Re: Shmetamerism
Re: Shmetamerism
- Subject: Re: Shmetamerism
- From: Peter MacLeod <email@hidden>
- Date: Tue, 6 Nov 2001 10:38:42 -0800
At 8:50 PM -0800 11/5/01, Robin Myers wrote:
> If we take the two outside then the 2000P print might match the
> Kodak card in some densities but not in all, there will be a shift in
> graybalance in some areas of the 2000P print.
>
> So is that not metamerism - or perhaps illuminant metamerism??
>
> Or do we need a new word to describe the above problem. If so let's
> coin one?? I'd rather work on ways to get rid of the problem than
> <argue> about how to describe it ;-}.
A term already exists to describe this phenomenon, "color inconstancy". When a
color is perceived to be the same when the illuminant is changed, it is "color
constancy" at work. When this human visual system mechanism fails then it is
"color inconstancy". There is no need to invent terms.
I'm not sure I agree--color inconstancy and illuminant metamerism
are not the same thing.
Illuminant metamerism (as defined, e.g., in [Berns 2000])
describes the situtation where a metameric pair of samples no
longer match (i.e. differ in tristimulus value) under a
change in illuminant. When I say differ in tristimulus value,
I mean compared to each other, not compared to the value under
the first illuminant.
Color (in)constancy refers to the tendency of a single sample
to change perceived color appearance or not under a change in
illuminant.
Think of two cases. First, think of the interior of my car. The
seat coverings are leather, and the dashboard is some sort of
plastic. They look about the same color, but the degree to which they
continue to match under a change in illuminant can be computed
by a metamerism index. If I drive through the tunnel in Yerba
Buena Island, which is between the two spans of the San Francisco
Bay Bridge, I notice that they look like different shades of yellow,
because that tunnel is illuminated by sodium lights, which are nearly
monochromatic. So under some illuminant they don't match, but under
most, they match acceptably well. So that's a case of metamerism.
The "acceptably well" standard is a tolerancing of the metamerism
index.
On the other hand, if I buy some red carpet at Joe's Carpet
Warehouse, which has nasty, spiky fluorescent or mercury lights,
and get it installed at my house and it's now orange, that's a
problem of color inconstancy.
Having a print and a gray card no longer match under different
lighting is an example of illuminant metamerism, not color
inconstancy.
BTW I looked a little more at the "it's a floor wax/it's a dessert
topping" issue in the definition of metamerism.
[Field 1999] has a slightly different take on the definition of
metamerism, which more closely describes what publishing people think
of as metamerism: "If two samples have different spectrophotometric
curves such that they match under one illuminant but not under others,
or that they are perceived as matching by one observer but not
another, they are said to be metameric. The phenomenon is known
as metamerism." Hunt [Hunt 1987] (there are more recent editions
of Hunt, but I don't have one) talks about metameric pairs
in the sense of different spectra that have the same tristimulus
values under a given illuminant, but then talks about (pg 181)
illuminant metamerism and observer metamerism in the sense that
Berns uses, though Hunt doesn't put them in italics and give them
an entry in the index like he does for metamerism.
My point is that what I'd consider some of the primary literature
in the fields of color science and graphic arts color reproduction
is not extremely consistent about when metamerism means a matching
pair under a single illuminant, and when it means a matching pair
that no longer matches after a change in illuminant or observer.
I'd tend to follow Berns in calling the latter cases "illuminant
metamerism" or "observer metamerism" just to make sure I knew what
I was referring to, but given the confusion in the literature,
it's reasonable to think about context when talking about
metamerism, and allow that sometimes saying "metamerism" really
means "illuminant metamerism."
--Peter
References:
[Berns 2000] Berns, Roy, "Billmeyer and Saltzman's Principles of
Color Technology," Wiley-Interscience, 2000
[Field 1999] Field, Gary, "Color and Its Reproduction," GATF Press,
2000
[Hunt 1987] Hunt, R.W.G., "The Reproduction of Colour in Photography,
Printing & Television," 4th ed., Fountain Press, 1987