Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!samsung!zaphod.mps.ohio-state.edu!wuarchive!husc6!m2c!jjmhome!cpoint!alien
From: alien@cpoint.UUCP (Alien Wells)
Newsgroups: comp.sys.ibm.pc.programmer
Subject: Re: 256 GrayScale Matching to 256 Colors
Keywords: Look Up Table Needed
Message-ID: <4158@cpoint.UUCP>
Date: 2 Mar 90 20:12:49 GMT
References: <858@dukempd.phy.duke.edu> <1990Mar2.043848.501@uncecs.edu>
Reply-To: alien@cpoint.UUCP (Alien Wells)
Organization: Clearpoint Research Corp., Hopkinton Mass.
Lines: 26

In article <1990Mar2.043848.501@uncecs.edu> cjl@uncecs.edu (Charles J. Lord) writes:
>It sounds possible, but the problem is that the color pallette
>on the VGA is 18 bits: 6 each R,G,&B.  To do a true gray scale
>with a color monitor, you can only define a 64-shade scale.

It seems to me that a similar thing could be done for a color VGA system. 
start with all three colors 0.  To fill suceeding registers, use the following
algorithm, cycling until all registers are complete:
	- increment R by 1
	- increment G by 1
	- increment B by 1, decrement G by 1
	- increment G by 1
Thus, you get colors like:
	- 00,00,00  01,00,00  01,01,00  01,00,01  01,01,01  02,01,01

The nice thing about using this is that the colors rendered would look like
a gray scale (they all approximate the vector running through the center of
color space).  The total number of colors is 4 * 2**6 = 2**8 = 256, thus all
registers are used.  The only 'bug' is in the one discontinuity where G is
being decremented, but 01,01,00 will at least be different than 01,00,01, 
and my suspicion is that the progression would look natural.

-- 
--------|	Then am I a happy fly, If I live or if I die.
Alien   |	                         - William Blake
--------|     jjmhome!cpoint!alien      bu-cs!mirror!frog!cpoint!alien