region.
 GATF Test Form User Guide 3-7

Gray Neutrals
The gray balance rendition in a photograph is most readily determined by examining the tones approximating
neutral grays. It is these areas in which color casts are most visible. Accurate color correction is dependent upon first
establishing proper gray balance. In the gray neutrals test image, the subject is primarily composed of gray tones. While
some appear neutral, other grays are slightly warmer or slightly cooler. The only saturated hues are those provided by the
color accents.
What makes the gray neutrals photograph difficult to print accurately is that when the neutral balance is optimized
for a given gray tone, any remaining color cast is readily visible in the surrounding gray hues.
In addition to its extreme sensitivity to color casts, the gray neutrals image is also low-key due to the preponderance
of dark tones with relatively few highlights. As with the red couch photograph, this image requires greater tone compres-
sion in the midtone-to-shadow portion of the tone reproduction curve in order to maintain sufficient contrast for good
tone separation. The combination of tone compression and gray balance requirements gives the gray neutrals photograph
the distinction of being the most difficult QC photograph to reproduce accurately.
3-8 GATF Test Form User Guide

Color Gamut/Fruit
The color gamut/fruit photograph was illuminated with electronic flash and serves the purpose of providing a broad color
gamut. However, the fruit photograph is shot from a closer perspective which includes fewer compositional elements. For
this reason, the areas of saturated color are larger, but the overall amount of resolved detail is less.
The fruit and fabrics were selected for their variety of saturated hues which correspond to the outermost points in
the color gamut of transparency films. With the narrower gamut of process color inks, reproduction of these hues is partic-
ularly difficult.
 GATF Test Form User Guide 3-9

Memory Colors/Covered Bridge

Although the covered bridge photograph has the look of a familiar postcard, its purpose is to evoke the viewers’
memory-color response. Hues that we have encountered many times before, in this case blue skies, green vegetation, a
faded red bridge, and a white rail fence, are stored in our visual memory and can be recollected with greater accuracy
than colors we see less frequently. When looking at a photograph that contains memory colors, the viewer subconsciously
evaluates them first to determine whether or not the color rendition is credible.
The covered bridge photograph is composed entirely of memory colors. It is a useful test of color reproduction qual-
ity because most viewers are well acquainted with these hues. Small variations in color rendition are quickly noticed, so
the color reproduction must closely approximate the original scene.
The covered bridge photograph also serves as an excellent color reference for a naturally lit outdoor scene.
Atmospheric conditions included unobstructed early evening sunlight from a predominantly clear blue sky with a scattering
of cumulus clouds. Because the sun was positioned behind the photographer, the scene was evenly lit from the front, creat-
ing maximum saturation in all the elements—bridge, fence, vegetation—with minimal shadows. The combination of direct
sunlight plus diffused light from a blue sky created a daylight color temperature approximating 5000 K, the reference for
standard viewing conditions.
 Analysis of Print
Attributes
This section describes the analysis and evaluation of print attributes.

The attributes include:

• Solid ink density
• Dot gain
• Print contrast
• Ink trapping
The use of concentric GATF color hexagons as an analysis tool is described. This allows the simultaneous monitoring
of several attributes in a visually sensitive format.
The analysis presented here is limited to measurable print attributes. Some important print attributes are not readily
quantified. Several print defects (i.e., hickeys or mottle) that can cause poor color reproduction may not be accounted for
in the measured attributes.
4-2 GATF Test Form User Guide

Color-Measuring Instruments
The measurement of color during press operation is performed
with a reflection densitometer. It is commonly known that densito-
meters do not always agree with each other. The instruments differ
in spectral response, reading aperture size, and illumination geom-
etry. The difference in spectral response between wide- and nar-
row-band densitometers is responsible for large differences in
measured densities (see Figure 4-2). Status-T response refers to a
particular wide-band response specified as a standard for graphic
arts by the American National Standards Institute (ANSI). (ANSI
PH 2.18-1984).
As more Status-T densitometers come into use, the inter-
instrument agreement can be expected to rise. A quality control
target called the GCA T-Ref provides a calibrated series of ink-on-
paper patches. The T-Ref allows the user to test a densitometer for
conformance to Status-T response.
Within a process control operation, the greatest accuracy is
gained by using the same densitometer for successive measure-
ments. If this is not practical, then having the same model of
instruments calibrated to the same plaque will provide increased
confidence in agreement between instruments. Either narrow- or
wide-band response can be used to evaluate the Test Form. A
microprocessor-assisted instrument is very useful for performing
calculations as well as measuring functions.
The reading aperture of a densitometer refers to the size of
the spot it will measure. Common sizes are 1⁄16 in. and 1⁄8 in. (1.6 mm
and 3.2 mm). Of these, the 1⁄8-in. aperture is preferred for analysis
of the GATF Test Forms. This is because there are no small-area
measuring needs on the test form and because larger-aperture
instruments have been found to give more consistent dot area
readings.
Interfacing the densitometer with a printer or, better still, a
personal computer can speed the analysis task. Ideally, a two-direc-
tional scanning densitometer can be used to input measurements
into a custom software program that would analyze the readings
and produce a complete profile of the printing system in minutes.

Figure 4-2. Narrow-band (top) and wide-band

(bottom) spectral responses.
 GATF Test Form User Guide 4-3

Figure 4-3b. Spectrophotometric curve.

Figure 4-3a. CIE L*A*B color space.

