Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

What decides the colour of light?

  1. Nov 12, 2015 #1
    Consider a beam of light passing through a slab of some refractive index.
    We know that the speed and wavelength of the light changes, but its frequency remains the same.
    Since the wavelength of the light changes, does its colour change, or does it remain the same as its frequency remains the same.
     
  2. jcsd
  3. Nov 12, 2015 #2

    DrClaude

    User Avatar

    Staff: Mentor

    Color only has meaning when the light reaches the eye. All that counts is the frequency.
     
  4. Nov 12, 2015 #3

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Although people tend to talk in terms of the wavelength of light, that quantity keeps changing as it passes through different media (in particular, it is different whilst it is actually in the eye). Frequency is not changed (as Dr Claude pointed out). But pretty well all the light we see consists of a mixture of frequencies. We seldom come across monochromatic light in nature. Apart from lasers and some electric discharge lamps, the light is far from pure.
    It is vital to distinguish between colour and wavelength at all times.
     
  5. Nov 12, 2015 #4
    Correct me if I'm wrong, but colour is, essentially, the minds way of detecting frequencies. Instead of X Hz or Y Hz, our brain recognizes different frequencies as different colors.
     
  6. Nov 12, 2015 #5

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    A partial correction. There is a combination of frequencies involved with most colours.( the ones that are not "spectral" colours). Think in terms of musical chords rather than just tones.
     
  7. Nov 12, 2015 #6

    phinds

    User Avatar
    Gold Member

    HUH? Frequency and wavelength are inverse properties. How can one change and not the other?
     
  8. Nov 12, 2015 #7

    phinds

    User Avatar
    Gold Member

    I think it's probably more correct to say that color is how the brain categorizes the electrical signal sent to it by the optical nerve rather than that the brain "detects frequencies" directly.
     
  9. Nov 12, 2015 #8
    Love the analogy . Thanks.

    Because velocity decreases. Sophiecentaur was referring to light travelling through different media.
     
  10. Nov 12, 2015 #9
    The proportionality constant (speed of light) changes so that the frequency stays the same.
    The wavelength changes due to the changes in speed of light in various media.
     
  11. Nov 12, 2015 #10

    phinds

    User Avatar
    Gold Member

    Got it. Thanks. I'm slow today. Well, OK, I"m slow every day, but ...
     
  12. Nov 12, 2015 #11

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I always say that human colour perception is a really poor spectrometer. It is sooo easy to fool. And it doesn't matter at all.
     
  13. Nov 12, 2015 #12

    phinds

    User Avatar
    Gold Member

    I disagree that it doesn't matter. There are situations where subtleties in color do matter.
     
  14. Nov 12, 2015 #13

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Of course - but the eye is still a lousy spectrometer. That is what doesn't matter because it is never called on to do that job. Colour is not wavelength, is it?
    The eye is very good at resolving small differences in perceived colour (when it matters) but that doesn't involve measuring wavelength but combinations of the outputs of just three groups of sensors.
     
  15. Nov 12, 2015 #14
    This seems similar to Edwin Land's Retinex theory:

    "The Retinex Theory of Color Vision

    Λ retina-and-cortex system (retinex) may treat a color as a code for a three-part report from the retina, independent of the flux of radiant energy but correlated with the reflectance of objects "

    The Retinex Theory of Color Vision SCIENTIFIC - CiteSeer
     
  16. Nov 12, 2015 #15

    phinds

    User Avatar
    Gold Member

    Ah. That I agree with. There are people who can reliably discern very subtle differences in color (which is what I mean that mattered) but if they were to look at one of those colors one day and one close to it the next day I doubt they could tell the difference.
     
  17. Nov 12, 2015 #16

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I remember reading that paper a long time ago when I was involved in colour TV. The well known tristimulus theory of colour vision seems to overlap the Retinex Theory. Reflectance is not the only thing that counts for a lot of our visual input these days (TV displays and projected film) and it is altogether a very complicated business. The eye manages to process out things like the illumination in assessing the colour of an object. That is truly amazing and the description of the process - 'integrating to grey' is a bit of an oversimplification. It's about all that your automatic digital camera colour correction can manage. The eye seems to extract information at a far deeper level, based on context and memory. Brilliant and the Land paper makes a good effort at describing what goes on.
     
