Re: Linearizing a press (was Re: Max K generation question)
Re: Linearizing a press (was Re: Max K generation question)
- Subject: Re: Linearizing a press (was Re: Max K generation question)
- From: Marc Levine <email@hidden>
- Date: Tue, 28 Sep 2004 13:04:23 -0400
There's a lot of good stuff in here.
>
You are now getting into where I'm at. I'm not 100% convinced that this
>
additional or preliminary press run should take the form of a colorimetric
>
linearization target. Although, I don't see why not.
Ok.......
>
But, as you pointed
>
out, it's just that the whole prepress world has never been in the mindset
>
of questionning their press tonal behavior like we have for years when
>
calibrating and linearizing an inkjet printer for proofing separations. Two
>
points come to mind. 1st, if I have a superlative linearization routine
>
inside my printer software à la BestColor or BlackMagic v2.6+ then I don't
>
believe running Monaco's colorimetric linearization (or anybody else's) on
>
top of BestColor will do anything for me.
Correct, If the RIP linearizes colorimetrically, an additional linearization
will provide no additional benefit. It is more useful when used in
conjunction with density-based linearizations. (or in applications where
there is no linearization mechanism).
>
2nd I tend to think that running a
>
colorimetric linearization on a press is not the end be all? Maybe I'm wrong
>
and maybe it's a simple place to start. But that's what I'm after. I want
>
what you describe below:
>
>
> a homogenized color
>
> space that minimizes profile interpolation between color samples and enables
>
> the profiling engine to deliver maximum detail throughout the press's
>
> dynamic range.
Think of your ICC target not as a series of recipes that you print and
measure, but as a series of color samples from which you will try and
estimate a device's color behavior. The device's printing condition is what
dictates the spacing of these color samples and how evenly they are spread
throughout the device's colorspace. A uniform distribution of color samples
will equalize the color distance between target samples and will reduce the
amount of interpolation that the profiling engine needs to do when modeling
the remaining device colors.
Think about the Hutch scanner target vs. an IT8 target. The reason it works
is that (in addition to having a very accurate measurement file) it provides
a more uniform color sampling of a scanner's colorspace and produces a
profile with increased detail in both highlight and shadow areas. The same
logic applies to prints - although the issue is not as dramatic when you
consider that a press colorspace is much smaller than a scanner and you are
using many more patches to meter it.
>
>
With today's CTPs, I suspect it's possible and perhaps more desirable to
>
carry this preliminary linearization steps with traditional means, using
>
plain ole densitometry instead of colorimetry: we have to do it anyway! So
>
why not make an initial press run with linear plates, first, and study the
>
results.
It is common to run a raw plate first and then build a linear correction
that makes the plate print "linear". It is also reasonable to then run the
"corrected" plate on press to identify the press behavior and then use that
info back at the plating stage. It is important to clarify corrected vs.
uncorrected plates. What you are correcting to can be linearity in terms of
density, color, or can be you own special printing condition. The point is
that that linearity just represents a single condition based on a certain
metric that everybody kinda "gets". The important this to stay focused on is
that you are after a certain "corrected" behavior on both your plate and
your press, and that you will probably will need to make at least 2 sets of
plates and perform 1 press run (2 if you want to build your ICC targets) in
the process.
>
I would trow in my favorite PrintOpen characterization target in
>
the mix, and get a feel for the press tonal behavior. PO will analyze the
>
data and show me the curves it will use to build neutrals. At that point,
>
it's plain to see any crossovers and the spatial distribution of patches,
>
how well spaced they are in Lab or even xyY.
A profile will develop it's own curves based on the behavior of the device,
the application settings you select, and the algorithms that the software
employs to build ICC profiles. The resulting gray construction curves may or
may not look like behaviors which have been defined when calibrating the
press. Be aware that the press is the press and the profile is the profile.
>
we never
>
talked about calibrating and linearizing a press the way we always talked
>
about calibating and linearizing an inkjet printer. Why? Because we always
>
took presses for granted. Not anymore here.
Calibrating the press has always been an issue, but it has not been one that
prepress has been comfortable with historically. As workflow solidifies and
interim stages (such as film) disappear, it is clearly becoming more
important for prepress and the pressroom to cooperate on their activities
and not take anything for granted. Understanding of all the variables (or
families of variables), how they can be directed to meet certain criteria,
and implementing structures to control these variables will ultimately
determine how effectively a business can meet their color quality and
production objectives. Optimizing a press condition to better color tonality
can definitely have an impact on your quality ceiling. However, selecting a
target printer behavior that is "pressroom friendly" will likely yield more
consistent results.
Marc
--
Marc Levine
Sales Guy
Technical Guy
X-Rite Incorporated
Email email@hidden
www.xrite.com
_______________________________________________
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