From:	George Foot <mert0407 AT sable DOT ox DOT ac DOT uk>
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: Colors
Date:	11 Feb 1998 20:58:21 GMT
Organization:	Oxford University, England
Lines:	43
Message-ID:	<6bt3dd$42g$4@news.ox.ac.uk>
References:	<887209728 DOT 858627 AT nn1>
NNTP-Posting-Host:	sable.ox.ac.uk
Mime-Version:	1.0
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp

On Wed, 11 Feb 1998 16:11:15 +0100 in comp.os.msdos.djgpp Anders
Clerwall <anders DOT clerwall AT swipnet DOT se> wrote:

: Hi,, I have tried to convert rgb values to a color value in 800x600 (64k
: cols) using the formula: (r << 11) | (g << 5) | b; what's wrong?!.. when I
: put in for example the rgb values 60,60, 60.. the pixel on the screen is
: blue!.. it should be white, no?

I have never done anything in high colour depths, but your formula
implies that R is a 5-bit number, G is a 6-bit number and B is a 5-bit
number.  So, R ranges from 0 to 31, G from 0 to 63 and B from 0 to 31
also.  Note: (1) G has a different range to R and B so you'll need to
scale its values accordingly, and (2) the ranges aren't from 0 to 63
as for a 6-bit DAC in 256 colour mode.

So, make sure your R, G and B values are clipped to the correct
ranges.  You might like to express them all as 8-bit quantities then
scale them down.

/* untested code -- for illustration only */

#define MAKE_IN_RANGE(x,l,h) x = (x < (l)) ? (l) : ((x > (h)) ? (h) : x)

int get_colour (int r, int g, int b) {
	MAKE_IN_RANGE (r, 0, 255);
	MAKE_IN_RANGE (g, 0, 255);
	MAKE_IN_RANGE (b, 0, 255);
	r >>= 3;
	g >>= 2;
	b >>= 3;
	return (r << 11) + (g << 5) + b;
}

Then you can call `get_colour' with r, g and b strengths all ranging
from 0 to 255 (automatically restricted to that range) to get your
colour.  I believe this is something like what Allegro does.  Using
ranges from 0 to 255 makes it simple to write 15-bit and 24/32-bit
equivalent functions.

-- 
george DOT foot AT merton DOT oxford DOT ac DOT uk

ko tavla fo la lojban

- Raw text -