  18. Nov 12, 2015 #17

    DaveC426913

    User Avatar
    Gold Member

    Or four, in some rare cases...
    :smile:
     
  19. Nov 13, 2015 #18

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Tetrachromacy is one extreme of colour vision, I guess. Colour vision is a very personal thing and those tristimulus response curves (look em up folks) are the result of a lot of statistics, conducted on a lot of subjective results with a lot of people (probably a limited racial spread, though), I believe. There is much more spread in the responses between different people than your average conversation acknowledges. The system was aimed mostly at getting a good enough display and printing method to satisfy enough people with the 'accuracy' of copied colours. The proof of the pudding seems to suggest that RGB and CMY depiction of colours is 'near enough'.
    One of these days, perhaps, someone will come up with a TV system that uses more than three analyses and more than three basic phosphors. That could be really impressive and we would then start to realise the limitations of what we have at the moment. Colour printing just can't get away with three colours when the colours really count.
    There are parallels with Stereoscopic displays, which are impressive but very limited, in fact and multi channel surround sound systems which do better than bog standard stereo sound.
    Our brains are definitely on the side of the manufacturers, though. They desperately try to get sense out of these artificially presented sensations, despite the imperfections in the reproduction system.
     
  20. Nov 13, 2015 #19
    Its still pretty damn impressive for all that. And as a spectrometer it serves us very well.
     
  21. Nov 13, 2015 #20

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Of course, our colour vision serves us very well - in terms of our survival, and we should not expect anything more than that. (Evolution / Nature never does more than necessary). As a spectrometer, the eye is actually totally inadequate. It cannot even tell the difference between spectral Yellow and a combination of two monochromatic Red and Green lights. If I bought a spectrum analyser that could be fooled as easily as that then I would send it back to the shop.
    I don't understand why people get all defensive about their bodily system when someone points out its inadequacies. Our vision is what it is. It has no evolutionary advantage in being a spectrometer - so it never developed to be one.
    "It doesn't mean you're a bad person." :smile:
     
  22. Nov 13, 2015 #21
    The cones in the retina of the eye are stimulated this way, 64% react to red light of frequency centred about 650 nanometers about 33 % react to green light centred on 540/550 nanometers and just 2 % react to blue light centred on 450nanometers. SO if you look at a banana for example, all but the frequencies between 570 and 580 are absorbed and the 570/580nanometer light is reflected to your eyes. The cones respond as indicated and the signals in terms of amplitudes from the cones are transmitted down the optic nerve to the brain. The brain interprets these amplitudes/ decodes them if you will, and responds with the result that you are looking at something with a colour we call yellow.
     
  23. Nov 13, 2015 #22

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    A very nearly monochromatic bannana skin? Hardly likely. It's surface colour, under white illuminant will probably sit around half way between white and spectral yellow on the CIE chart. The pigment will probably be a mix of several natural dyes - at least it could well be. It sure ain't spectral.
    If I'm being picky, it's to raise a bit more awareness about the nature of colour.
     
  24. Nov 13, 2015 #23
    Some people actually study reflection spectra of bananas. :)
    http://ucce.ucdavis.edu/files/datastore/234-953.pdf
    The graph in figure 3 tells it all.
    For a well ripened banana, the reflection coefficient (for visible light) is maximum in the range 550-680 nm. And is at least 20% for the rest of the visible range.
    One of the pigments is chlorophyll but obviously, not the only one.
     
  25. Nov 14, 2015 #24

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    And probably the curvature, too. :smile: (I just watched the Hugh Fearnley-Whittingstall programme on Cosmetic vegetables; we are so very fussy about curvature in veg)

    That spectrograph is interesting as it shows a peak that is 'identifiable' as 'a yellow' (as expected). We would all agree that bannana colour could be described as a bright or strong yellow. However, looking at the total area of the rest of the curve, in the visible range, it is nearly the same as the area of the portion that you could describe as yellows. So, it's fairly desaturated and far from spectral.
    I was looking in my massive photo library for a convincing picture of a bannana so that I could look at the RGB components of its yellow colour. I haven't found a picture yet but I may take the trouble to photograph one later today. If it's anything like the other bright coloured objects I have on file, there will be very significant B contributions, along with the G and R (which produce the recognisable yellow). Looking at a brightly clothed audience at an outdoor sport event on a sunny day, it is hard to find objects that are actually 'saturated' colours. Of course, your TV display will never give you spectral colours because they lie outside the gamut, encompassed by the phosphors.
     
  26. Nov 16, 2015 #25
    Are you saying that it's more proper to note that we are seeing
    red at 400–484 THz rather than seeing at 620–750 nm. I know these are equivalent, but in terms of perceived color we are responding to frequency?
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook