why is hue circular?

Everyone is familiar with the colour spectrum. If you pass white sunlight through a prism then it splits into the component wavelengths. The shorter wavelengths appear blue, the longer wavelengths appear red, and in between we have the familiar colours that I learned as school as Richard Of York Gave Battle In Vain, for the sequence red, orange, yellow, green, blue, indigo and violet, and that I have since understood is taught in the US as a person: Roy G Biv. I wonder if there are any other mnemonics that people know of? Of course, many people believe that Newton was in error when he identified 7 colours in the spectrum – he was probably influenced by Aristotle who wrote about there being 7 fundamental colours as there are 7 tones in the musical octave. I’ve posted about the indigo issue before – http://colourware.wordpress.com/2009/07/20/indigo-a-colour-of-the-rainbow/ – so won’t repeat that here.

Newton was probably the first person to create a hue circle (others, such as Forsius, created colour cicles but often included white and black in the circles). Newton created a true hue cirlce where he took the colour spectrum and wrapped it around, noticing that the two ends of the spectrum (where the reds become bluish and the blues become reddish) look rather similar.

Of course, there was a gap because the two ends of the spectrum did not quite match and thus Newton had to add in some purplish colours – these are hues that are never seen in the spectrum (and are sometimes called extra-spectral hues or non-spectral hues). The hues in the spectrum can be created by a single wavelength; however, the extra-spectral hues only occur when we see several wavelengths at the same time. For example, when we see short and long wavelengths together we can see purple.

In my lecture at the University of Leeds (www.leeds.ac.uk) this week someone asked “Why do the two ends of the spectrum look similar at all when the light is so different physically (at one end the waves are short and high energy and at the other they are long and low energy)?” Very very good question – if changes in wavelengths change the hue why should wavelengths that are so different look so similar?

So, why is hue circular? The answer is that it has very little to do with wavelengths and physics and more to do with human physiology. The human visual system captures light with three classes of cell (called cones) in the retinae of the eye. The signals from these cones are processed by the human visual system to create opponent signals (red-green and yellow-blue). This puts red and green opposite each other and yellow blue opposite each other and results in the perception of hue being circular. It also explains why some hues particularly contrast – sometimes called complementary colour harmony.

10 thoughts on “why is hue circular?

  1. Dear professor,

    I’m editing a new hungarian wikipedia page, the title is ‘Color vision’, about the human color perception.

    I would ask you, that can I paste the above image from the newton colour round to this article?

    Your sincerely,
    Balázs Németh

  2. Read Over Your Good Books In Verse (an old one from a group visiting us from Cumbria) and Rowntree’s Of York Give Best In Value (from an adult accompanying a school from York – clearly feeling more proud of a sweet maker than a potential king of England)

  3. I could not understand the last para:
    “The signals from these cones are processed by the human visual system to create opponent signals (red-green and yellow-blue). This puts red and green opposite each other and yellow blue opposite each other and results in the perception of hue being circular.”

    Could you please elaborate? Thank you!

    1. You’ll note in the original post above that there is a diagram showing how Newton noticed that although the short and long wavelengths couldn’t be further apart (physically) there are sort of similar visually. At the short end of the spectrum we have a purplish colour and at the long end we also have a purplish colour? Why is this? And why is hue circular?

      The circularity of hue is implicit in lots of colour spaces such as Munsell, CIELAB, NCS etc. Where does it come from?

      In fact, the CIELAB colour space – see http://apachetechnology.in/KC/Multimedia/DIP/Segmentation/Images/cielab.gif – results from modeling of the human visual system. The CIELAB a* and b* coordinates both result from opponent modelling. If you look at the equations that transform CIE XYZ into CIELAB a* and b* you will see this. So a* and b* are opponent or antagonistic signals. Once you put them together , as in the CIELAB a*-b* plane the circularity of hue just pops out.

      I am sorry if this is still not perfectly clear but I hope it helps a little.

      1. Thank you Professor.

        I also have confusions regarding the number of colors in the rainbow. While you have mentioned the 6 colors and self-observations of a rainbow to validate, I tried to google images of color spectrum of rainbow and that from a prism. I find that some images show a violet band while for others there is deep blue going into black. This wonderful photo collection has some images in which there is no violet while most have it (https://people.rit.edu/andpph/exhibit-spectrum.html). Other images on internet seem to be artificial rather than actual spectrum images.

        Then there is this video from Minute Physics (https://www.youtube.com/watch?v=9udYi7exojk).

        All this has caused a lot of confusion for me (as well as the surprising-frustrating feeling that even a seemingly simple concept is controversial but is taught in a very black & white manner in textbooks without acknowledging the gray areas or the historical development like Newton’s 7th color).

        From all these I find the colors in the rainbow to be ROYGCBV or ROYGCB (and that is when Orange and Cyan are considered as separate colors and not mixtures of RY and GB). Can you please give your comments on the number of colors in the rainbow and the above video?

        1. The MinutePhysics video you linked to is excellent and explains a lot. Thanks for that. Part of the problem is that when we use labels such as blue and violet these are quite crude ways of talking about colour. It’s a little bit like the argument that the best subtractive primaries are cyan, magenta and yellow but all the text books tend to teach the primaries as red, blue and yellow. Is there a confusion between people using blue and cyan names to represent the same thing? I think it is possible. I think this could account for some people calling one component of the rainbow blue and others calling it cyan. We’re just speaking crudely and with insufficient precision.

          I share your frustration about the field that is called colour theory – this presents a whole lot of stuff in an overly simplistic way. Some of it is not even true. I guess we just have to accept that not everything we read in textbooks (never mind the internet) is true.

Leave a Reply

Your email address will not be published. Required fields are marked *