7
$\begingroup$

The spectrum of the Sun as seen at sea level can be seen at https://commons.wikimedia.org/wiki/File:Solar_Spectrum.png so we can see that wavelengths around green to yellow are the ones that are the most present. The human eye seems to be more sensitive to green wavelengths (around 555 nm which is plain green) compared to others: see https://en.wikipedia.org/wiki/Color_vision#/media/File:Eyesensitivity.svg.

However the Sun appears mostly white/yellowish. I don't understand why. The human eye sensitivity for orange/yellow/blue is lesser than for green. Much lesser in case of red wavelength. So even though the spectrum is a continum of all these wavelengths why is the eye fooled toward white?

And why more toward yellow than green?

Thanks.

Improve this question
$\endgroup$
1

4 Answers 4

Reset to default
1
$\begingroup$

Note the vertical scale on the two graphs you gave: The solar spectrum at sea level is given as an intensity (power per area), and it is very nearly flat over most of the visual range. The eye sensitivity is given as a percentage, which the wikipedia page where it is used does not explain beyond calling it "normalized" and "relative brightness sensitivity." If this percentage is akin to a quantum efficiency, the probability that any one photon gets detected, there is a very natural possible explanation for this effect: If rather than a power-based intensity we use a photon number based intensity for the solar spectrum, its maximum will be at lower photon energies (redder colors) where the same power corresponds to more photons.

Perceptions are always tricky: Neither our eyes or our brains tend to function quite the way one might naively expect, in lots of ways. Hence much of the full explanation might not depend much on (photon) physics at all.

Improve this answer
$\endgroup$
4
  • $\begingroup$ I don't think the percentage is akin to quantum efficiency so the fact that there's a possibility that there are far more "red photons" than "green photons" reaching our retina does not imply that we'd see an equal amount of red and green and hence a white light from the Sun. The reason I don't think it is so is because of green and red lasers having the same power. The green laser appears far more brighter than red lasers even if their power is the same, hence many more "red photons". The eye has a much higher sensitivity for green light. So I still don't know the answer to my question... $\endgroup$
    – untreated_paramediensis_karnik
    Apr 3, 2015 at 0:15
  • 1
    $\begingroup$ There is no contradiction here: If you take a 670 nm red laser, accordinging to the eye sensitivity plot you linked to, the eye only has 10% sensitivity for that wavelength (a reduction of a factor of 10), which dominates the photons/energy difference that gives you about 1.2 times as many photons at 670 nm compared to 555 nm. For yellow light, say at 580 nm, they eye still has nearly 90% of its peak sensitivity acc. to that graph, and the story may well be different. The overall sensory impression obviously is a (possibly nonlinear) combination of all contributing spectral regions. $\endgroup$
    – user73762
    Apr 3, 2015 at 7:28
  • $\begingroup$ You are right. By the way I think I may have found out the answer to my original question: the solar spectrum as seen on Earth's surface has much less violet and blue than both green and red according to the first link I posted. So mixing all colors should be slightly like mixing mostly green and red (and yellow/orange) than mixing blue/green/red, hence the yellowish color we see instead of greenish or white. $\endgroup$
    – untreated_paramediensis_karnik
    Apr 3, 2015 at 14:31
  • $\begingroup$ Yes, that works in the right direction and is hence plausible. Plus, compared to a solar spectrum that may be integrated over all the sky, the sun will have even less blue considering the contribution from the rest of the sky. $\endgroup$
    – user73762
    Apr 3, 2015 at 17:33
1
$\begingroup$

And why more toward yellow than green?

Human vision and perception of colors is a complex process. It is safe to say humans are not very good at determining the actual spectral distribution of light they see. The sunlight may very well have frequency distribution that has maximum in green and humans may still see it as having different coloration. That is because the frequency distribution characterizes square of the Fourier component of electric field of the sunlight and this need not have necessarily any simple relation to how humans perceive colors for light that is not monochromatic, like sunlight.

Human perception of color depends in a non-trivial way on the whole spectrum of the light, not just the part where the maximum is.

Well-known daily example of this is perceiving color of light emitted by the RGB pixels of a TV / computer monitor. When you look on the bright bar above, you probably see bright yellow, but in fact the Fourier component of the electric field of the light coming from this rectangle has almost negligible component at frequency corresponding to yellow monochromatic light. The perception of yellow is accomplished with combination of red, green and blue light regulated to appropriate intensities.

