Color standards

Beatrix Arendt <barendt [at] monticello__org> writes

>... So we would like to say, "these
>'x' blue chips that we pulled from Munsell's color books are going
>to be the range for 'blue.' And these 'y' chips are the range for
>light blue." That way, when an archaeologist looks at other's
>reports from different sites, they will be able to know exactly what
>range of blues they were talking about. We are hoping it will take
>out some of the subjectivity of color choice. It is a first attempt
>at speaking the same "language."

I'm sure you will get a lot of great answers from the DistList, but
here are my thoughts on the matter. I do not believe you can "take
the subjectivity out" of color viewing. Color viewing is an
inherently subjective event. The solution, instead, is to "control
for the subjectivity". This is because there is a problem here that
goes deeper than the imprecise or inconsistent language that is used
to describe color. The problem lies in the very nature of the
color-viewing event.

Even if you were to choose a standard color-chip set as a reference,
you would run into two problems: One is the "two-socks problem",
named for the situation when you pull two socks from your drawer and
think they are the same color, only to see in another light that
they are definitely not. The other is trying to control for the fact
that a viewer working at a different site (even the same
archaeologist) is liable to see colors differently than in the
reference conditions. These are actually two aspects of the same
basic problem; our color sensing ability is great at distinguishing
between colors, but not as good at absolutely identifying colors.
This is, in part, because it always works in a context. You can
point to a particular chip from Munsell as a reference, but the
experience of that color is going to be affected by things like the
nature of the light in the viewing area and the surrounding colors
(sometimes referred to as "background noise", but in fact they are a
very important part of how we identify color--that is, by
differentiating it from the other visible colors which surround it).

You must try to control for the nature of the light in effect during
the color-viewing event. Your reference would then become "this chip
from Munsell viewed under the standard conditions". Typically you
would reproduce the "standard conditions" in a special light box.
Unfortunately to properly match you need to also see the artifact
under the same conditions and there will be times when you cannot
physically stuff it into the light box (or, for that matter, it
might be impractical to bring the light box to the artifact under
observation). In these cases a handheld colorimeter might not be a
bad solution. A scholar would bring it along and use it to compare
the Munsell chip to the artifact, or simply record the L*a*b* values
and do the comparison later (depends on the features of the
colorimeter).

Best of luck, it sounds like an interesting project.

Bob Savage
Media Preservation Unit
Stanford University Libraries


                                  ***
                  Conservation DistList Instance 14:64
                  Distributed: Thursday, May 24, 2001
                       Message Id: cdl-14-64-003
                                  ***