|
Server : Apache/2.4.62 System : FreeBSD fbsdweb2.web.rcn.net 14.1-RELEASE FreeBSD 14.1-RELEASE releng/14.1-n267679-10e31f0946d8 GENERIC amd64 User : www ( 80) PHP Version : 8.3.8 Disable Function : NONE Directory : /domains/nostrada/ |
Upload File : |
<html>
<head>
<meta http-equiv="Content-Language" content="en-us">
<meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
<title>Nostradamus and the wonderful world of color</title>
<link href="color_theory.css" rel="stylesheet" type="text/css">
<script src="https://use.typekit.net/mfy1nby.js"></script>
<script>try{Typekit.load({ async: true });}catch(e){}</script>
</head>
<body>
<div class="head-box"><h1>Nostradamus and the wonderful world of color</h1></div>
<div class="body-box">
<p>I wanted to learn more about color theory — things like primary colors, split primaries, complementaries, color harmony, and the like. But what I read either seemed arbitrary or involved 'way too much calculus for my taste. So I decided to investigate color theory in my own way, beginnning with its scientific basis and how we perceive it.</p>
<h2>What is color?</h2>
<p>Color is what we perceive when our retinas are exposed to visible light.</p>
<h2>So what's "visible light"?</h2>
<img src="color/fotolia-comp-74283953.jpg" alt="the visible light spectrum" width="500" height="468" class="float-left img-border" ></p><p>Visible light is the portion of the electromagnetic spectrum that's . . . visbile.
Which usually means visible to humans.
<p>For the average human, that's radiation whose wavelength is between about 380 and 750 nanometers (for reference, a nanometer is about 2.18723 x 10<sup>-9</sup> cubits).
<p>Of course real humans are not average and some see more or less than the 380-750nm band.
<p><em>Color blindness</em> is a different phenomenon. People who are color blind can see pretty much the same light as everyone else, they just cannot differentiate between color <em>A</em> and color <em>B</em>
<p>On the other hand, some bees, butterflies, and birds can see well up into the ultraviolet range, whereas snakes can detect infrared light, but not through their eyes.
<h3>The spectrum -- ROY G. BIV</h3>
<p>We start with the color spectrum -- the familiar color series produced by a
prism. These are the same colors that are in the rainbow, and form a small
portion of he electromagnetic spectrum.</p>
<p>The spectrum can be represented by a line. Other color models include the
wheel and the cube. </p>
<ul>
<li>Why are these colors components of <i>white</i> light?</li>
<li>Why is brown not in the spectrum?</li>
<li>Why can you make white light using just RGB?</li>
<li>What happened to the "I" in Roy's last name?</li>
</ul>
<h3>Red, green and blue -- RGB -- the light primaries</h3>
<p>You can make all the visible colors using just red, green and blue
light. </p>
<p>My understanding of RGB will work but the actual processing is more complex
([physiology:] current theory states that color perception involves both
trichromatic and opponent color processing).</p>
<p>SML cones are roughly BGR, though current science says the L cones are most
sensitive to Y -- which will come as a surprise to all our RGB TV sets and
monitors.</p>
<p>Magenta (a non spectral color) is a combination of red and blue</p>
<p>Metamers</p>
<p>It's my opinion that more effort has been put into studying color perception
than color emission -- probably because you can't charge for light but you can
charge for Peach Passion lipstick.</p>
<p>Everybody uses "colorimetric assays" as a lab technique for
deteremining what's in something.</p>
<p>freq vs wavelength</p>
<h3>Blue, red and yellow -- the painter's primaries</h3>
<p>Most color theory written for artists is based on the BRY primaries. Color
wheels, too.</p>
<p>One definition of primaries is that:</p>
<ol>
<li> a primary color cannot be produced by mixing two colors
together</li>
<li> The primary colors can be combined to produce all colors</li>
</ol>
<p>BRY fails the test. The true pigment primaries are CMY and schools are
beginning to teach them that way.</p>
<ul>
<li>Why BRY? Because painters are conservative</li>
<li>Because painters are philosophical</li>
<li>because pigments were limited</li>
</ul>
<p>The Pantone mixing system uses 11 colors and still has gamut limitations.
