On 19 November 2010 09:46, James Morris <[email protected]> wrote: > I'm working a lot with a program that generates random palettes of > colour where a palette consists of 256 colour entries. Each entry > consists of the R, G, and B components. The program only allows you to > work with a range of colours such that you can say "I want (256 * > 0.05) random R components, (256 * 0.08) random G components, and (0.2 > * 256) random B components" (specifying how many random entries > expressed as ratios of the total number of entries is one of this > program's many idiosyncrasies). > > Taking the above example, means (roughly) generate 12 random numbers > to be used as Red components and spread them evenly across the 256 > entries, then interpolate to fill in the remaining entries between > each of them. Do the same for the Green components, only use 20 random > numbers instead, and for the Blue components use 51 random numbers. > > Attached are three examples of the colour palettes typically > generated. The filenames of the attachments indicate how many random > entries for each component are generated. > > The question is, if there was a colour palette for the all the colours > in the world (I'd increase the number of entries in the palette to > 16777216) and this palette was filled with all the colours of the > world... and one wanted derive from that colour palette a random > colour palette, what would one use as the random ratios for each of > the R, G, and B components?
Actually the real question is, given all the colours in the world, which of the colour components has the greatest frequency of variations , R, G, or B ? -- _ : http://jwm-art.net/ -audio/image/text/code/ _______________________________________________ NetBehaviour mailing list [email protected] http://www.netbehaviour.org/mailman/listinfo/netbehaviour
