What is color?

Discussion in 'Physics & Math' started by Asexperia, Nov 24, 2019.

  1. exchemist Valued Senior Member

    Messages:
    8,689
    Many heartfelt congratulations on this burst of lucidity!
     
  2. Google AdSense Guest Advertisement



    to hide all adverts.
  3. James R Just this guy, you know? Staff Member

    Messages:
    33,474
    There are objective measures of the "colour" of light, which are its frequency and its wavelength (which are related to one another). But the electromagnetic spectrum, of which visible light is only a tiny part, is continuous. Clearly we do not perceive a colour difference between light of wavelength 500.1 nm and 500.2 nm, for instance, so our perception of different colours must map to ranges of wavelengths. But since the spectrum is continuous, there are no physical boundaries between, say, yellow and green light. Instead what happens is that as the wavelength of the light is gradually increased, the yellow light becomes greener bit by bit, until at some arbitrary point we say it looks more green than yellow.

    It gets more complicated than this when we look at how the human eye actually perceives colour. Most of us have only three distinct types of chemical pigments in our eyes, each of which is sensitive to a different range of the colour spectrum. When we look at something, our brain essentially gets a three-valued signal corresponding to how much the colour cones of each type are excited by the light. Then the brain processes those signals and gives us a subjective interpretation of the colour of the light. The important point is that the colour perceived is not a "raw" wavelength value, but rather a combination of 3 different values determined by the physical structure of the receptors on the retina of the eye.

    Finally, there is the matter of how we choose to label the various perceptions we get from those three signals. That is partly cultural, and it turns out that different cultures can label the colour spectrum differently. Some cultures don't recognise certain distinctions between colours. For example, there might be cultures that do not distinguish yellow and green in their language; they use the same colour term for both. There are cultures whose words suggest that the sky is black, not blue.

    Colour is a complex thing. But the most important thing to note, I think, is that the colours we see and talk about are not inherent in the things we're seeing. Our perception of colour is tied up in our biology and even our culture.
     
    Asexperia likes this.
  4. Google AdSense Guest Advertisement



    to hide all adverts.
  5. exchemist Valued Senior Member

    Messages:
    8,689
    Yes. Evidence that colour is a perception and not simply related to the physics of light is that a mixture of long and short wavelength light (red and blue) is seen by the brain as purple, whereas violet light, which looks almost the same, is shorter wavelength than either red or blue.
     
  6. Google AdSense Guest Advertisement



    to hide all adverts.
  7. James R Just this guy, you know? Staff Member

    Messages:
    33,474
    Yes. Similarly, if you're reading this on a colour LED display (such as on a mobile phone, tablet or LED monitor) be aware that there are no yellow LEDs in the device. Any yellow you see on the screen is due to pixels putting out a suitable mixture of red, green and blue light, sufficient to excite the cones on your retina in such a way as to produce in you the sensation of "yellow", despite the fact that no light in the usual "yellow" wavelength range is involved.
     
  8. Asexperia Valued Senior Member

    Messages:
    1,403
    Excellent James R.

    Is it correct to consider red, green and blue as fundamental colors and that others are formed from a combination of these three ?
     
  9. arfa brane call me arf Valued Senior Member

    Messages:
    6,568
    Electronics is the science behind the display you are probably reading this on (obviously).
    The technology of color displays depends on more than a passing understanding of how humans perceive colors.

    So not that surprising then, that how displayed (electronically generated) colors mix together is important. There is therefore, a standard reference for this.
    The standards are a visualisation of the underlying vector space of color, which is a Hilbert space (lots of physical things are, though).
     
  10. James R Just this guy, you know? Staff Member

    Messages:
    33,474
    In general, no. Red, green and blue are no more fundamental that yellow or orange or purple in the sense of the physics of light. All colours are on an even footing as far as the light is concerned.

    But as it applies to the specifics of human vision? Maybe. R, G and B are "fundamental" in that by mixing them in various combinations we can generate the same perception in a person as any of the colours that are perceivable by human beings. However, you should be aware that "red" cones on the retina do not only detect red light, or anything like that. It is more the case that they work most efficiently when they are detecting red light. There's actually not a huge difference between red cones and green cones in the human eye. The difference between those cones and the blue ones is larger, which is one reason why red-green colourblindness is much more common than the other varieties of colour blindness.
     
    Asexperia likes this.
  11. Seattle Valued Senior Member

    Messages:
    5,265
    There's also the concept of "additive" vs "subtractive" color.
     
  12. gmilam Valued Senior Member

    Messages:
    3,144
    No. As Seattle said there are additive and subtractive primary colors. Your TV and computer screens use Red, Green and Blue for the primary colors. But when using ink, as in your printer, the primary colors are Cyan, Magenta and Yellow.
     
  13. Seattle Valued Senior Member

    Messages:
    5,265
    To look at it another way, if you take two flashlights and put a green filter over one and a red filter over the other and shine then so they overlap, you will get yellow light.