Colorimeters are currently being promoted for use in the graphic arts for process control. A colorimeter is simi-
lar to a densitometer in that it measures reflected color, but colorimetric measurements are based on the spectral
response functions of a standard observer. These values can be plotted in visually consistent color diagrams, like the
CIE L*A*B* space (see Figure 4-3a).
The CIE L*A*B* color space is a convenient means of plotting colors for comparative purposes. Whether colorime-
try will come into common use in the pressroom remains to be seen. One system is being used for making readings within
the printed image. These readings are compared between the OK sheet and samples taken during the pressrun. Changes
in the colorimetric measurements can be interpreted in perceptual terms. Changes in ink density are recommended based
on shifts in the inter-image measurements.
Colorimetric values do not completely characterize the complex reflectances of sample colors. For this level of
color measurement, a spectrophotometer is required. A spectrophotometric curve (Figure 4-3b) is produced based on
the reflectances of the sample at closely spaced (e.g., 10-nm) intervals across the visible spectrum.
Spectrophotometry is used to verify that inks match a standard, such as in the NAPIM/SWOP Ink Verification
Program. If two colors match spectrophotometrically, they will match under all conditions of viewing.
4-4 GATF Test Form User Guide

Solid Ink Density

Examine the density variation across the test form, around the printing cylinders, and between the various samples select-
ed for analysis.
First, measure the solid ink densities across the test form to test conformance to the target ink densities. Use the
solid color control bar across the trailing edge of the test form and the color patches following the digital ladder targets for
this purpose. There are ten solid ink patches for each color across the test form. Read the fifth and sixth colors with the
densitometer filter that provides the highest numbers. This will vary depending on the light absorption properties of the
specialty colors being used.
The densitometer used for this analysis will affect the readings. If the densitometer used during the pressrun is also
used for this analysis, differences in density can be attributed to dryback of the inks. If two different densitometers are
used, their spectral responses should match; otherwise, the readings will show significant differences.
Enter the density readings in the table labeled “Densities Across the Test Form” in the Sheetfed Press Test Analysis
Form. Also enter the density aimpoint, upper tolerance limit, and lower tolerance limit for each color into the table.
Establish aimpoints for the fifth and sixth (specialty) colors by reading a color sample, perhaps from a color atlas book,
with the densitometer used for press control.
A graphic representation of density profiles can
be useful to communicate this information at a glance.
Ink density profiles can also be constructed for this
purpose. The aimpoint and tolerance levels for each
ink should be indicated on the graphs as in Figure 4-4.
These density profile graphs show how much variation
there is in ink density across the test form and the
readings are to the aimpoints.
Measure the density variation around the print-
ing cylinders for each color on both sides of the test
form. Locate the solid ink stripes between the digital
ladder targets and use them for this purpose. Record
the density values in the table labeled “Densities
around the Printing Cylinders” in the Press Test
Analysis Form. There are twelve small black marks
evenly spaced along the cyan solid stripes to indicate
points for measurement for the solid ink stripes.
Record the average densities, high density readings,
and low density readings. This data should be carefully
examined to see if there are any identifiable trends Figure 4-4. Ink density profiles.
such as consistent falloff of densities from the leading
to the trailing edge of the form.
Analysis of density variation during a pressrun is of interest in establishing process controls for printing. Variation
may occur between several consecutive sheets or between successive samples taken from a long run. The latter variation is
more indicative of the conditions found on live jobs, but the cost in time and materials for testing long-run variations is
higher. When variation is tested over time, the press adjustments made by the operators affect the overall variability found
in the printing system. Ideally, judicious adjustments counteract changing conditions on the press and reduce variability;
however, this is not always the case. It is a useful experiment to balance the press after makeready and then make no
adjustments whatsoever during the course of a long run. Perform a variability study on samples taken under these condi-
tions to identify the natural variations occurring in the printing system. Next, make a second pressrun where the operators
adjust the press during the run as they normally would. A comparison of the results from these two pressruns will show
how effective the press operators are at reducing variation during press operations.
When sheet-to-sheet variation is tested, use the same measuring target for each consecutive measurement. A solid
ink patch over the center portion of the test form should be selected for this purpose. Record the data in the process vari-
ability data table in the Sheetfed Press Test Analysis Form. The data table can accommodate twenty-five readings for each
color. If more than twenty-five samples are taken, use multiple copies of the table.
 GATF Test Form User Guide 4-5

