Re: Color constancy and metamerism
Re: Color constancy and metamerism
- Subject: Re: Color constancy and metamerism
- From: Marco Ugolini <email@hidden>
- Date: Wed, 21 Sep 2005 22:07:39 -0700
In a message dated Wed, 21 Sep 2005 05:10:31, Dennis W. Manasco wrote:
>> I have achieved near perfect neutral appearing prints under 5000K
>> lighting, but under cool white deluxe, the prints appear slightly
>> sepia. That is metamerism.
>> ...
>
> Actually that is metameric failure.
I don't wish to be butting heads, but actually it's illuminant metamerism
failure only if you are viewing two identical copies of a print (same paper,
printer and inks, cut from a tiled print on a larger sheet) side by side in
a darkened room, and each of them is under a different illuminant.
In this case you would have print A under a D50 illuminant and print B under
illuminant F2 (cool white fluorescent, 4230K CCT): the first one would look
neutral, the second would show a sepia cast. That is an instance of
illuminant metamerism failure.
Again, you always have to have two concurrent instances of something to
establish a metameric match or failure (two samples and one observer in
sample metamerism; two illuminants and one observer in illuminant
metamerism; two observers and one illuminant in observer metamerism, etc.).
But simply taking one print from one room under illuminant D50 to one under
illuminant F2 and noticing that now it looks sepia is not metamerism: it's
an instance of color inconstancy.
> Color constancy is something else entirely: It's basically your mind
> fooling you about colors.
Mark Fairchild says that, strictly speaking, color constancy does not exist
in humans ("Color Appearance Models," 2nd ed., pp. 132-133). Basically,
color constancy CAN be seen as a "trick," in much the same way as chromatic
adaptation.
But he also goes on to quote R.M. Evans (1943): "...in everyday life we are
accustomed to thinking of most colors as not changing at all. This is due to
the tendency to remember colors rather than to look at them closely."
Obviously "color constancy" is a construct of the mind: we expect colors to
look a certain way under different illuminants, though they may differ if we
"look at them closely." Still, the whole of color perception itself is also
a product of the mind. In the end, then, what does it mean to say that
"color constancy" is a way in which the mind fools you? It doesn't seem to
me that it adds much to our knowledge just to say that. Whole philosophical
edifices are built upon the ultimately indeterminable nature of perception
and of the human mind, and so forth ad infinitum...
To come back to Earth, if a sample's color seems to shift enough from one
illuminant to another to be PERCEIVED AS DIFFERENT FROM ONE'S EXPECTATION OF
ITS APPEARANCE, that color is NOT perceived as constant by the mind, even
after it discounts the illuminant and performs a chromatic adaptation to the
new illuminant. If a grayscale print that looked neutral under illuminant A
now looks sepia under illuminant F2, the mind perceives a lack of color
constancy in that print (in this case, a lack of the expected neutrality),
even without "looking at it closely."
Clearly, perceived inconstancy also depends on the observer's level of
training in recognizing varying degrees of inconstancy: Jim may see the
print as neutral under either illuminant, but Joe may see it as slightly
sepia under F2, enough to bother him, because his EXPECTATIONS of constancy
have been heightened by his visual training.
In closing, if "color inconstancy" is an inappropriate way to describe this
effect, that still does not make "metamerism" the correct term purely by
default: a third term would then need to be introduced (neither "metamerism"
nor "color constancy").
Please offer suggestions. Thank you.
Sorry for the long-winded tirade...pheew!
--------------
Marco Ugolini
Mill Valley, CA
_______________________________________________
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