From: David Miller <dm2363@...>
To: datacolor_group@yahoogroups.com
Sent: Mon, August 16, 2010 1:06:48 PM
Subject: Re: [datacolor_group] Problems building profiles for an EPSON 4000 and
Hahnemuhle Papers
On Aug 15, 2010, at 10:09 PM, brucepanock@... wrote:
> I am the person referred to in a response to "keith."
>
> My principle problem is scanning the targets. I am following the instructions
>in detail (I even stop the oline video as I move from step to step). The problem
>is that cell 10(j) always scans black rather than what appears to be white. I am
>attempting to build profiles for Hahnemuhle Photo Rag Bright WHite 310 gsm, Fine
>Art Pearl 285 gsm, Gaguerre canvas 400 gsm, as well as creating better profiles
>for Epson Matte and canvas.
>
Yes, I see that you have a support ticket open on that, and someone from
tech support should be handling this with you now. I've looked at your ticket,
and I'll summarize a few things here, but please continue with this in the
support
ticket itself...:-)
Your paper white patch measuring as black, instead of white, can only be from
one thing:
you need to calibrate the spectro properly on the white tile on the calibration
base.
With the SR (scanning) spectro: "proper" calibration means that you have to push
down firmly
on the spectro nose when you have it in the base, before you press the button on
the spectro
itself. If you don't "press-before-you-push-the-button" during spectro
calibration on the tile,
you may not get firm enough contact with the tile for the calibration to work
correctly.
What happens when this goes wrong? Specifically: whites, and near whites, can
measure incorrectly
as either black, or with a very strong cyan cast.
> Should I be using the EZ targets or the Classic Targets? I used the Classic,
>High Quality Target.
>
You can use either, but this won't affect your white patch measurement.
> I use the plastic white in Spectro cradle to determine white point.
>
> If I cannot get past the scanning I am stuck .. please help.
>
I'm sorry that tech support hasn't responded directly and correctly on this one
yet; it's a question
that comes up occasionally (not many people run into this, because the
instructions in the software
say to press down on the nose, but it's possible to miss this) and the answer
for it is very simple:
just press down on the nose, during calibration, before you press the button.
It should be simple for you to test this. Run the software, go back to your
measurement file, click
View/Measure, recalibrate the spectro (see all of the above); then click on the
white patch in the
Target window. Switch the measuring mode for a single measurement (use the popup
in the lower right
corner of the window). Put the spectro on the white patch in the target and
press the button. You should
now get a proper measurement for white (should come up as a very light gray -
not black, and not a strong
cyan cast).
Your other problem is your target prints. Assuming that the rest of your target
actually looks like the
measurements, they're printing as (fairly horribly) color managed. Look at the
first group of 5x5
patches in the upper left corner of the target - those are the full black to
full blue to full green to
full cyan patches. Patch 5A (full blue) should be a very strong blue; it's the
most fully inked blue that
the driver can print (and even if it's biased towards magenta, as may be the
case with the older 4000 driver,
it should still be a very strongly inked patch; this is the most saturated blue
possible with the driver
settings you're using to print the target). Same comment about the green patch
(5th down in the left column)
and the cyan patch (5th down and across). All of the patches in between those
should also be very dark and
saturated.
In your target print, they're not. That either means:
- The driver settings are wrong
- You've got a big problem with your inks (nozzle clogs, most likely, or there's
something wrong with
what's in your ink carts). Best thing to do here is a very, very careful nozzle
check. If your inks
are all working correctly, and nothing is missing, then you either have bad
inks, or (implicitly bad)
3rd party inks, or (see above) your driver settings while making the target
print are wrong.
So you have 2 problems:
- Measurements won't let you build a proper profile because the spectro is
calibrated incorrectly. That's
the first issue to address.
- Even with this addressed, the target print is wrong and you won't be able to
build a proper profile from
it, even if every patch is measured correctly.
David Miller
Senior Software Developer, Digital Color Solutions
Datacolor