Identify the sampling plan when you complete the process variability data table. Less variation is expected from fifty
consecutive sheets than from fifty sheets pulled at regular intervals during a pressrun of 10,000. Deviations of density
between different colors are not directly comparable. A density difference of 0.10 in yellow is a larger percentage of the
average density value (about 1.05) than a difference of 0.10 density units in black, where the average density value might
be 1.60. Unfortunately, it is not sufficient to calculate a simple ratio (coefficient of variation) to put these values on the
same scale because density readings do not represent ratio level data; instead, they are a logarithmic function of the
attribute of reflectance. Comparison of density deviations for different ink colors is further complicated by two physiologi-
cal factors. First, people are not as sensitive to changes of reflectance at low intensity (high densities) as they are to
changes of reflectance at high intensity (low densities). Second, the situation is complicated by human observers with
varying spectral sensitivity. The just noticeable difference (JND) thresholds between colors vary widely in the average
observer. For example, the JND values are low for neutral and near-neutral colors. Maintaining gray balance is a particu-
larly important and difficult aspect of high quality color reproduction. This is due, in part, to the low tolerance for devia-
tion on the part of observers. (Gray balance here refers to more than reproducing a neutral scale with cyan, magenta, and
yellow inks; it refers to maintaining the hues of all near-neutral colors within an image.)
The standard deviation for a set of ink density values estimates the amount of deviation one would expect between
the group mean and any single density reading selected at random. The standard deviation is useful for establishing the
process control limits for each ink density. Traditionally, density tolerances, such as +0.05, have been assigned somewhat
arbitrarily as process control limits. This can cause problems if the normal variability of the printing system produces sam-
ples that fall outside of the tolerance limits as a function of random sampling. Knowing the standard deviation of the den-
sity readings allows the user to estimate the percentage of randomly selected samples that will be within tolerance limits if
the printing conditions have not changed. If the population of density readings describe a normal distribution, 68% of the
samples will fall within one standard deviation of the mean and 95% of the samples will fall within two standard deviations
of the mean. It is not advisable to set density tolerance limits at 1 standard deviation because under unchanging press
conditions nearly one-third of all sample readings will show an out-of-tolerance condition. This would cause the press
operators to make unnecessary adjustments. This, in turn, will increase the variability of ink density and, therefore, the
variability of color appearance being experienced during the pressrun.
More reasonable process control aimpoints are 2 standard deviations from the mean. In this case, an out-of-toler-
ance reading would only have a 5% chance of occurring due to the random variations of the process. If the density win-
dows described by 2 standard deviations seem unacceptably wide, decrease the variability of the system before tighter
process control limits are specified. This is accom-
plished by eliminating sources of variability.
To create a graphic representation of density
changes during the run, construct an attribute vs.
time graph, as shown in Figure 4-5. The attribute in
this case would be ink density, which is plotted along
the y-axis. Plot successive samples for each sample
reading, and then connect the points to complete the
graph. More than one color can be plotted on a set of
axes, or individual graphs can be for made each color.
An examination of the attribute vs. time graph
will show trends that are occurring during the press-
run. You may discover either periodic disturbances in
the system or slow trends. For instance, if the ink
fountains are refilled at several points during a long
run, there may be a density jump immediately fol-
lowing the addition of new ink. Or, if there is a grad-
ual density falloff over time, it should be evident
from this graph. If there is more variability early in a
run than there is later on, it can be seen from this
graph. This condition may be caused by changing
rheological properties of the inks in the early stages
of the pressrun. Figure 4-5. Attributes vs. time graph.
4-6 GATF Test Form User Guide

You can obtain a graphic representation of the

distribution of densities for each color by plotting a
frequency polygon. To undertake a plot first determine
the range of densities. Then divide the range into
intervals of equal size. Calculate the number of sample
readings falling in each interval and use this informa-
tion to construct the frequency polygons. Use the
attribute frequency data table in the Sheetfed Press
Test Analysis Form to record the intervals and fre-
quencies for each color. The data table can also be
used for other print attributes aside from ink density.
First label the table with the attribute to be analyzed
and the sampling plan. Then record the interval mid-
points in the table, and the number of density readings
found within each interval.
Plot the frequency polygons from the data
recorded in the attribute frequency data table. First, Figure 4-6a. Frequency polygon
label the axes of the graph. Label the x-axis with the
interval midpoint values, and the y-axis with the fre-
quency levels. Then plot a point above each interval
corresponding to the number of density values that fall
within that interval. Connect the plotted points to cre-
ate the frequency polygon, as shown in Figure 4-6a.
The shape of the polygon should approximate a
normal distribution. In fact, using the standard deviation
to estimate the percentage of samples that deviate from
the mean by chance alone assumes that the population
of density readings represents a normal distribution.
You may wish to calculate the numerical indices
of skewness and kurtosis, to estimate the extent to
which a density distribution fits the shape of the nor-
mal distribution. The skewness describes the degree of
asymmetry in a distribution. A skewness of zero is a
symmetrical distribution with equal concentrations of
Figure 4-6b. Skewness of distributions.
higher-than-average and lower-than-average density
values (Figure 4-6b). If the skewness coefficient is
negative, the population is skewed towards higher val-
ues (that is, the mean is greater than the median).
Positive skewness coefficients describe distributions
skewed towards lower values in which the mean is less
than the median.
The kurtosis coefficient describes how tightly
clustered the values of a distribution are about the
mean. The normal curve has a kurtosis coefficient of
zero (Figure 4-6c). Negative kurtosis coefficients indi-
cate distributions that are less clustered than the nor-
mal distribution, and positive kurtosis coefficients indi-
cate more closely clustered distributions than normal.
To calculate skewness and kurtosis coefficients,
first normalize the data to the standard z-scale. The
Press Test Analysis Form contains a z-score transfor- Figure 4-6c. Kurtosis of distribution.
mation table to assist the user in organizing
 GATF Test Form User Guide 4-7

this data. Convert the individual density

readings to z-scores then use the z- Z-value: Skewness Kurtosis equation:
scores to calculate the two indices of equation:
xi - x  z i3  z i4
skewness and kurtosis for each ink 2 =
zi= 1 = –3
color, according to the formulas shown s n n
in Figure 4-7.
The skewness coefficient for distri- Figure 4-7. Skewness and kurtosis coefficient formulas.
bution of ink densities should have a mini-
mum value. If the values reach zero, the variations exhibited by the densities of the samples are distributed symmetrically
about the average value. A negative skewness means that there is more density variation in the values below the average
value than there is in above the average values. The further the skewness coefficient is from zero, the more pronounced the
effect will be. Positive skewness values bear the same relationship in the opposite direction: there is more variation in densi-
ty values above the mean. When a density distribution shows a high skewness value in either direction, a hypothetical expla-
nation should be formulated to explain the anomaly. For example, if a set of density readings is showing a low negative
skewness, it might indicate that the average value is close to the maximum density that can be achieved with a given ink.
This would cause densities on the high side to have a ceiling or limit. Adjust density aimpoints downward in this case since
dot gain will be high and difficult to control when printing at the maximum density attainable with an ink.
Different criteria are used to interpret the kurtosis coefficient. Higher values are better, since high coefficients of
kurtosis indicate tight clustering of the samples around the mean value. Positive kurtosis values are better than negative
ones. A large negative kurtosis coefficient implies that the density readings do not exhibit a strong central tendency. It is
difficult to establish process control limits in this case because many of the sample readings are distant from the popula-
tion average.
Studying printing process variability is an expensive and time-consuming pursuit, but there are benefits for the dili-
gent printer. Both troubleshooting and process control are improved by careful analysis of the variability information of
ink densities and other print attributes.
Using a computer with statistical software can greatly reduce the time and effort needed to make calculations of
mean, standard deviation, skewness, and kurtosis.
Ink density variation is only one of several attributes that you should monitor during printing. Others include dot
gain, ink trapping, and print contrast. The study of variability described for density deviations can be repeated for any of
these print attributes. The interpretation of standard deviation, range, skewness, and kurtosis will aid the quality assurance
personnel in characterizing the variability inherent in the printing system.
4-8 GATF Test Form User Guide

Dot Gain
Dot gain refers to the difference between dot sizes on color separation films and the apparent printed dot area of those
same dots on paper. Dot gain is known to have a more pronounced effect on the visual appearance of a print than ink den-
sity. A slight shift in dot size of a few percent will produce a visually noticeable shift in hue for many colors. All printing
systems have some amount of dot gain resulting from a number of factors. First, it varies for different original film values,
meaning that highlight dots are subject to a different magnitude of dot gain than midtone or shadow dots. Second, the
screen ruling influences dot gain. Finer screen rulings will experience more dot gain than coarse screens. Third, dot gain is
comprised of both physical and optical components. The light scattering properties of the substrate have an influence on
dot gain. Fourth, true dot gain is easily confused by a host of other printing problems including slur, doubling, or toning.
Finally, dot gain or loss occurs at every point of image transfer. Great effort may be expended in troubleshooting a printing
press to alleviate dot gain that is actually due to excessive platemaking exposures.
Several targets are included on the GATF Test Forms for measurement of dot gain. To characterize the dot gain of a
printing system, it is referenced to the midtone. Read the fifty percent tint patches on the printed sheet with a densitome-
ter then calculate apparent dot area using the Murray-Davies equation:

1 – 10–DT ADA = apparent dot value

%ADA =  100 DT = density of tint
1 – 10–DS DS = density of solid

Many densitometers are equipped with microprocessors to make this calculation automatically. It is important to cal-
ibrate the densitometer to an adjacent solid ink patch and to the substrate when making these readings. Some confusion
may arise from the use of the Yule-Nielson equation. The Yule-Nielson equation is a special form of the Murray-Davies
equation that is used to isolate mechanical or physical dot gain from optical dot gain. This is performed by an “N-factor”
that relates to the light scattering/absorbing properties of the substrate. If the N-factor is set at one, the Yule-Nielson
equation becomes the Murray-Davies equation. When dot gain is measured for process control, it is most convenient to
measure total dot gain (physical plus optical) using the Murray-Davies equation.
Although the 50% film dot is used to describe the dot gain level of a printing system, it is not sufficient information
for the color separator who must prepare films to fit the conditions of a printing system. For this purpose, a dot gain curve
is more useful. The dot gain curve plots the dot gain as a function of original film dot area. A halftone step scale or tone
scale is used to measure the dot gain at several discrete steps from highlights through shadows. The Sheetfed Press Test
Analysis Form provides a dot gain table to record the dot gain values for each color.
The Sheetfed Press Test Analysis Form also has a dot size comparator data table to compare the dot gain values that
occur at four screen rulings—150, 175, 200, and 300 lpi (60, 69, 79, and 118 lines/cm). When color separations or halftone
films are prepared for printing, the amount of dot gain to allow for is influenced by the screen ruling. There are both
advantages and disadvantages to using finer-than-normal screen rulings.
The principle advantage is the increased ability to render fine detail, and
the principle disadvantages are reduced press stability and increased lev-
els of dot gain.

Dot Gain Curves

The information from both the dot gain table and the dot size
comparator data table is used to construct dot gain curves. The Sheetfed
Press Test Analysis Form contains axes for dot gain curves. The dot gain
curve depicts the original film dot (x-axis) vs. the total dot gain (y-axis).
More than one dot gain curve can be plotted on a set of axes. For
instance, the dot gains of several different ink colors or several different
screen rulings can be compared directly by putting them on the same
axis, as shown in Figure 4-8.
Dot gain curves between colors are directly comparable (unlike ink
density levels) because dot gain calculation involves the relationships of
tint density to solid density. Dot gain is expressed as a percentage which Figure 4-8. Dot gain curves.
 GATF Test Form User Guide 4-9

is a common scale for all colors. Ideally, the dot gain curves for various colors are the same, but different colored inks have
different characteristics. A midtone dot gain difference of no greater than 4% was established by the SWOP Committee as
the allowable process control specification for proofing.
The most meaningful target dot gain levels are internal shop standards developed over time as the lowest and most
consistent dot gain that can be maintained. As a starting point, it may be useful to consider some industry-wide averages.
The North American Commercial Print Survey, conducted by E. I. duPont de Nemours & Co., Inc., reports the following
average midtone dot gains from sixty-one participating printers:
Yellow 18%
Magenta 20%
Cyan 20%
Black 23%

The 1986 Recommended Specifications for Web Offset Publications (SWOP) sites dot gain levels for sheetfed press
proofing, off-press proofing, and web production presses as follows:
Sheetfed press proofing 17% ±4%
Off-press proofing 22% ±3%
Web production printing 24% ±4%

Printers who are using positive-acting plates will have about 6% lower dot gain than printers using negative-acting
plates. This is because positive-acting plates cause dot sharpening (loss), while negative-acting plates cause dot gain. Any
gain or loss during platemaking is included in the total dot gain values since these express the net dot-size difference from
film to print. Taking density readings on printing plates is unreliable because of the reflectance characteristics of the
grained aluminum base. The Dot Gain Scale II© is a useful target for assessing the amount of gain on negative-acting
plates because no densitometer is needed.
The shape of the dot gain curve should be evaluated. The usual curve shape will not be symmetrical; instead, it will
have slight positive skew. There will be larger dot gain levels in the 25-50% range than in the 50-75% range. Ink colors
exhibiting dot gain curve shapes that differ noticeably from the other inks should be the subject of further inquiry.
The dot gain curves comparing screen rulings are only available for cyan, magenta, yellow, and black. Each ink color
is plotted on a different set of axes. All four screen rulings for a given ink are plotted on the same axis. This facilitates eval-
uation of the dot gain increases associated with finer screens.

Variability Studies
If a variability study of dot gain is undertaken, the procedure is the same as that described for studying density varia-
tion. In fact, the same samples are used. The study would usually be limited to the midtone dot gain only. All of the fol-
lowing graphs and charts would be applicable to the study of dot gain variation:
• Process variability data table
• Attribute vs. time graph
• Attribute frequency data table
• Frequency polygon
• Z-score transformation table

When two print attributes, such as density and dot gain, are both studied from the same samples, it is beneficial to
test the correlations between them. The correlation index is a measure of the strength and direction of the linear associa-
tion of two variables. The densities and dot gains are correlated for each ink color. It would not be fruitful to pursue corre-
lations between yellow density and magenta dot gain; only magenta density should be correlated with magenta dot gain.
There are many correlation coefficients that can be used with different types of data. The Pearson Product-Moment
Correlation Coefficient fits the requirements of the application to print attributes. It is useful to examine a scatterplot of
the two variables when testing the correlation between them. The scatterplot is a graph where each axis is used to repre-
sent values of one of the attributes. Points are plotted for each associated pair of readings; for example, the density and the
dot gain for a selected color, as seen in Figure 4-10a.
The scatterplot is examined to identify extremely deviant scores known as outliers. Often such points represent
4-10 GATF Test Form User Guide

errors in measurement rather than real values. They will cause the cor-
relation coefficients to be artificially low. Also the type and strength of
the relationship can be seen from the scatterplot. If the relationship
between the two variables is nonlinear but still not random, it can be
seen in the scatterplot.
Computation of the correlation coefficients can be accomplished
from the transformed z-score values of the two attributes, according to
the following formula:
 zx i zy i
rxy =
n–1

The Press Test Analysis Form contains a z-score product table to

assist the user in calculating correlation coefficients.
When the correlation coefficient is calculated from transformed z-
scores, it is equal to the covariance of the two attributes. The correlation
values can range from –1 to +1. A correlation of zero indicates that
there is no relationship between the two variables and that they are
changing independently of one another. A correlation of +1 means that
there is a perfect positive relationship; that is, as the value of one vari-
able increases the value of the second variable will increase in an
absolutely predictable way. A correlation of –1 describes a perfect nega-
tive relationship where increases in one variable correspond with
decreases in the second variable. Correlation values that are between
zero and +1 describe various degrees of imperfect positive relationship
where both values tend to increase together but not in a perfectly pre- Figure 4-10a. Scatterplot of density vs. dot gain.
dictable manner.
When two attributes, such as density and dot gain, are examined
together, it may be desirable to predict levels of one variable from a
knowledge of the second variable. For example, it may be desirable to
predict the amount of dot gain to expect if the process control aimpoint
for ink density is set at a specified level.
Regression analysis can be used to make such predictions. The
accuracy of the prediction will be dependent on the correlation that
exists between the two variables. By convention, the y-axis value is the
predicted value, while the x-axis is used for the predictor variable.
Linear regression is used when the relationship between the two vari-
ables is linear. This relationship is graphically illustrated by a best-fitting
straight line that is drawn through the scatterplot of the two variables, as
shown in Figure 4-10b.
The equation for the linear regression is also shown in Figure 4-
10b. To make a prediction for a Y value, the corresponding value ofis
substituted into the equation. In the example shown here, the desired
ink density is substituted into the regression equation. The resulting
predicted Y value is the dot gain level that one would expect from that
ink density. As with descriptors of cultural tendency (standard deviation,
skewness, and kurtosis), the correlation and linear regression calcula-
tions can be quickly made by a personal computer equipped with the
appropriate software. Figure 4-10b. Scatterplot and regression line density
and dot gain.
 GATF Test Form User Guide 4-11

Print Contrast
A print attribute that is closely related to dot gain is print contrast. It is a numerical index based on the relationship of the
printed density of a 75% film dot patch vs. the solid density. The formula is:
Ds – D75 PC = print contrast
PC =  100 Ds = density of solid
Ds D75 = density of 75% tint

Some microprocessor-assisted densitometers perform print contrast calculations automatically. The print contrast
values for the six ink colors can be conveniently measured in two locations using the six-color single-tier control bar on the
tail edge of the test form or the six-color two-tiered control bar on the lead edge.
Enter the print contrast values into the print contrast table of the Sheetfed Press Test Analysis Form. The average of
four print contrast values is calculated for each sample. Conventionally, print contrast is only calculated for the single ink
targets, however, contrast values for the overprint colors are being examined as attributes to monitor during printing.
Higher print contrast values are better than low ones. The higher the print contrast, the greater the number of tones
that can be distinguished between the three-quarter tones and solid ink density. For most original photographs, the tone
scale has been severely compressed in the shadow region during the color separation process.
Therefore, printing systems that maintain a maximum number of shadow tones are closer in appearance to original
photographs. The aimpoints for print contrast, like dot gain aimpoints, are best based on shop standards. The average
print contrasts reported by duPont in the North American Commercial Print Survey are reasonable starting aimpoints for
process control. They are as follows:
Yellow 37
Magenta 41
Cyan 39
Black 43