    If you mix red and green paint you will get grey.
     
  14. James R Just this guy, you know? Staff Member

    Messages:
    33,474
    The difference between mixing light and mixing paints is that the colours we see from paints are reflected colours, whereas the colours we see from light sources are directly emitted.

    For example, red paint reflects red light while absorbing green and blue. Green paint reflects green while absorbing red and blue. So if we mix red and green paint we get paint that absorbs all colours to a similar degree and it looks grey. On the other hand, if we mix yellow and blue paint, we get green. Actually the primary paint colours that are usually used are cyan (reflects blue and green), magenta (reflects red and blue) and yellow (reflects red and green). Mix cyan and yellow and you get paint that reflects green light while absorbing the red and blue.
     
  15. Seattle Valued Senior Member

    Messages:
    5,265
    The whole red, yellow and blue primaries (for paint) and the other argument for yellow, cyan, and magenta is just that, an argument.

    They both work for painting. Some say you can mix blue and red from cyan and magenta therefore cyan and magenta must be the primaries.

    However, you can also mix cyan (cool blue) and magenta red (cool) just as easily. Add a little blue to red and you get magenta. Add a little green to blue and you get cyan.

    The only paint color that is a true primary is yellow.
     
    Last edited: Dec 5, 2019 at 4:26 AM
  16. James R Just this guy, you know? Staff Member

    Messages:
    33,474
    Actually, it occurs to me that I'm thinking of ink colours with magenta, cyan and yellow. If you have a colour inkjet printer, chances are that it takes cartridges with those colours.

    The principle is sound, though. Colours, whether ink or paint, are produced reflectively, so if you want to make green paint by mixing other colours, then whatever you do has to produce paint that ends up reflecting green light.
     
  17. arfa brane call me arf Valued Senior Member

    Messages:
    6,568
    The question what is color is very much like asking how many ways are there to map the real line to itself.

    Already we have this map from the set of colors, to paint and to light-emitting devices, or to reflective and emittive sources of color, from a set of colors.
    We can ask how many different distinct colors (the set is a set of pointlike objects we choose to call colors) we can mix together, etc. The space of colors is additive, so all we need is scalar multiplication.
     
  18. exchemist Valued Senior Member

    Messages:
    8,689
    Gibberish.
     
  19. arfa brane call me arf Valued Senior Member

    Messages:
    6,568
    Colors are additive, so if you can choose a set of scalars from some field, you have the beginnings of a vector space. The space of colors (that is, what we perceive as different colors), is a Hilbert space.

    Some people will say a Hilbert space isn't a vector space, it's a space with rays through an origin. Some will say Euclidean space is a Hilbert space.
    Color, although it's a perceived thing, is a Hilbert space. There is a reasonably good wiki page explaining why.

    Of course physical things have a map to this space, if the things are colored or emit colored light.

    exchemist: shouldn't you be up a tree?
     
  20. arfa brane call me arf Valued Senior Member

    Messages:
    6,568
    Color perception
    Main article: Color vision ยง Mathematics of color perception
    Any true physical color can be represented by a combination of pure spectral colors. As physical colors can be composed of any number of spectral colors, the space of physical colors may aptly be represented by a Hilbert space over spectral colors. Humans have three types of cone cells for color perception, so the perceivable colors can be represented by 3-dimensional Euclidean space. The many-to-one linear mapping from the Hilbert space of physical colors to the Euclidean space of human perceivable colors explains why many distinct physical colors may be perceived by humans to be identical (e.g., pure yellow light versus a mix of red and green light, see metamerism).

    --https://en.wikipedia.org/wiki/Hilbert_space

    Please Register or Log in to view the hidden image!



    The white curves are the three 'resonance curves' for human pigments. This map says nothing about the neural level of perception, but represents the bare photochemistry, say.
     
    Last edited: Dec 6, 2019 at 9:47 PM
  21. arfa brane call me arf Valued Senior Member

    Messages:
    6,568
    Applying a bit of pedantry: saying color "is" a Hilbert space might be corrected by saying the Euclidean space which is the abstraction of color perception, is a vector space. The Hilbert space, as the quote above says, is the space of physical colors.

    But it's common to confuse an abstraction with the physics, most engineers understand this (ahem).
     
  22. Asexperia Valued Senior Member

    Messages:
    1,403
    In general, color is the subjective sensation produced by the reflection of light on objects.
     
  23. QuarkHead Remedial Math Student Valued Senior Member

    Messages:
    1,638
    What on Earth are you talking about? Hilbert space is a generalization of Euclidean space, both of which are therefore vector spaces
    And then what is this inverse of the physical colour "red"? What is the identity such that "red + (-red) = no colour"?

    Maybe you mean the space of all EM frequencies is an Hilbert space, which has nothing really to do with their perception - this is physiology.

    I suggest it is you who is confused.

    Advice to you: do not try to "learn" science from Wikipedia; you might learn some here, but a good text would be better
     

Share This Page