On Tue, Mar 20, 2012 at 12:57 AM, Greg Ewing
<greg.ewing@xxxxxxxxxxxxxxxx> wrote:
Christopher Night wrote:
The idea is that if we could independently stimulate each of the three colors receptors (cones) in our eyes, then we could reproduce any visual sensation and thereby any color. The problem is ... there's no such thing as a wavelength of light that stimulates the middle (green) cones without also stimulating either of the outer two.
But that means we never experience the sensation of
having just the green cones stimulated, so there is
no need to reproduce it.
Unfortunately that's not the case. Here's a simplified example to show the problem. Consider four variables red, yellow, green and blue, representing the intensity of four different wavelengths of monochromatic light. Any of them can have any positive value. Say the amount of stimulation each cone receives is:
LOW = 1 x red + 0.5 x yellow + 0.2 x green
MEDIUM = 0.2 x red + 0.5 x yellow + 1 x green + 0.2 x blue
HIGH = 0.2 x green + 1 x blue
Now, the question is, how can you reproduce the response you get with yellow light, with just red, green, and blue? Yellow stimulates the low and medium cones with equal intensity, and the high cone not at all. So if (red, yellow, green, blue) = (0, 1, 0, 0), then (LOW, MEDIUM, HIGH) = (0.5, 0.5, 0). Using only red, green, and blue, the only way to get LOW = MEDIUM is to use equal parts red and green. But if green > 0, then HIGH > 0.
Of course you might say, maybe a different set of three wavelengths would be able to produce the same response as any monochromatic light. And some sets do better than others. But as I understand it, no set of three is perfect. (Also, these are obviously completely made-up numbers.)