1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

B The truth behind true violet

  1. Mar 30, 2017 #1
    IMG_8530.jpg

    As you can see the red line fades out and stops at roughly 420nm.

    Computer generate violet by mixing blue and red, so how do we see natural violet (405-420nm) if only blue cones react to it??

    First I supposed that violet affects blue and red cones, but that seems not to be the case on the graph. The interesting fact is, that the increasing intensity of that violet light makes it look "whiter" aka affects the green cones too- staying in my hypothesis.
    Please proof me wrong if it's possible.

    Thanks
     
  2. jcsd
  3. Mar 30, 2017 #2

    Drakkith

    User Avatar

    Staff: Mentor

    According to wikipedia. the red cone actually responds to two different regions of the spectrum. The first is the yellow-red end and the 2nd is a small region overlapping the blue cone near the violet end of the spectrum. So violet light stimulates both blue and red cones, allowing it to be approximated by RGB color systems like those in your computer screen.
     
  4. Mar 30, 2017 #3

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Basically, there are TWO Violets!! :eek: But it's nothing to lose sleep over.
    The 'Violet' that a computer monitor can present cannot be the same as 'spectral' violet. If you search for links to the CIE chromaticity diagram you can see how the three analysis curves in your post can be condensed onto a two dimensional chromaticity chart. The three curves have very low values around violet and I think the errors are quite high in that region - the lines sort of peter out due to human response and the shortcomings of the tests which have to be used. No one can put a meter on the sensor outputs and get reliable Voltage values, as with a photoelectric sensor. This link is a bit long winded but has lots of info about colorimetry. In particular, it shows how the phosphors of a TV display can reproduce a range of colours - but only the colours within the Gamut ( the triangle) of the three phosphors when plotted on the CIE chart. Any colour within the gamut can be represented by an appropriate mix of the three phosphors. The term "violet" is just a name for a colour somewhere down in the bottom left hand area of the CIE chart.
    The Curved border of the CIR chart represents the spectral colours and the tristimulus values that (average) people assign to those colours. There are very few instances where the actually see 'saturated' colours in real life (on the edges of the CIE chart). Rainbows are particularly useless as sources of saturated spectral colours because they are heavily desaturated by the grey / white / blue off the sky behind them. If you look at the spectrum from a 'white' light source, the luminance of the blues (including indigo and violet) is very low so it's all a bit of a muddle down that end and it's very hard not to mix up the two violets. Other areas of the CIE chart are much more useful; Greens and 'skin tones' are higher luminance colours (evolutionary advantage, of course)
    That link I posted is pretty good - better than Wiki because it is fuller. It makes the point that the Red Green Blue description of our analysis is far over simplified.
    There is also the point that a TV phosphors are not very saturated colours - the triangle is far away from the edges. Why? I think it's because no one has found 'spectral' phosphors that are bright enough to produce a useful TV picture.
     
    Last edited: Mar 30, 2017
  5. Mar 30, 2017 #4

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor
    2016 Award

    Purple, like pink, brown, and pastels, are not spectrally pure colors and thus produced by the visual cortex. There are a lot of interesting mathematics involved when relating the (Newtonian) linear spectrum to the planar 'color wheel' that is taught in elementary school. One interesting read is here:

    http://link.springer.com/chapter/10.1007/978-94-017-0934-7_32

    Other, stranger optical effects are well-known: impossible colors and Benham's disk:

    http://science.sciencemag.org/content/221/4615/1078
    https://www.exploratorium.edu/snacks/benhams-disk
     
  6. Mar 30, 2017 #5

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    I would go further than that. You don't see "spectrally pure" colours anywhere but in a lab or, perhaps from a laser source.
     
  7. Mar 30, 2017 #6
    You should define what the plots are that you posted. Apparently they come from a Wikipedia article and are identified as "The normalized spectral sensitivity of human cone cells of short-, middle- and long-wavelength types."

