#621 overall
Color Course Notes (Maureen Stone)
SIGGRAPH
1 recommender · 1 thread · recommended in 1 year, 2010 · score 0.8
A set of SIGGRAPH course notes by Maureen Stone covering color science and how human color vision works, aimed at practitioners who work with color in graphics. The notes are presented as an accessible, rigorous alternative to conventional artist color theory.
Commenters describe it as something pretty accessible and good that gives a more accurate grounding in color than traditional color-theory material.
Science & Engineeringintermediatefreeself-paced text#color science#human color vision#color models#graphics#stone
Recommendations by year
20062026
To be honest (and I mean no offense), a lot of that material is somewhat misleading. For example, this statement,
comes straight out of the 19th century (or the 17th) – for coverage of primary colors see http://www.handprint.com/HP/WCL/color6.html – and this:
is not the way those terms are used today by people trying to use precise terminology. Intensity is reserved for counts of photons, chromaticity encompasses two dimensions, and as you describe it what you’re really talking about is chroma, not saturation.
All the stuff about triadic and tetradic and alternate complementary schemes, while it seems to be popular among some “color theorists” (whatever that means) is pseudoscientific creation of jargon divorced from any evidence-based underlying structure.
Edit: definitely take a look at the Maureen Stone SIGGRAPH course notes for something pretty accessible and good, http://graphics.stanford.edu/courses/cs448b-02-spring/04cdro...
I agree that this blog post takes a sort of silly tone, and also that many artists are very sophisticated in their understanding and use of color.
He goes wrong in a few places, for instance in using quite confusing coloration of his diagrams, and he really shouldn’t put the labels R, G, B on cone responses: these are too evocative and are likely to confuse as much as they enlighten.
But to be fair, basic paint mixing models in artists’ heads often aren’t very good, and many artists would really benefit from studying human color vision more seriously. Explaining how vision works to visual artists, especially amateurs, is one of the more rewarding things I ever do: you can see these lightbulbs going off in their heads as a bunch of stuff that they know “intuitively” and experientially is given proper names and ordered into a comprehensible model.
I think these spaces have no place being taught to non-specialists. They are not intuitive to humans, being based on particular technology rather than on human vision. Any user interface which exposes these color spaces to non-experts is letting them down. Just as bad is using HSL or HSV or some similar trivial derivative space. Image editing software, in particular, is dramatically harder to use than it would be if the dimensions were at all relevant to vision.
Specialists looking for models related to vision mostly use open, publicly developed and specified models. At root, most of modern colorimetry is based on the CIE system, which has served very well since the 1930s. Recent fairly effective models that would be useful for software include CIECAM02 and IPT, which behave better in some ways than CIELAB, but are a bit more computationally expensive.
That’s reasonable. Let’s start with some pointers to good sources!
For this audience, Maureen Stone’s SIGGRAPH notes from 2001 are a nice first start, with lots of pointers to other good resrouces: http://graphics.stanford.edu/courses/cs448b-02-spring/04cdro...
For anyone who wants to really dive deep, I recommend http://www.handprint.com/LS/CVS/color.html (edit: Actually, I see that the blogger does link there. So that’s nice.)
Also, anyone looking for pointers to books, feel free to email me (address in profile), and I can point you in some direction based on your specific interests.