[Tig] color perception of lights

Dominic Case dominic_case
Fri Oct 19 00:13:29 BST 2001


I sent this before but failed to address it ti tig (d'oh!), so it only went
to Rob. Also forgot to sign it. Brainfade?
Dominic Case
============
>streetlights in
> the U.S. will be quite blue or quite amber 

I don't think Martin's explanation accounts for the the exaggeration of
perveived colours that Rob describes. The graph of atmospheric absorbtions
shows that most visible spectrum is transmitted quite well - as we know, UV
is absorbed, and it's also true that the blue wavelengths are scattered,
accounting for the yellow colour of the sun and the blue of the sky when we
look upwards from the ground. But the same atmosphere would make _all_
terrestrial light sources look more orange if that were the main effect.

I think Rob's own answer is the better one: we tend to take ambient light as
reference: a white sheet of paper has a different spectrum depending on what
it's illuminated by, but we "know" it's white, so the brain autocalibrates
its responses to other colours on the basis of the observed spectrum from
the paper. The light source that illuminates the white paper is also
illuminating everything else we see. 

Not so from an aircraft, the light sources - streeet lights etc - are just
points of light, to be compared with other point sources of different
illuminants. So we are better able to notice the difference.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dominic Case
Group Technology & Services Manager
The Atlab Group
87 Mars Road, Lane Cove 2066
ph +61(0)2 9906 0100
fax +61(0)2 9906 7048
email: cased at atlab.com.au
web: http://www.atlab.com.au
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~






More information about the Tig mailing list