    Standard observer color matching functions are generated by finding the combination of R, G and B light sources that visually match the color of each wavelength in the 400nm to 700nm wavelength range. See the section Color Matching Functions at the same Wikipedia article: https://en.wikipedia.org/wiki/CIE_1931_color_space

    For a wavelength such as 420nm, some mixture of blue and red is required for a visually match. The matching occurs in your brain.
     
    Last edited: Mar 30, 2017
  8. Mar 30, 2017 #7
    oh no how am i supposed to sleep tonight.
     
  9. Mar 31, 2017 #8

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    The eye is not a spectrometer and it has no need to be. That is important to bear in mind.
     
  10. Mar 31, 2017 #9

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor
    2016 Award

    Yes, but that's not really what I mean. For example, λ=555±0 nm is a spectrally pure color most of us would call 'yellow'. There is no such assignment for purple/brown/pink, etc.
     
  11. Mar 31, 2017 #10

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Yes - that's certainly one way of looking at things but that is treating all the de-saturated spectrals as different beasts from all the de-saturated non-spectrals, which are allocated all sorts of special names. I think this is more of a cultural thing - very deep rooted - and I, for one, think that we should wean people (at least the better informed PF clan) away from the colour equals the wavelength idea, because it doesn't, except on the outside curve of the CIE diagram.
     
  12. Mar 31, 2017 #11
    Brown is a low intensity yellow/amber/orange. It can only be seen in comparison to a higher intensity (brightness) color. In that sense it is related to gray.

    White, gray and black are all achromatic (without perceived color). The intensity relative to the surrounding visual field is how our eyes/visual system determine whether it is white, gray or black (even if they all have the same relative spectral power distribution).

    You can't project brown light any more than you can project gray light.
     
  13. Mar 31, 2017 #12

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    The Luminance value is a third parameter that needs to be added to the co ordinates on the CIE Chrominance plot (z axis). Your eyes analyse the colours a scene by trying to eliminate the mean luminance. Over a wide range of light levels, this works pretty well but it fails in very bright or very dim scenes (cones stop functioning properly).
    There is an area, somewhere in the middle of the CIE chart which is chosen, in any particular system, to be white. Just like the Deluxe colour charts and the light bulbs you can buy, they can present you with a range of 'Whites', each of which will look different and distinguishable from the others. So 'whites and greys' are not truly achromatic. RGB values on a TV will be equal for a good grey but only when coding and decoding are 'in step'. Your brain does its best to eliminate the wide range of illuminating 'whites' and does a good job of carrying a colour in memory so it can be matched at another time and under different illumination. (not perfect, of course). The colours we 'see' are a combination of illuminant and reflection /transmission characteristics. In TV, the colour is all illuminant (the glass is not tinted).
    If you look at a colour TV (or a projected film) you will be able to see browns in amongst all the other colours on the screen. A low luminance yellow, placed beside a high luminance yellow (identical R:G:B ratios) will look brown. That colour is in your head and, like all other colours, except spectral colours, consists of a spectrum of different wavelengths. So I would say that you cannot 'project' any colour. All you can do is 'suggest' to the viewer the sensation of that colour by a suitable combination of primaries which you project on a particular area of a screen, in the presence of other, reference areas or objects.
     
  14. Mar 31, 2017 #13
    Now you are touching into the difference between chromaticity (2 dimensional) and color space (3 dimensional and includes intensity).

    There is also a difference between looking at a light source directly vs how we normally perceive light: source --> object --> eye (the visual triangle). That is starting to get off topic.

    As for achromatic, it literally means without perceived color or hue. If you want to say that gray has chroma (or hue) then I would hazard to guess that your position is at odds with decades of color research. Just because we think of gray as a color (and you can buy gray paint for example) does not mean that it has color. You are talking about a difference in luminousity which is not the same thing as chromaticity.
     
