Path: utzoo!utgpu!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!bloom-beacon!bu-cs!purdue!decwrl!sun!pitstop!sundc!seismo!uunet!portal!cup.portal.com!Classic_-_Concepts From: Classic_-_Concepts@cup.portal.com Newsgroups: comp.graphics Subject: Re: Palette Optimization Message-ID: <12744@cup.portal.com> Date: 18 Dec 88 19:08:18 GMT References: Organization: The Portal System (TM) Lines: 16 Drew Olbrich writes: > "....Can anyone give me some references...on optimizing a color palette > (That is, if I have a 24 bit color image, but my video buffer only handles 56 colors at a time, how do I go about choosing the "best" 256 colors..." Drew you've bitten off a big one here, since the answer lies not only in mathematics, but in perception and biology, as well. You'll find that colors which are mathematically close or distant will not match perceptually in the same degree. That is, palette colors which are separated by 10 or 15 numbers may appear the same to the human eye in one part of the palette and quite different in another, particularly if it hits a 'color boundary'. To compli- cate matters, each system relies on a different color selection system and mapping system--a solution on one won't work on another. For the time being, writing a 'visual selector' and running through the palette with 'yes/no' re- sponses after the field has been narrowed mathematically to 350 or so might be the best strategy. It will at least eliminate those which are visual 'duplicates' or unacceptably unattractive colors. (continued...)