You may wish to construct frequency polygons that depict fre-

quency vs. print contrast. The print contrast data can be normalized,
and the skewness and kurtosis coefficients for the distribution can
be calculated to assist in troubleshooting and process control in a
way similar to that used with solid ink density.
Calculate the correlations between print contrast, ink density,
and dot gain. A negative correlation will exist between dot gain and
print contrast. As dot gain increases, print contrast will decrease. The
relationship between density and print contrast should be a curvilin-
ear one; that is, as ink density increases, print contrast increases until
a point is reached where further increases in density cause lower
print contrast values. This is due to the excessive ink film thickness
needed to raise densities beyond this point which cause dot gain to
rise at an accelerated rate, thereby decreasing print contrast.
An example of a scatterplot of ink density and print contrast
shows this relationship (Figure 4-11). The relationship shown in
here can be used to establish or evaluate the process control aim-
points for ink density and print contrast. It is desirable to obtain the
highest print contrast that the system can deliver in a consistent
manner. This value will be near the apex of the curve. The associat-
ed ink density and print contrast values can be deciphered from a
best fitting curve drawn through the scatterplot. This value may
need to be adjusted when consideration of other print attributes,
such as ink trapping, are taken into account. Also, printing problems
like doubling will become more pronounced at higher ink film thick-
nesses and may necessitate further adjustment to the process control Figure 4-11. Scatterplot of ink density and print control.
aimpoints.
4-12 GATF Test Form User Guide

Ink Trapping
Ink trapping is a print attribute that relates to the two-color overprints of green, blue, and red. It estimates the amount of
ink that is transferred to a previously printed ink film, compared to the transfer of ink to an unprinted substrate. The print-
ing sequence must be known to make trapping measurements. The Preucil equation for calculating ink trapping is
T = trapping
Dop – D1
Dop = density of overprint
T=  100
D2 D1 = density of first-down ink
D2 = density of second-down ink
Microprocessor-assisted densitometers can automatically perform trapping calculations. It is important that the den-
sity measurements be made with the filter appropriate to the second-down ink color.
The ink trapping values should not be interpreted as identifying the percentage of the second-down ink that trans-
fers to the first ink film. The failure of the law of additivity of ink densities prevents this relationship from holding true.
Process control aimpoints for trapping are based on internal standards. The highest achievable values are sought.
There are substantial differences in trap values based on the printing sequence. There are differences in trapping
due to the spectral response of the densitometer. Several different equations can be used to calculate trapping values; e.g.,
the Childress equation, the Hamilton equation, and the Brunner equation. Therefore, process control aimpoints should be
based on the highest achievable values that can be attained with a given printing system.
The Press Test Analysis Form contains an ink trapping table to provide the user with a means of recording trap val-
ues. The single-tier control bar across the top of the GATF Test Form and the two-tiered control bar across the bottom
contain single-color patches and two-color overprints that facilitate measurement of ink trapping at five locations. The cal-
culations of trapping for each color are averaged together to arrive at a trapping value for a sample sheet.
Ink trapping is dependent on the rheological properties of the inks, which are dynamically variable during a press-
run. Ink trapping can be monitored as a run control
attribute. As ink trapping shifts, so do important hues in the
shadow section of the color gamut. Many originals have
important greens, blues, and reds as well as dark near-neu-
trals and tertiary colors making the transfer of ink to previ-
ously printed ink films an attribute worth monitoring during
a pressrun. When a variability study of a printing system is
undertaken, the trapping values are graphed against time as
shown in Figure 4-12. This enables the user to see any
trends that are occurring in ink transfer during the pressrun.
For example, the green trapping might be decreasing con-
tinuously during the run, perhaps due to excessive levels of
water take-up in the yellow ink. Under these conditions, the
dark green vegetation in a color reproduction would be
gaining a bluish cast as the green trapping of yellow ink
decreased. The press operator might reduce the cyan ink
density to alleviate the problem, but the blues in the repro-
duction would shift towards magenta as a consequence.
When analyzing the ink trapping variability data, it is
desirable to test correlations between ink trapping and den-
sity levels of first-down inks, or between trapping and the Figure 4-12. Ink trapping vs. time.
other print attributes. Also, off-line test results, such as
ink/water emulsification levels or ink tack ratings, can be
examined in relation to ink trapping.
 GATF Test Form User Guide 4-13

Concentric Hexagon Analysis

The data from the GCA/GATF Proof Comparator III
can be used to plot a series of four concentric GATF
Color Hexagons. To plot color hexagons the red, green,
and blue filter densities are needed from the cyan,
magenta, yellow, red, green, and blue patches. Enter
the density data in the color hexagon density data table
of the Press Test Analysis Form. A method for plotting
points on color hexagon graph paper is shown in Figure
4-13a.
Plot points for cyan, magenta, yellow, green, blue,
and red at each of three tint levels (25%, 50%, 75%)
plus solids. Then draw line segments to connect the six
points at each level. The result will be four concentric
irregular hexagons of increasingly larger size, as shown
in Figure 4-13b.
Also plot the paper densities on the hexagon as a
reference point. Draw arc segments between the points
of the same hue with increasing dot area. Plotting hexa-
gons by hand is a laborious task; if your densitometer
interfaces with a personal computer, the plotting can be
expedited.
The principle advantage of the concentric hexa-
gons for a print analysis tool is that all the print attribut-
es commonly used for process control are represented
visually on a single diagram. These attributes include Figure 4-13a. Plotting points on a color hexagon.
cyan, magenta, and yellow ink densities; dot gain; print
contrast; and ink trapping. If concentric hexagons are
plotted from the color OK sheet, this graph can be
compared to others plotted from samples taken during
the pressrun. A light table can be used to superimpose
the concentric hexagons from the sample over the con-
centric hexagons plotted from the OK sheet. The
assessment of changes taking place between the OK
sheet and the sample is done visually and quickly. The
analysis of the superimposed graphs can include evalua-
tion of the following print attributes:
Solid ink density changes are visible as differences
between the 100% cyan, magenta, and yellow plotted
points. As density increases the prints will fall farther
from the center along the same radial lines. If the solids
shift radially, a shift in hue is indicated. If a change is
seen on the superimposed hexagon graphs, examine the
related readings from the color hexagon density data
table. A change in values that is approximately propor-
tional across the three densitometer filters results from
a change in inking level. A nonproportional change in
density readings would result from ink contamination.
Dot gain changes are easily seen at the three tone
levels (25%, 50%, and 75%) as shifts in the locations of
the cyan, magenta, and yellow points at those levels. Figure 4-13b. Concentric color hexagon.
4-14 GATF Test Form User Guide

Changes in the locations of the 50% points will be the most evident, since the dot gain is near a maximum value for mid-
tones. Shifts in the positions of these points, particularly changes in the quarter-tone (25%), should prompt careful exami-
nation of the color reproduction. Rendering of highlight-to-midtone values is critical for most subjects.
Print contrast changes can be seen as shifts in the relative locations of the solid and 75% tint points. The farther the
distance between solid and 75%, the greater the print contrast. Changes in print contrast resulting in lower values indicate
reductions in the amount of shadow information conveyed in the color reproduction. This information was conveyed as
halftone film dots between 75% and the maximum printing dots of about 97%. It is usually the case that print contrast has
a strong negative correlation with dot gain, so that as print contrast decreases dot gain will increase. In instances where the
two indices of print contrast and dot gain are not following this pattern, an explanation should be sought. For example, if
high dot gain is experienced in a given color, the press operator may reduce the ink level in that color well below tolerance
limits to compensate for the dot gain. This will cause the dot gain to come down, but it might also result in a low print con-
trast index.
Ink trapping can be deciphered from the plotted positions of the red, green, and blue overprint points. The satura-
tion of the overprint colors is indicated by their plotted distances from the center of the hexagon. Ideally, the hue angles of
the overprint colors at increasing tint levels lie on a straight line bisecting the angle described by the hues of the two com-
ponent ink primaries. This represents 100% transparency and 100% transfer of the second-down ink. In the ideal situa-
tion, the saturation of the overprint color, as with the primaries, would plot an equal distance from the center. Under real
printing conditions, the overprint hues are shifted towards the first-down ink color because the second ink has not trans-
ferred at 100% efficiency. The points of increasing tint levels do not lie on a straight line. They curve toward the first-
down color as ink coverage increases. The saturation of overprints does not fall an equal distance from the hexagon center
as do either of the component primaries. Instead it is less saturated, causing a flat and off-center portion of the 100% hexa-
gon.
When sample hexagons are superimposed on the OK sheet hexagons, any changes in trapping will be apparent by
changes in the overprint point positions. Color hexagon plots are a good means for comparing the printing conditions of
successive print tests. The method of superimposition provides immediate visual comparison of density, dot gain, print
contrast, and ink trapping.
 GATF Test Form User Guide 4-15

Color Reproduction
Image Register
Do the images fit? Examine the reproduction for register of the halftone dots. Misregister can be mistaken for image soft-
ness in the original, either intentional or accidental. Image softness may result from inaccurate lens focusing, subject
movement during long exposures, climatic conditions such as fog, snow, rain, high humidity, and/or particulate air pollu-
tion, or by lens attachments such as diffusion filters.
Use a hand magnifier to aid in estimating misregister, even though the effects of misregister are clearly visible to the
unaided eye. Pay close attention to the edges of the image. Are any of the process colors distinguishable as extending
beyond the image limits? Maintaining good register is increasingly difficult as more color reproductions are added to the
printing form. Compromises are inevitable when two or more color images are printed simultaneously. The dimensional
instability of the substrate can also cause misregister difficulties.

Image Sharpness
How sharp is the detail in the color reproduction? This issue is related to but not the same as color register. A misregis-
tered image will not have sharp detail, but a perfectly registered image also may not have sharp detail. Image sharpness is
influenced by several factors, from the sharpness of the camera lens to the resolving power and contrast of the photo-
graphic materials. The color separator can influence the perception of image sharpness with the unsharp masking function
on the scanner. Detail in a poorly-focused original can be greatly improved with electronic image enhancement, but too
much peaking can become distracting—detail is essentially outlined with dark and light edges.
When evaluating the effectiveness of the detail enhancement in a reproduction, the color proof and original photos
can serve as points of reference. The final judgment of “how much is enough?” is a subjective call based largely on the
requirements of the individual photograph. When the detail in a reproduction looks unnaturally prominent, there may be
too much electronic peaking. Examine the edges surrounding the compositional elements in the image with a hand magni-
fier to determine the amount of edging that has been added electronically.

