Home' FLEXO Magazine : February 2016 Contents Mark will present on EG color profiles at Forum 2016 as part
of the session titled “New Tools for Your Prepress Work-
flows,” scheduled for Sunday, March 6 at 3 p.m . Catherine
Haynes, All Printing Resources, Inc. and Brad Taylor, Du-
Pont Packaging Graphics, will chair the session.
Other presentations in this session include:
• What Does a 50 Percent Pantone 281 Look Like?,
from John Seymour, QuadTech
• Color Communication Across the Silos, from Ray
going beyond eight samples is rarely worth the cost, as random varia-
tion is already reduced about as much as it can be.
For most production printing processes, measuring multiple charts
from within a sheet yields better data than measuring the same chart
at the same location on multiple sheets. Chart 1 is an eight stage de-
scending, two page chart, printed three up on a sheet and run on press
using a sequential press characterization run. This is from an actual
pressrun at Clemson University conducted for an EG seminar in No-
vember 2015. Decreasing patch set size from nine stage descending to
eight stage descending enabled increasing the test chart from two up
to three up per cylinder. In this specific case, we decided to accept the
“cost” of a smaller patch set to get the benefit of having more charts
STARTING WITH MANY MEASURE CHARTS
While we may never know the exact trade off between patch set size
and charts per cylinder, we do know that the starting point of creating
a good profile is to have multiple sets of measured data—six to eight
total charts of each of the four sectors (CMYK, Omyk, cGyk, cmVk).
The task now is to convert this into a single data set—a 7-color profile.
There are six steps in the creation of a 7-color profile. Part one of this
article focused solely on the first step. Here, we will address steps two
Steps to 7-color Profile Creation
1. Create, print and measure the profile charts
2. Visually inspect/repair the data
3. Average the measurement files
4. Smooth the single averaged file
5. Synchronize the data to a curve specification
6. Merge the color sectors into a single profile
Good test target measurement software will include an inspection
function (step two) as part of the measurement function (step one),
so the user can “see” bad data immediately after scanning each chart.
Good profile creation software will also enable the importing of data
measured in other locations with other software. Hence, steps one and
two may be tied together or separate. However, the concepts are most
easily understood if we consider data inspection as a separate step that
occurs after test chart measurement.
VISUALLY INSPECT / REPAIR DATA
After measuring the data on an automated measurement device, you
will have 24 or more separate measurement files to inspect (six samples
x four color sectors). This can only be done with preview software to
view each measurement file. A critical aspect of such preview software
is the ability to display the data in either the exact format as measured
or in a “visually ordered” format. Inspection in a visually ordered
format is critical to visual detection of errors. Measurement error tends
to be random in nature and it is not possible for a human observer to
see random error in a random target. It is not until the target is put in
visually ordered form that a human can “see” the bad data.
The samples in Image 1 and Image 2 are from the EG seminar at
Clemson University in November 2015. Of all the samples measured,
only two—which happened to be in the orange sector—had bad data.
In both cases, the bad data could not be seen when viewed in mea-
sured format, but was immediately visible when viewed in visually
Procedures for handling bad data are beginning to emerge in the
industry. The first rule is to repair only the extremely obvious bad
data points. Most “bad” data that is not extreme will be repaired in the
Image 1 & Image 2: Measurement inspection software should include
the ability to inspect the data in either the measured format (above) or in
a visually ordered format (below).
FEBRUARY 2016 | FLEXO 123
Links Archive January 2016 March 2016 Navigation Previous Page Next Page