Re: Real world experience w/ GMG and Oris RIPs
Re: Real world experience w/ GMG and Oris RIPs
- Subject: Re: Real world experience w/ GMG and Oris RIPs
- From: Graeme Gill <email@hidden>
- Date: Wed, 01 Dec 2004 02:02:34 +1100
- Organization: Color Technology Solutions Pty. Ltd.
Andre Schützenhofer wrote:
from my understanding, the profile grid points have been computed
already - they were interpolated form say an ECI2002 chart. If we send
LAB ECI2002-testchart values or others through the profile, those
initial "grid points" (specific test chart values) most likely do not
exist anymore and have to be interpolated form an interpolated profile
grid. Therefore, tweaking grid points seems to make sense to me.
Yes, depending on the type of interpolation algorithm used, the original
test points may not be perfectly reproduced via the forward profile (the
A2B table). This is not necessarily an error though. The test points
have some level of random errors in them, and filtering the A2B grid
values may result in a higher error to the particular test points,
but a better fit to the actual device response.
Putting this consideration aside however, there is still absolutely
no need for actual iteration (printing more test charts and measuring
them). The interpolation algorithm can use whatever process it's creator
would like to invent, to make the forward interpolation better match
the test patches. Anything like that would be invisible to the user.
[Some of the work done by Done Bone (D.J. Bone, "Adaptive Colour Printer
Modelling using regularized linear splines," Proc. SPIE Vol. 1909,
p. 104-115, Device-Independent Color Imaging and Imaging Systems Integration,
Ricardo J. Motta; Hapet A. Berberian; Eds.(1993))
used an approach that helped minimize the number of samples that needed
reading with a "one at a time" manual instrument. It was an adaptive
approach which started by printing and reading a smallish number of samples,
and then by estimating the variability of the device response between those
samples, chose more points to print and measure. Some of the stuff I've been
working on has a similar flavour, but instead of an interactive print/measurment
approach, I've been simply using an existing device profile to create an
efficient new set of test samples.]
Graeme Gill.
_______________________________________________
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