Re: Editing profiles [was New EyeOne ruler and soft case]
Re: Editing profiles [was New EyeOne ruler and soft case]
- Subject: Re: Editing profiles [was New EyeOne ruler and soft case]
- From: Klaus Karcher <email@hidden>
- Date: Mon, 14 Nov 2005 22:31:55 +0100
- Organization: schrift & form GmbH
Graeme Gill wrote:
I've played around with this idea recently, while staying within
the ICC framework (see <http://www.argyllcms.com/doc/refine.html>),
but so far I'm unconvinced that it does much, unless you start
out with fairly inaccurate profiles. If you start out with
reasonable profiles, iterating doesn't get significantly
better - you're still limited by the repeatability of the
instrument and the device.
Thomas Holm/pixl wrote:
Well the name of the game in Europe these days is pretty much proof
verification, IE using the Fogra Mediawedge v2 to verify that your
simulations (proofs) are as good as they can be.
When using standard ICC tools (GMB, Heidelberg and a few others) and
verifying the proofs, I get an average deviation of around 2 and a
maximum deviation of between 4-6 depending on what I simulate. THis is
measured with two different, recently recalibrated, Eye One's which
pretty much agree.
If, on the same printer, using the same paper and same spectro as above
I use GMG, or CGS for that matter, and do iterative measurements, My
average usually drops to below 1 and the max deviation is usualy around
2-2,5 (A notable difference is Epson 4800 where the MAgenta primary
usually ends up at dE 2,7 when simulating Fogra 27 as it is essentially
out of gamut).
[...]
sorry for the late reply, but I think it can still be interesting:
some time ago I tried to figure out how well Heidelbergs iterative
correction works. Therefore I took the 27x27 CMY-filds from the top left
corner of ECI2002-visual and iterpolated the steps between them to the
full 8-bit Range (256 steps per chanel)
you can find this image at
http://digitalproof.info/iteration/reference.tif
http://digitalproof.info/iteration/reference-with-circles.jpg
shows small white circles aroud the ECI2002 Grid points.
Then I convertet the reference from ISO Coated to Lab (abs) and
a) once to a PO5-Profile *before* iteration and
b) once to the same Profile *after* iteration
and calculated the difference between a and b
here ist the result:
http://digitalproof.info/iteration/result.jpg
You can see that some of the differences look like a pearl necklet --
and the pearls are the ECI2002 grid points. Therefore I think that
iteration can be dangerous: even if the measuered results seem to be
much better, the reliability and smoothness of the profile can be worse
after iteration. This finding complies with my visual impression with
the iterated profile.
regards, Klaus
_______________________________________________
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