About press sampling
About press sampling
- Subject: About press sampling
- From: Roger Breton <email@hidden>
- Date: Tue, 08 Mar 2005 08:05:03 -0500
I'm working on profiling a press that exhibits an abnormally large
variability -- don't ask me what the printing process is, there was no room
to put any color bars. Despite this seemingly large variability, I believe
this print process can still be profiled.
Usually, when building press profiles, I take a random sample of press
sheets out of the press run and build a profile that way. Since the sheets
usually cluster within a narrow range of density, this approach works well
for me. But this time, since the variation is so large, common sense is
telling me I should proceed differently.
My thoughts on this, first, was to study the variability of the
'population', something that, as you'll see, current profilers are not well
adept at. I was planning on bringing a set of 25 ECI2002 measurements in
ProfileMakerPro MeasureTool, and simply average them. I was counting on the
fact that MeasureTool would supply me with ample detailed statistics on the
sample, such as mean DeltaE and so on. That way, I thought, armed with the
mean and standard deviation, my job would have reduced to discard the sheets
that deviated too much from the mean and use what's left to build my
profile. Alas, it does not work that way. To date, Measure Tool or any other
measuring package I know of have this capability :(
Stubborn, I turned to my favorite color management utility, MS Exel, and
proceeded to import a few sets of measurements on my own. I was about to
calculate separate mean L*, mean a* and mean b* but it then occurred to me
that the results of this would be meaningless. Why? Because Lab triplets
representing 3D coordinates. So, it's the average of the 3D coordinates I
want -- not the individual L, a, b averages. Which inevitably means turning
my data into matrices. Anyone ever done that and could offer some simple
guidance? Unless I am going about this the wrong way.
In the meantime, I'm back with MeasureTool and PrintOpen, analyzing two sets
of data at a time.
Regards,
Roger Breton | Laval, Canada | email@hidden
http://pages.infinit.net/graxx
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden