Graphics Programs Reference
In-Depth Information
Sidebar
Averaging Measured Data:
Even though we may be using Spectrophotometers that measure
more than one sample of color per patch, there are cases where averaging multiple targets
from a single printer can produce increased accuracy in a printer profile. For devices that drift,
due to any number of issues like ambient temperature, humidity, or other factors, averaging a
group of targets is useful. If a device drifts slightly over a number of days but drifts in a con-
sistent fashion, averaging the measured data can produce a printer profile that attempts to hit
the middle of the preverbal barn. For a device like a printing press that has deviations from the
beginning to the end of a press run, averaging measured data taken from multiple targets can
produce results that are more predictable. Naturally, a device has to conform to some level of
consistency or no number of data averaging will produce a printer profile that adequately
describes the device.
The first task is to examine how the device varies over time. This can be accomplished by
comparing several measured data files output over a period of time. One tool that I use
routinely for this task is the GretagMacbeth
MeasureTool (Compare)
module. If a user opens
two measured data files with the same number of patches and patch layout, this module will
show some very useful statistics, as seen in Fig. 6-16. I could output and measure a standard
color target used to build a profile or for that matter, any target of any patch size (as long as
the two targets I want to compare are the same) and analyze the differences. Once I compare
two or more targets and see that the average deltaE is sufficiently low (something in the neigh-
borhood of 6 or less), I can then take the two data files and average the data to build a profile.
I can use as many measured data files as I wish to produce a new set of average data. This
work of averaging can be done in an application like Excel; however, using a tool like
MeasureTool is easier and provides the deltaE statistics necessary to see if this undertaking is
worthwhile.
The other useful function of the
Compare
mode in
MeasureTool
is for process control.
If, for some reason, the color management for a printer seems to be suspect, I can output a
new target, measure, and compare the older data file to the newer data file. If the deltaE is
low, this is an indicator the color problems are elsewhere in the color management
pipeline. For output devices that use silver papers and chemistry, process control is critical to
consistent results. Using this technique allows a user to see how consistent their printing
process is from day to day or any other length of time. An automated Spectrophotometer is
ideal for these types of environments. With packages that produce small postlinearization charts,
a user could print out such a target each morning and compare the measurements from previ-
ous day. If the deltaE is low enough, proceed with printing. If not, use this target to update the
profile.
Using the compare mode in
MeasureTool
is also an excellent way to evaluate when your
print is dry and stable, which is useful for ink-jet users. Measure a target after it has been
printed, then several hours later and perhaps the next day. By comparing the deltaE from the
same targets, you can get an excellent idea of what changes, if any, occur as the print dry. The
compare function can be used on the same target over time to evaluate print fading or other
changes in the colors originally measured. Alternatively, measure the same target several times
in a row and compare the readings to see how consistent your Spectrophotometer is, especially
if the unit takes only a single reading per patch.
