Re: Color Temperature -- The ICI (CIE) Chart -- The I See It Rules

[Wright Huntley <jwwiii at pacbell_net>]

> Date: Sat, 15 Feb 2003 06:59:43 -0800 (PST)
> From: "S. Hieber" <shieber at yahoo_com>
> Subject: Re: Color Temperature -- The ICI (CIE) Chart -- The I See It Rules
> 
> Wright Huntley referred readers to
> http://www.photo.net/photo/edscott/vis00020.htm to learn
> about color perception.
> 
> Throw in a few fourier transformations and and I think
> everyone will have it licked ;-) 

Not until you have expanded it with some prolate spheroidal wave 
functions. :-)

> 
> This was pretty good as facts go but not very easy to
> understand.    

I hope you ignored the "cube" arguments, as they require a lot of 
background to appreciate. I just wanted folks to see the "white" curve 
with respect to normal color space.

 > Some rules of thumb:

But lets get them right. OK?

> 
> 
> The apparent color of a complex light beam is made up of
> one or more various different color light beams.

Sometimes, but not usually. Most light sources are a smooth continuum, and 
don't have spectral peaks (different, separate colors).

> 
> Different combinations of colors at different intensities
> will yield the same color temp rating but not the same
> apparent overall color.

One can make up any particular apparent color by adding different source 
colors in the right amounts.

The same is *not true of color temperatures.  They are not white and 
should not be described as having a color temperature unless their 
coordinates lie on (or pretty near) that black line in the CIE chart. If 
they do, they always will look exactly the same.

The habit has been to allow tube manufacturers to apply a guess of color 
temperature, even when they make a fairly non-white tube. That causes the 
confusion. Plant and Aquarium bulbs of a magenta hue have no real color 
temperature. They are just not white. Period. Likewise, green Cool White 
tubes are fairly far off the white curve. Our brain compensates, but a 
color photo taken under them looks bizarre (if accurate).

> 
> Some overall colors of complex light beams make things look
> natural, others make them look funny-colored.

Yes. This is complicated by the fact that we have tricky ways to 
compensate, in our brain, for illuminations of different color.

A green shirt looks green at noon and at sunset, even though the 
illumination has gone from bluish white (sun plus sky) to rather 
orangeish. 7000K to 2000K.

We tend to mentally "force" things to look their right color under lots of 
different lamps. An aquarium looks very different in a dark room (where it 
is the main source of light) and in a brightly lit room (where it is not). 
The latter situation is where bad lamps can make tank colors look 
particularly bizarre.

> 
> With darkness, color fades before shapes.  Depth is less
> detectable with distance and with darkness.

With brightness, the same is true. That's what I said -- that the chart is 
a mid-slice through a shape like a sweet potato. Color space collapses at 
*both* ends.

> 
> Next step, reviewing Gestalt research on visual perception
> - -- let only the timid tarry long  :-0 .

I think I did! [I spent some years at a University researching color 
displays and human perception thereof. Guess I got carried away. ;-)]

> 
> 
> Scott H., snowed in again but determined to buy more fish.

Sympathy. It sprinkled on my way to the killifish club meeting last night 
but it is dry now -- chilly 52 degrees outside. :-) Another tough CA winter.

Wright

-- 
Wright Huntley -- 209 521-0557 -- 731 Loletta Ave, Modesto CA 95351

       "Everything is relative. In most of coastal California,
       Ted Kennedy would be politically middle of the road --
            and, in San Francisco, right of center."
                                    -- Thos. Sowell  JWR 2/14/03
                     www.sfbaka.net