Re: Lab is always "just right" p1
Re: Lab is always "just right" p1
- Subject: Re: Lab is always "just right" p1
- From: Steve Upton <email@hidden>
- Date: Tue, 11 Sep 2001 02:05:23 -0700
OK, I've been thinking about this one a little and I think I can
explain things - at least to some extent...
Let me start with the first part of the thread.
Neil's question:
"And do we know that the PCS is only as big as Lab?"
, I believe focuses on my comment:
When I view with the 3-D Grapher a Lab profile gamut
overlaid with other color spaces I work in,
I notice that the Lab space does not completely enclose them.
The reds of Adobe RGB (1998), EktaSpace PS 5, J. Holmes,
my custom scanner RGB profile, my custom ink jet profile.
I do not know the answer. Is the PCS only as big as Lab?
Well, first, the PCS IS Lab. So there are no limitations placed on
Lab by the PCS. The PCS for the profiles of today is D50 Lab. (or
sometimes it's older sibling D50 XYZ) - boy I've been hanging out on
this list with Henrik a little too long. I've started using
personification to describe color space variants. (!)
Now, I never intended for Lab profiles to be graphed in 2D or 3D. It
was a little surprising to me that people were routinely dragging Lab
profiles into the grapher to use as an outline for reference.
Honestly it was not until I was a little more careful in my coding
and I revisited the graphing routines shutting Lab graphing off that
I heard some noise. Henrik was not impressed!
So, after playing with Lab profiles in 3D a while I do see the
benefit in having them around. I am now working on ColorThink 2.0's
ability to graph Lab profiles properly.
That said, in CT 1.1 I have not been able to graph, in 3D at least,
any of the standard spaces and have them extend beyond the confines
on Lab. Perhaps I am doing something different? (or, heaven's
forfend, wrong)
Why does even a modest (reflective, not transparencies) scanner
profile appear to enclose a larger space than Lab?
I don't know.
I don't either but I am willing to accept a bug (mine or others) as
an explanation. I have been relying on the CMM to keep the range of
Lab inside the normally accepted limits. Perhaps some CMM's are
breaking the rules?
Or, in the spirit of recent good advice,
perhaps I should instead ask: what am I assuming here
that perhaps I should not be assuming?
please see part II of this post
--