Confused about human sensitivity to color

[atrix621]

I am trying to design an optimal colormap for display interval data.

I would like to create a scale that is linear in perceived luminance while adding a hue component to help mitigate the simultaneous contrast effect. However, I am confused on what hue to pick.

The longest curved path I get through CIELAB space that is still in the sRGB gamut is with hue=325 degrees -- that gives me a path where the chroma is maximized. On the other hand, several different sources tell me that the human eye is most sensitive to yellow-green (http://deron.meranda.us/ruminations/purple/ http://www.physicsclassroom.com/class/light/u12l2b.cfm) or magenta (http://www.atmos.ucla.edu/~fovell/AS3/theory_of_color.html ). What is actually the correct transfer function?

Additionally, Pizer and Zimmerman in "Color display for ultrasoundography" say the human visual system is most sensitive to luminance changes in orange-yellow hues but humans are most sensitive to hue changes in magenta, but they list no sources.

 

[Ibix]

I can't really help you, but I do know that John Barbur at City University in London was actively researching in this area around ten years ago. That might be a name to start searches in the literature. I recall that he used a green-on-nicotine-yellow colour scheme in his PowerPoint slides, which looked horrendous but was definitely easy to read. :D