Highlight and Shadow Reproduction

Is the full tonal range used for the color reproduction? Examine highlights, shadows, and fully saturated colors in the pic-
ture. Specular highlights should be reproduced as unprinted paper, the lightest possible value on a press sheet. Diffuse
highlights should reproduce as minimum printing dots, the first tonal value darker than unprinted paper. The Digital
Sheetfed Test Form 4.2 provides a basis for measuring the minimum printing dots of all colors in the highlight/shadow
patches of the GATF Plate Control Target. If the minimum printing dots in a reproduction are too big, the value differ-
ence between specular and diffuse highlights becomes unnaturally large and the visual appearance of the reproduction is
compromised. Larger dots may create too much highlight density and cause the image to appear less lifelike.
At the shadow end of the tonal scale, the darkest value should equal the maximum printing density of the system.
Use the ink coverage target to measure the maximum density. If this point has been reached, and simultaneously low dot
gain is achieved, the shadow contrast will be good. The measured attribute of print contrast should be correspondingly
high. This will enable the printing system to register a maximum number of tonal levels in the shadows which, in turn, will
cause the color reproduction to look more lifelike, since it will better approximate the tonal range of the photograph.

Tone Reproduction
Is the tone reproduction good? If your examination of the extremes of the tonal scale found good highlight and shadow
rendition then the reproduction has good contrast, but may still lack good tonal separation in the midtones. Does the pro-
portion of light and dark tones look realistic? Is the transition between tones smooth and even? Are the tonal differences
and sense of lighting from the original captured in the reproduction?
Evaluate tone reproduction under standard viewing conditions. If the photograph contains people, pay particular
attention to their skin tones. Harsh lighting of people in the original may be exaggerated in the reproduction which causes
high contrast and strong shadows. If proper tone reproduction is lacking, it may be necessary to decrease midtone density
values to attain good skin tones.
4-16 GATF Test Form User Guide

The tonal range of the original nearly always exceeds the range of the reproduction. Therefore, a color proof limited
to the tonal range of the reproduction is used to interpret the tones of the photograph to the printing scale. It is useful to
have both the photographic original and the color proof available for comparison when evaluating tone reproduction. The
original tonal values are determined by the lighting in the scene, the color response of the film, and any deliberate manip-
ulations performed on the photograph.
Under some circumstances, the evaluator may prefer a tone reproduction that does not match the original faithfully
but repositions certain tones. For example, it may be preferable to reproduce someone’s face lighter than it is found on
the original if that person was standing in shadows in the original scene.

Gray Balance
Is the gray balance good? Viewing the color reproduction under controlled viewing conditions together with a color stan-
dard such as the original, a proof, or an OK sheet is the final test of the success of the total effort of color reproduction.
The last evaluation is necessarily subjective, since color perception is itself a cerebral function. The evaluation of gray bal-
ance relies on the human interpretation between reality (the color standard) and expectations (the memory colors). For
some scenes, favoring the expected colors over the actual colors will improve the human response to the reproduction.
Examples of this include brightening and saturating a blue sky more than in the original, or reproducing an element that
the viewer would expect to be neutral, as neutral gray, in spite of the fact that it was off-neutral in the photograph.
To evaluate gray balance, two aspects are examined. Does the reproduction have the same gray balance as the origi-
nal? Does the gray balance of the reproduction viewed alone look realistic, or is there a hue shift in one or more elements
that seems off? The color distribution of the light source that illuminated a scene influences the hues in a photograph of
that scene. For example, a photograph made outdoors late in the day will have a warm reddish cast from the setting sun.
Two scenarios are possible. First, if the photograph contains substantial white elements, it may be desirable to eliminate
the warm cast and maintain neutral balance. On the other hand, if the photograph contains people, the warm glow impart-
ed by the light may be emphasized in the color reproduction. There are no absolute rules to determine which approach
should be taken with a given original.
 About GATF
The Graphic Arts Technical Foundation is a nonprofit, scientific, technical, and educational organization dedicated to the
advancement of the graphic communications industries worldwide. Its mission is to serve the field as the leading resource
for technical information and services through research and education.

For 73 years the Foundation has developed leading edge technologies and practices for printing. GATF’s staff of
researchers, educators, and technical specialists partner with nearly 2,000 corporate members in over 65 countries to help
them maintain their competitive edge by increasing productivity, print quality, process control, and environmental compli-
ance, and by implementing new techniques and technologies. Through conferences, satellite symposia, workshops, con-
sulting, technical support, laboratory services, and publications, GATF strives to advance a global graphic communications
community.

The Foundation publishes books on nearly every aspect of the field; learning modules (step-by-step instruction booklets);
audiovisuals (CD-ROMs, videocassettes, slides, and audiocassettes); and research and technology reports. It also publishes
GATFWorld, a bimonthly magazine of technical articles, industry news, and reviews of specific products.

For more detailed information on GATF products and services, please write to us at 200 Deer Run Road, Pittsburgh, PA
15143-2600 (Phone: 412/741-6860) or visit our website at http://www.gatf.lm.com
 Other GATF Products of Interest
Order
Number
19×25-in. Sheetfed Test Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7297*
Digital Test Form 4.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7246D
Newspaper Test Form Kit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7187*
Digital Newspaper Test Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7187D
QC Photographs (Prints) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7125PP
QC Photographs (Trans) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7125PT
QC Photographs (Slides) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7125PS
QC Photographs (Digital) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7125PD
GATF Digital Plate Control Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7120D
GATF Plate Control Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7120
GCA/GATF Digital Proof Comparator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7145D
GCA/GATF Proof Comparator III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7145*
GATF/RHEM Light Indicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7065R

Products with * are available in either positive or negative form.

Negative is reflected by suffix of N, positive reflected with P.