  15. Mar 31, 2017 #14

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    If you are trying to be accurate then you must take into the account the illuminant, and the white balance. By definition, of course, an area of an image with no (very low) chrominance would be called Grey (white / black) but I have to ask you about the viewing conditions, background illumination and the rest of the picture. The perception of your 'grey' area could be affected by those conditions. It would be a "grey area" in more than one respect.
    Yes, that's true but which 'white' is 'white'? A surface that reflects all wavelengths equally could be called white but that doesn't mean it will look white. The chromaticity of that surface will depend on the illuminant used and there is a fairly small region of the CIE chart that will have 'no chromaticity'.
    In my view you are artificially assigning something 'special' to whites and greys, instead of treating them the same as all other parts of colour space. As with many other quantities in Science, it is relative.
     
  16. Mar 31, 2017 #15
    Surfaces have no color. They have a reflectance spectrum. What we perceive as color is a result of how our visual systems interpret a spectral power distribution. It also depends on the adaptive state of the visual so even the same spectral power distribution can be interpreted differently at different times.

    As for white, the chromaticity range that defines white depends on your application and source. If you are talking about daylight then you have to refer to the appropriate D-series illuminants, fluorescent lighting then ANSI/NEMA C78.376, and for solid-state (LED) lighting ANSI/NEMA C78.377.

    There is a fairly small region of any topic that is technically accurate. I was trying to give generalized information that might be useful to a wider range of members of this forum. If this is in fact a forum for spectroscopic and especially photometric experts then I apologize and will use appropriate terminology and references in the future.

    If you are trying to argue about a single exact white/achromatic point that applies to all situations then the whole discussion is useless since as I mentioned the definition of white depends on the application or market niche. A rose growing in a wheat field is a weed.

    In my view you are trying to nit pick my answer. I am not assigning anything special to whites and grays (you forgot about blacks) - I am using the term achromatic as it is generally used in color science and photometry.
     
  17. Mar 31, 2017 #16

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Actually, I thought you were doing the same. :wink:
    We are not far apart here. I agree with (and have made the same points) that colour is what the brain makes of the effect of light reflected from a surface or emitted from a light source - in the context of the surrounding light situation. I can see no particular significance in a region perceived as white / grey / very dark grey or a region perceived as bright yellow / brown / dark brown. You seem to say there is something special about 'lack of' chrominance. Isn't it just the same as velocities near zero; part of a continuum? In view of the adaptability of our colour vision to circumstances and to how (exceptionally well) we perceive colours in different circumstances, the concept of white / grey seems to be fairly incidental.
     
  18. Mar 31, 2017 #17
    haha - I too sense agreement.

    I wouldn't exactly call the concepts incidental - they are terms that really only have meaning in context.

    I know companies that have literally spent millions arguing over definitions of white light . . .
     
  19. Mar 31, 2017 #18

    Drakkith

    User Avatar

    Staff: Mentor

    Huh. I would have given them an answer for a measly few grand. :rolleyes:
     
  20. Apr 1, 2017 #19

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Since the (relatively) recent explosion in artificial for lighting, they are surely onto a loser. We spent millions of years evolving a colour sense that relies on Sunlight and we are unlikely to be fooled totally by an alternative source of illumination. That brief foray into fluorescent lamps was a real 'low' in the science and LEDs are doing a lot better - at least their spectrum looks continuous. Colour temperature specs for LEDs are very speculative, I think.
    But (imo) it won't be many years before LEDs are used for every form of lighting and we will hardly see Tungsten sources at all. And no one will need to look ill in a car park at night, once Sodium or Mercury have gone.
    You'll always have to take clothes outside of the shop in order to get a really good match with your favourite tie.
     
  21. Apr 1, 2017 #20
    Thanks guys
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: The truth behind true violet
  1. Is it not true (Replies: 7)

  2. The Truth (Replies: 35)

  3. Is this true? (Replies: 3)

Loading...