Xref: utzoo comp.windows.x:35751 comp.graphics:17519
Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!island!hue
From: hue@island.COM (Jon "Beer - It's not just for breakfast anymore" Hue)
Newsgroups: comp.windows.x,comp.graphics
Subject: Re: changing color intensity in X
Message-ID: <3963@island.COM>
Date: 25 Apr 91 19:09:13 GMT
References: <1991Apr20.204325.2078@eng.ufl.edu> <728@modulex.dk>
Sender: usenet@island.COM
Followup-To: comp.windows.x
Organization: Island Graphics Corp., San Rafael, CA
Lines: 27

In article <728@modulex.dk> ssn@modulex.dk (Svend Skafte Nielsen) writes:
>   Intensity (also called lightness or luminance) is usually calculated
>   as
>
>         I = (R+G+B)/3

Oh really?  I've never done it like that.  Say, does anybody else do it like
that?

When I know the chromaticity coordinates and white point of the RGB device, I
plug them into a program which spits out the RGB->XYZ matrix coefficients and
use the 'Y' line.

[284]coney:hue: ./calc
Enter red x and y: 0.625 0.340
Enter green x and y: 0.280 0.595
Enter blue x and y: 0.155 0.070
Enter white x and y: 0.283 0.298

X = 0.38106154R + 0.32025711G + 0.24834578B
Y = 0.20729748R + 0.68054637G + 0.11215616B
Z = 0.02133945R + 0.14297193G + 1.24172890B

When I don't, I cheat and use Y = .30R + .59G + .11B, even though it doesn't
correspond to any device you'll likely run across.

-Jonathan		hue@island.COM