(Gives good gamut)</p>
<h3>CMYK</h3>
<p>CMYK -- why is the K needed? In RGB light, if your green is a bit off center,
you can just make it a bit brighter. But the CMY pigments <i>absorb</i> light;
if the wavelengths each absorbs isn't the exact complement of the eye's sensitivity
curve, unwanted light will be reflected. As a practical matter, those ideal pigments
either don't exist or are too expensive for commercial use. The additional
light absorption is accomplished by adding some black pigment.</p>
<h2>What's the best color model?</h2>
<p>You might as well start with RGB because it's a reasonable approximation of
how the eye works. It's used by computer monitors and TV sets and is the usual
way to specify colors for the Web. For those who grew up mixing paints
rather than colors of light, RGB is not intuitive. </p>
<p>To choose or mix colors more like an artist might, you can use HSV. You pick
a color ("hue") then add varying amounts of white and/or black. And
HSV converts neatly to RGB. (See <a href="http://www.buena.com/articles/hsvspace.pdf">http://www.buena.com/articles/hsvspace.pdf</a>
though note that the author describes HSV as a cone. I'm pretty sure that HSV is
"really" a cylinder but that so many colors ***see other cites*** </p>
<p>If you're printing in full color, you need to use CMYK, the industry
standard. For consistent color proofing, the American process color proofing
standard is SWOP (outside the US, there are different proofing standards). </p>
<p>For a broader gamut than is possible with CMYK, Pantone
has a proprietary 6-color system they call Hexachrome. </p>
<p>For spot color, the
Pantone Matching System is the standard for specifying color inks (again, there
are other standards for spot color outside the US). </p>
<p>[disclaimer] This essay is designed to help you -- and me -- think more
clearly about color. But you should understand that I am not a color scientist
and have not doubt made errors. In other words: If you rely on this for your
thesis, you're screwed.</p>
<p> </p>
<p>additive and subtractive:</p>
<p>Light is usually described as "additive" because you can add the
primary colors of light and get white. Paint is usually called
"subtractive" because you can add the primary colors and get . . .
well, it should be black. But you could put a blue and a green gel together in
front of a light source and be pretty close to black. If not, try adding a red
gel. In other words, you can subtract colors of light. And you can paint a disk
red, green and blue. When spun quickly, it should look white, proving that
pigments can be additive. If it looks gray instead, just chalk it up to
experimental error. </p>
<p>James Clerk Maxwell described spectral color (light) and
perceived color (eyeballs) around 1860. Also invented color separation.</p>
<p><img border="0" src="images/color/Spectrum4websiteEval.png" width="600" height="72"></p>
<p>Color spectrum simulation digitally created by Deborah S Krolls (http://www.nationmaster.com/encyclopedia/Image:Spectrum4websiteEval.png).</p>
<p>note: Cecil on color terms</p>
<p>Most real colors are non-spectral just like most real objects don't have
integer dimensions.</p>
<p><img border="0" src="images/color/Cone-response.png" width="550" height="346"></p>
<p>Spectral absorption curves of the short (S), medium (M) and long (L)
wavelength pigments in human cone and rod (R) cells. After Bowmaker J.K. and
Dartnall H.J.A., "Visual pigments of rods and cones in a human
retina." <i>J. Physiol.</i> <b>298</b>: pp501-511 (1980). (http://en.wikipedia.org/wiki/Image:Cone-response.png)</p>
<p> </p>
<hr>
<p>9-26-2013 Tony Zak asked me whether one could match the color of his trousers to a PMS chip.</p>
<p>My answer:</p>
<p>No.</p>
<ol>
<li>His trousers are made up of many colors.</li>
<li>There are only 1114 Pantone solid colors. If his trousers were only a single color, it may not be one of the PMS colors.</li>
<li>Printing on paper won't have the same reflectance as Tony's trousers.</li>
</ol>
<p>That said, you could try matching the trousers with CMYK colors. CMYK can easily produce millions of different colors. However</p>
<ol>
<li>There are still many visible colors outside of the CMYK gamut (i.e., that CMYK pigments cannot reproduce). You could use a 6-color process to expand the gamut, but Hexachrome has been discontinued and CcMmYK is not widely available.</li>
<li>You'll still have the problem of "illuminant metameric failure" -- meaning that you have a perfectly good color match to Tony's trousers under one particular light source, but when you change to different illumination, it no longer matches.</li>
</ol>
<p>
<st1:date Month="8" Day="22" Year="2004">8-22-04</st1:date>
Color</p>
<p >See: <a href="http://www.siggraph.org/education/materials/HyperGraph/color/coloreff.htm">http://www.siggraph.org/education/materials/HyperGraph/color/coloreff.htm</a></p>
<p><a href="http://davis.wpi.edu/~matt/courses/color/">http://davis.wpi.edu/~matt/courses/color/</a></p>
<p >color wheel vs. the spectrum line</p>
<p >Spectrophotometry</p>
<p>This page started with my own attempts to better understand color theory, and
my desire to clarify for others the parts of conventional color theory that I
think are wrong.</p>
<p>color blindness and cats</p>
<p>SWOP</p>
<p>5500 Kelvin = (In "customary measure," that would be 9900 Rankine)</p>
</div>
</body>
</html>