Re: Nimrina colors updated

On Sun, 3 Sep 2006 15:58:02 -0500, Herman Miller <hmiller@...> wrote:

>I haven't made much progress with Nimrina yet -- I sort of got
>sidetracked reorganizing the color chart.
>
>http://www.io.com/~hmiller/png/nimrina-colors.png
>
>Unlike some of the other charts (Tirelat, Lindiga, Minza) I haven't made
>any attempt to line up complementary colors opposite each other.
>Instead, this chart is based on the idea of a red-green axis and a
>yellow-blue axis, with colors spaced as equally as possible around the
>circle.

I don't see why you should have to line up "complementary" colors opposite
each other. That arrangement is based on the most unintuitive part of the
way human vision works (the tristimulus originated at the cones, which does
not reach our brains in that form). It's only useful if you are dealing with
pigment or light mixes, while it's very misleading if you want to describe
how colors relate to each other _perceptually_ (I'd even say the ubiquitous
and almost exclusive use of this kind of arrangement in the teaching of
color theory is detrimental to clear thinking, in that it alienates people
from the true nature of color perceptions as experienced at the brain
level). OTOH, the red-green and yellow-blue axes deal with the kind of color
data that our visual cortex directly works with (the elementary color
percepts), and arranging them opposite each other shows the actual way our
brain organizes color perception.


>So every day I open the color chart and want to move the colors
>around again. I tried coming up with some more quantifiable way to pick
>these colors, but the results ended up worse.

What do you mean by "quantifiable way" to pick the colors?

You could define colors in terms of proportions between its percept
components. Make a chart with combinations for easy proportions; for
example, those that can be expressed on a pie chart divided in twelve
sectors (i.e. 30 degrees each, like those between the hours on a clock
face), which include those based on halves, thirds and quarters, like 100%,
75%/25%, 66.6%/33.3%, 50%/50%, 50%/25%/25%, 33.3%/33.3%/33.3% and
25%/25%/25%/25%. Don't forget some other less elementary proportion, so as
to include colors with only a tinge of some percept; for example, one
twelfth (91.6%) of red diluted into eleven twelfths (8.3%) of white, which
would give a shade of pastel pink (as a side note, it's far easier to notate
thirds, sixths and twelfths as percentages if you use duodecimal). I
recommend you order the chart both by proportions and by type of color:


Saturated colors: entirely chromatic (the hue wheel) or entirely achromatic
(the achromatic scale)

1 percept (100%)

- white/black (pure achromatic)
- yellow/blue/red/green (unary hue)

2 percepts (75%/25%, 66%/33%, 50%/50%, etc.)

- white + black (gray scale)
- yellow/blue + red/green (binary hue; e.g., orange, purple)

Insaturated colors: achromatic and chromatic parts mixed (the colors filling
the space between the hue wheel and the achromatic scale)

- white/black + yellow/blue/red/green (light/dark unary hue; e.g., maroon, navy)

3 percepts (50%/25%/25%, 33%/33%/33%, etc.)

- white + black + yellow/blue/red/green (grayish unary hue; e.g., grayblue)
- white/black + yellow/blue + red/green (light/dark binary hue; e.g., peach,
teal)

4 percepts (25%/25%/25%/25%, etc.)

white + black + yellow/blue + red/green (grayish binary hue; e.g., taupe)


Then pick up a subset of colors from there, taking into account that some
percepts are more salient than others (especially red, which is why brown,
orange, purple or pink --all of them red-based-- are more frequently
encountered as basic colors than non-reddish colors like turquoise or azure).

Reply