I must admit, while the community aspect is neat, I find all these tools for creating color schemes (like “ColorSchemer”) pretty uninspiring. Every one I’ve seen has a pretty artificially inflexible interface, and most are based on device-oriented models of color, rather than anything based on (these days quite advanced) scientific understanding of human color perception; basically, they ignore most everything that color scientists have learned in the last century of research. Few seem to have any particularly insightful or interesting interface ideas either. Adobe’s Kuler is kind of neat, but it too fetishizes these nearly completely arbitrary (w/r/t perception) relations, in its definitions of various color scheme “types”.
Any specific requests on what you would like to see in a color scheme app? It seems you have some knowledge in the color science space and we'd love hear your ideas.
Sure. First, a big part of the point of a color scheme interface is to be able to orient and relate the colors in the scheme, but that’s only very useful in a perceptually relevant model such as Munsell’s, CIELAB, or CIECAM02.
The best analogy I’ve thought of is to the interface of a sink or shower. If you have two knobs for adjusting amounts of hot and cold water, then changing either temperature or pressure requires adjusting both. Given that there’s pretty much no situation in which you just want to make adjustments in precisely the way that one of your hot/cold knobs accomplishes, the interface ends up not being particularly intuitive (of course, since there’s feedback, you can still get where you want without too much trouble, just like in color; but the dimensions of adjustment end up being incidental rather than instrumental). Saving records of “hot 10, cold 6” or whatever are less useful than “pressure 16, temperature 5/8" would be (these numbers are made up; even better would be pressure «foo» pascals, temperature «bar» degrees).
The color dimension most important for perception of fine detail, for instance in making text readable against a background, or spatially identifying patterns in data, is “lightness”, the perception of which is pretty well modeled by Munsell value, CIELAB L*, or CIECAM02 J.
If you’re committed to making great color tools and can take some math, I recommend reading Mark Fairchild’s book Color Appearance Models; it has great material about human color vision generally, and is the best up-to-date color science textbook I know of. Especially w/r/t uses like cartography/information design, you might take a look at some of Cynthia Brewer’s work (here’s a presentation to the ASA http://www.personal.psu.edu/cab38/ColorSch/ASApaper.html which has a nice bibliography at the end). Here’s an IBM page on roughly the same subject:http://www.research.ibm.com/people/l/lloydt/color/color.HTM . As ever, Edward Tufe’s books are wonderful. Envisioning Information has a part about color. Tufte himself recommends this book: http://books.google.com/books?id=cVy1Ms43fFYC but I don’t know anything about it beyond the title. Bruce MacEvoy’s handprint website has great explanations, but is long enough to maybe be a bit overwhelming http://www.handprint.com/LS/CVS/color.html .
For learning about the historical use of color in art, and our changing understanding of it, John Gage’s book Color and Culture is just lovely. http://books.google.com/books?id=oq_GtjmoTNgC
I spent a bunch of time recently re-writing http://en.wikipedia.org/wiki/HSL_and_HSV which hopefully explains why these models shouldn’t be the basis of this sort of tool. :-)
----
As for my own specific interface ideas: I’ve thought a lot about it, but to really flesh them out will just require building my own tools and iterating. I’ve wanted to for quite a while; hopefully it’ll happen sometime. (Beyond color/color scheme pickers, I really want to work on tools for adjustment of color in photographs, since the ones offered by, e.g. Photoshop, do very little of what I want them to.)
how well does "the general public" get the idea that colour alone is worth thinking about? won't this suffer from being perceived as an "incomplete design place"? or is the idea that there's a niche for customizing sites (like twitter) where all you can really change are (css) colours and background images?
Alas.