Singin' the blues

[Bob Buddemeier]

I can't stand it any longer.

The perceived 'color' of an object is the response of a sensor to the spectrum
of the light reflected by the object.  The strength and quality (wavelength
mix) of the signal reaching the sensor is a function of:
1.  The spectrum of the source of incident light (e.g., sun, strobe, flashbulb);
2.  The pathlength and optical characteristics of the medium that transmits
the incident light to the object;
3.  The pathlength and optical characteristics of the medium that transmits
the reflected light to the sensor.
4.  The spectral reflectance characteristics of the object; and
5.  Background or scattering contributions to the incident light signal (along
both paths).

But we're not through.

All sensors are not equal.  E.g., human eyes -- ever have an argument with
your significant other about what tie/shirt/suit looked good together? But
especially, film.  Most commercial films are tuned to generate a consensus
perception that is not too far from what people think they see in air, in
sunlight.  But the recipes for doing that vary, and so you may get quite
different responses-- between films as well as between environments -- in
other environments (e.g., flash, or indoor lighting, or low-light, or
underwater).  Special-purpose films (e.g., ultra high speed) may have special
responses. And, the colors in developed film (especially prints) may change
with time.  Further, now that digital processing and images are coming in,
computer adjustment of hues makes it impossible to know how the image relates
to what you were looking at in the first place(or at least, to convince anyone
else of some objective reality).

One point is -- of course the bloody things look different in and out of
water, at different depths, in pictures and in life, etc. etc.

Another point is -- verbal transmissions, and even comparison of images of
unspecified origins, can't possibly resolve most of the things people are
talking about.  You need (sorry) a quantitative, experimentally repeatable
definition of "blue," or at least one which controls for the largest of the
major variables listed above.  (Photosynthesis types at least specify PAR --
trying to figure out what a blue coral means is tougher, since we haven't
really specified the question as far as I can tell.)

I have no objection to people using the electronic medium to trade just-so
stories, but please designate a satellite list or some node so that they don't
wind up in my mailbox until the collection has been edited for public consumption.

2.  My own just-so story is that my recollection from a decade (the 70s) in
the water at various places around the Pacific is that Porites comes in tan,
brown, green, olive-drab, and a range of blues ranging from rather light to
near-purple.  Take it, and do what you want.

3.  One of keenest memories of learning about corals is that once when I
started waving my arms about ecology and evolution and other things of which I
knew little, Bob Kinzie told me "you have to remember that basically color in
corals doesn't mean much of anything."  I do not suggest that either of us
Bobs would defend that as an absolutely true statement applicable to all
issues, but I suspect it's still a pretty good first approximation.

'Nuff?

Bob Buddemeier, self-appointed curmudgeon
--


Dr. Robert W. Buddemeier
Kansas Geological Survey
University of Kansas
1930 Constant Ave.
Lawrence, KS 66047

(913) 864-3965 w
(913) 864-5317 fax
buddrw at kgs.ukans.edu