Improve this answer
$\endgroup$
3
  • $\begingroup$ I understand. But the wavelengths distribution is rather flat albeit showing a maximum in the green wavelengths, adding the fact that our eyes sensitivity is much greater for green light than any other color, I still don't understand why the Sun doesn't appear greenish instead of yellowish. One explanation that I could guess is that there are somehow much less violet/blue light from the Sun than other colors (->the distribution of wavelengths is not so flat) and so the mix of colors would be more toward green/red (hence yellow) than blue/green/red (white). That's my explanation so far. $\endgroup$
    – untreated_paramediensis_karnik
    Apr 3, 2015 at 0:22
  • $\begingroup$ There is no reliable ground for holding on to the idea people should register the color of thermal radiation with spectral function $\rho(\lambda)$ as belonging to the color family based on the color registered for monochromatic light of wavelength equal to position of maximum of $\rho$. Spectral function is a technical concept, it is non-unique ( for example, position of maximum of wavelength distribution does not correspond to position of maximum in the frequency distribution) not the same ) and has nothing to do with human vision. $\endgroup$
    – Ján Lalinský
    Apr 3, 2015 at 19:32
  • 1
    $\begingroup$ The fact that thermal radiation has flatter spectral function $f$ than RGB elements have poses no apparent reason for us to expect thermal radiation to be easier to evaluate spectrally by human vision. It may be more sensitive to green light in a controlled experiment where such sensitiveness is measured, but this does not imply that we should see spectrally broad sunlight as green. The integral of $f\lambda)$ over range of wavelengths may be more important than the position of its maximum. $\endgroup$
    – Ján Lalinský
    Apr 3, 2015 at 19:40
1
$\begingroup$

Mostly working off this wikipedia article, so information would best be further confirmed.

From the article, "The color yellow, for example, is perceived when the L cones are stimulated slightly more than the M cones...." (Further questions here probably is best asked in the Biology StackExchange) Very much as Ján Lalinský says, colour perception in humans has many non-trivial behaviors - single-wavelength response is not generalizable to broad-spectrum response.

Looking at the Normalized responsivity spectra, the L and M cones both have very broad responses, extending out over 100 nm in either direction. L cells also seem to have a broader response, which might cause them to activate stronger in even-intensity-broad-spectrum light.

Aside from that, based off the spectrum of sunlight at Earth's surface (taken at one point, but should be roughly similar everywhere), while there is a slow drop-off going from green towards red, there is a very sharp drop-off going from cyan to purple. Therefore, since M cells respond more to blue-purple than L cells do, this probably overwhelms the weaker intensity drop-off on the overall shift of L cell response range towards the red range of sunlight.

Either effect, or perhaps a combination, would produce the slightly stronger stimulation of L cones than M cones necessary to produce the sensation of Yellow light.

Improve this answer
$\endgroup$
0
$\begingroup$

I am just rewriting what I wrote in a comment to make it more visible, since I think I found out the answer: the solar spectrum as seen on Earth's surface has much less violet and blue than both green and red according to the first link I posted. So mixing all colors should be slightly like mixing mostly green and red (and yellow/orange) than mixing blue/green/red, hence the yellowish color we see instead of greenish or white.

Improve this answer
$\endgroup$
3
  • 1
    $\begingroup$ If you were floating in space, the Sun would appear white. The reason it looks more yellow than white is because of the atmosphere, which scatters the blue wavelengths much more efficiently than the red. This is why the Sun becomes redder as it sets (since its light passes through more atmosphere), and also why the sky is blue (since the blue photons are scattered all over the sky). $\endgroup$
    – pela
    Sep 30, 2015 at 8:36
  • 1
    $\begingroup$ @pela I don't think that is right. The sun is considered a YELLOW star, it is G class. White stars are class A. $\endgroup$
    – Ambrose Swasey
    Aug 31, 2017 at 18:07
  • $\begingroup$ @AmbroseSwasey: The Sun is white indeed, but you're right that it's "on the border" of yellow stars in most HR diagrams. But the color coding is somewhat loosely defined, as if depends on human receptors. If for instance you color code according to where it emits most of its light, the Sun is a green star. $\endgroup$
    – pela
    Aug 31, 2017 at 19:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.