[argyllcms] Re: Determining paper color from a profile?
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Wed, 03 Feb 2010 13:54:20 +1100
Ben Goren wrote:
Is it safe to assume that the wtpt tag of an ICC profile is the actual paper
color?
Not really. The white point can be anything the profile creator decides
to put in there.
An Argyll-made profile of Red River Premium Matte paper, on the other hand, has
a wtpt
tag of X=0.859, Y=0.896, Z=0.849. According to Bruce, that's L=95.8, a=-0.920,
b=-9.09.
According to spotread from a sheet from a different batch from what I used to
make the
profile, the paper measures X=86, Y=89, Z=78 or L=95, a=1.6, b=-4.2. Not
exactly the
same, but certainly in the same ballpark. And it looks about right when I fill
the
screen with those values in Photoshop.
The Argyll printer profile white point is typically close to, but not quite the
same
as the white test patches. This is because the test points are modeled with some
degree of smoothing. The white point that ends up in the profile is taken
from the smoothed model value of no colorant (subtractive) or full colorant
(additive). You can make them closer by adding more white test patches
to the test chart, which will weight that more in the smoothed model.
So...is the Crane Museo Max paper really that pinkish? Or is their profile
somehow
b0rked? Or...?
Any of the above. The only sure way is to measure the paper.
Graeme Gill.
Other related posts:
