2007-09-14

An Observation About Perception of Color

The visible range of the electromagnetic spectrum covers the continuum of wavelengths between about 800 and about 350 nm (nanometers). In terms of energy, which is inversely proportional to wavelength, this is a progression from lesser to greater energy. And here there is an interesting disconnect between the way humans perceive color and the physical truth of the matter. School kids learn the mnemonic

R O Y G B I V

that is, the man's name, "Roy G. Biv," giving the colors of the rainbow in order from longest to shortest wavelength, which is the same as saying from lowest to highest energy. The interesting part comes from the observation that our perception of color is really circular rather than linear, more like

R O Y
V G
I
B

which is supposed to represent a circle. Perhaps a real image is called for:


The important thing to note is that we don't perceive the break in energy that comes between RED and VIOLET in the physical world. On the color wheel, our perception of the difference between, say, GREEN and YELLOW, is equivalent to our perception of the difference between RED and VIOLET; in energetic terms, however, green and yellow may only differ by tens of nanometers, whereas RED and VIOLET differ by hundreds. Anywhere on the continuum between ORANGE up to INDIGO, the infinitesimal difference we perceive between any two adjacent colors is truly minimal, in terms of energy, while at some point between RED and VIOLET, on the other hand, the same infinitesimal difference of perception is actually a maximal difference of energy, at least so far as visible frequencies are concerned. Perhaps the simplest way to express it is to say that we perceive the spectrum as a circle when in terms of measurable physical parameters it is a line.

Now the real challenge: Having observed this discrepancy, what we might call a "discontinuity" of perception, can we imagine a way to exploit it in a useful invention?

My first thought: we have two colors between violet and red which are infinitesimally different, which amounts to imperceptibly different, to the human eye, yet are dramatically different in terms of energy, which electronic devices should be able to detect. So if we oscillate a signal between those two optical frequencies we should be able to transmit data in the optical range in a way that is invisible to the human eye yet readily machine-readable. So we could have a light flashing coded information that a machine could detect but which would appear to be constantly shining to the human eye. Now, what good is that?

I dunno yet. Visual radiation has the advantage that the atmosphere is generally transparent to its propagation (which is why our eyes have evolved to see it). but so do radio and lots of other invisible waves. So the key is that we want a signal that is necessarily visible for some reason, but in which machine readable information can be transmitted invisibly. What is the application?


3 comments:

Anonymous said...

maybe somehow used in air traffic control? Visual signal plus distance and speed info passed between planes?

sean michael ragan said...

So it turns out that my inferences about the existence of too visually indistinct but energetically distant colors of visual light are probably flawed. The solution to the paradox that arises between the circular and linear presentations of the visible spectrum is this: There is no single wavelength of light which is "magenta" in color; rather, monochromatic light at the spectral extremes can be either red or indigo, and "magenta" is produced by a mixture of both colors of light. I'm not entirely certain this is correct, but it seems much more plausible than the paradox I've proposed. The door prize goes to my friend and colleague Ellie for suggesting this solution.

Don Jusko said...

Magenta is 3,800 angstroms.
The rainbow edgewise is the spectrum in a straight line. Both red and magenta go to clear where the two edges would meet if connected in a circle. That's where Michael's idea measures the energy levels that can't be seen.

Michael,
What you are calling Violet is Magenta #13 on a 36 color wheel.
Using my Real Color Wheel that matches the way color elements get darker as in the crystals they make was good. The world sees what the real pigment color looks like when converted to an RGB image.

This link is to an image map of that color wheel where all the dots match pigments with names and brands.
http://www.realcolorwheel.com/colorwheel.htm

Don Jusko, Real Color Wheel
djusko@realcolorwheel.com