Figure 1. Each curve indicates how well each of these two cone types absorbs light, plotted as a function of wavelength and scaled according to the absorption of the wavelength absorbed best.
In my introductory page and my page on variation in human color vision, there are graphs which show how animal photoreceptors absorb and/or respond to light. What’s that all about?
Figure 1. Each curve indicates how well each of these two cone types absorbs light, plotted as a function of wavelength and scaled according to the absorption of the wavelength absorbed best.
Photoreceptors as spectral pattern matchers
Let’s start simply with two cone types and monochromatic stimulation. We’ll use normal human S and M cones (I hope that doesn’t cause problems with adult content filters). The absorption spectra of these two cones are shown again in Figure 1.
We’ll let our monochromatic lights wander around a little bit... I’ll start off really simply to try to make sure we’re all on the same page. Figure 2 shows a monochromatic light with its energy concentrated at a wavelength of 533 nanometers. Admittedly this is not a terribly interesting figure. But it is a spectrum of a light that could be approximately produced in a lab, and its conceptual simplicity has made it a popular sort of spectrum to use for research. Its purpose here is to start getting you to think about the relationships between the physics of light and the physiology of color vision. To get a broad understanding, you might want to think of the spectrum in Fig. 2 and the two spectra in Fig. 1 and ask yourself, "does the shape of the spectrum in Fig. 2 more closely resemble the shape of the M cone spectrum or the S cone spectrum?". Go ahead. Ask yourself that...
Figure 2. The spectrum of a monochromatic light with its energy concentrated at 533 nm.
I’ll wait while you think of the answer...
Have you decided yet?
How about now?
Okay, I think I’ve waited long enough. The answer is the M cone spectrum. Don’t feel bad if you got it wrong; I’m sure you’ll do better next time. To make it easier, we’ll overlay the absorption and radiance spectra as in Figure 3.
Figure 3. Figs. 1 and 2 superimposed.
As I hope is now clear, 533 nm is part of the spectrum that normal human M cones absorb very well, and normal human S cones not so much. What if our monochromatic light carried its energy at a different wavelength? Rather than step through wavelengths one at a time, I’ll show you an animation in a separate page. The animation is almost a megabyte in size, so be ready for a wait if you’re on a dialup connection. Click here to see the animation.
I hope that you looked at the animation, because I will presume you now know some things I tried to explain on that page...
Figure 4 shows a more interesting spectrum. It is the radiance of sunlight reflected off of some distant trees. I acquired the spectrum when the sun was setting one evening while I was in Hoosier National Forest in Indiana. If you've been there, perhaps you recognize the spectrum. Either way, the question I’d like to put to you again is, which cone absorption spectrum does this radiance spectrum most closely approximate?
Figure 4. Human S and M cone absorption spectra with radiance spectrum of evening sunlight reflected off of leaf
We find the answer mathematically with a smidgeon of calculus. We multiply the absorption spectrum of each cone by the radiance spectrum of the light, and then integrate the result across wavelengths. The number that results from one of these integrations is proportional to the amount of light absorbed by the photoreceptor. Which of the two cone types, S or M, will absorb more light from the spectrum in Figure 4?
To make it easier to answer that question, I show you in Figure 5, the results of the multiplication of the spectra. To integrate means to calculate the area under the curves, so it should now be readily apparent that, relatively speaking, the M cone absorbs a lot more light than the S cone does from this example spectrum. (In case it isn’t obvious, the S cone*leaf light spectrum is the lavendar colored curve, the M cone*leaf light spectrum is the pinkish curve).
There is, however, another issue to be addressed. The spectral absorption curves in Figure 4 (and the computations used to generate Figure 5) were normalized. It is possible that in real terms, S cones could still absorb more than M cones. For instance, is the amount of visual pigment the same in the two types of cones? If there is a lot more S cone pigment than M cone pigment, then this would offset the fact that more of the leaf spectrum’s light is in the middle of the human-visible spectrum rather than at the short wave end of the spectrum.
Figure 5. Human S and M cone absorption spectra multiplied by radiance spectrum of evening sunlight reflected off of leaves
Go back to main Comparative Color Vision page
Go back to Comparative Human Color Vision page
Remainder under construction...
This page was last edited on February 15, 2005.
© 2004 - 2006 Mickey P. Rowe