RE: calculating from abscol to relcol from L*a*b*
RE: calculating from abscol to relcol from L*a*b*
- Subject: RE: calculating from abscol to relcol from L*a*b*
- From: "tl" <email@hidden>
- Date: Wed, 22 Apr 2009 07:38:58 -0400
Hi to terry and all other folks!
As marco so correctly said:
" In other words, one must convert from the measured Lab values to XYZ, then
from XYZ back to Lab -- but this time using the XYZ coordinates of the
measured reflective media white for the transform from XYZ to Lab, instead
of D50 white point coordinates."
This is correct algebra, but incorrect physics. If the paper white is the
result of florescence, then each of the printed colors is affected by the
colors specific transmission of UV. In other words, the color of the paper
is physically changing under the Ink, so most of the math falls apart in a
practical sense. The problem with the ICC conversion is that is both
cumbersome and physically incorrect as is the basic chromatic adaptation
used in CIELAB. If you had spectral data, you would measure the media white
and divide that spectra into all the other spectra. If you look at the
math, that is exact process you are doing with the Media White XYZ
coordinates in the LAB calculation Note that this still will not work in the
presence of fluorescence with or without spectral data. For operations like
this, you need to measure with UV cut, but keep in mind that this is will
not relate to what you physically see which, in theory, is the goal of using
Lab.
One portion of your last message confused me. You wrote:
" I have a method currently that "works" but I'm getting deltas that
average in the .30-.50 range with maximum deltas in the 2.0-3.0 range
compared to a true relative ICC conversion. "
How did you determine the Delta E values?. I would be very satisfied with
those.
Regards,
Tom
_______________________________________________
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