Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

[Physltest] [Phys-L] Re: primary colors



Frohne, Vickie wrote:
Technically, there are several sets of possible "primary
colors." A set of primary colors can be loosely defined as
three colors, that when mixed, produce all the other colors in
the rainbow.

Yes, loosely ... very loosely.

1) It depends on what you mean by "mixing". The physics
of mixing colored inks is radically different from the
physics of mixing colored lights. I was tempted to say
"diametrically" different ... but it's even worse than that.

Cyan, Magenta, Yellow is a primary set,

Those are subtractive primaries, for mixing inks.

and so
is RGB, which is the complment of CMY.

Those are additive primaries, for mixing lights.

Presumably one could
define orange, green, violet (the compliments of RYB) as a
"primary" set.

Not without significant sacrifices. For starters, using
orange, green, and violet lights, how are you going to
make red? How are you going to make blue?

We can agree that there is *some* discretion in the choice
of primaries ... but it's not a free-for all.

2) Also I find it alarmingly loose to suggest that any
three primaries can cover "all the other colors in
the rainbow". In either system (additive or subtractive)
there are going to be many colors of the rainbow that
are left out, i.e. not well approximated by mixing the
three primaries (under mild assumptions such as non-
negative amounts of each primary).

There is no compelling physical reason to
favor any of these options over another.

Au contraire, there are compelling reasons for using
RGB for lights and CMY for inks ... or something not
much different.

I have no problems with teachers calling RYB "primary colors",
since this is the common designation of primary colors in art
classes.

Somewhat common, but not universal. It's a moving
target. It's changing toward more use of the CMY names.
This is a change in the good direction IMHO. I think
the change is inevitable as more and more art teachers
and k-6 teachers get ink-jet printers and have to buy
ink cartridges.

. Some color-mixers give the
options of working with an RGB system, a CMYK system, a color
wheel system, or a hue-saturation-brightness (HSB) system.

Interpolation in RGB space is a decent model of the physics
of mixing colored lights. Interpolation in CMYK space is
at best a very rough model for the mixing of real inks. The
fact that the black locus in this figure
http://www.av8n.com/imaging/color-management.htm#fig-cie2w
is not triangular should serve as a word to the wise. It
was drawn using real data, not some wishful-thinking theory.

In general the mixing of inks is a nightmare. There are a
lot more than three relevant variables. Any three-variable
theory is doomed before it starts. Nowadays it's easy to
buy printers with six inks (C PC M PM Y K) or more.
_______________________________________________
Phys-L mailing list
Phys-L@electron.physics.buffalo.edu
https://www.physics.buffalo.edu/mailman/listinfo/phys-l