Newsgroups: comp.os.msdos.djgpp
From: tob AT world DOT std DOT com
Subject: Re: Allegro loads graphics funny...
Message-ID: <EK6tBr.7Ju@world.std.com>
Sender: tob AT world DOT std DOT com (Tom Breton)
Reply-To: tob AT world DOT std DOT com
Organization: BREnterprises
References: <m3pvnrccoo DOT fsf AT H-Bukin DOT inp DOT nsk DOT su>
Date: Tue, 25 Nov 1997 05:45:27 GMT
Lines: 40
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Precedence: bulk

Michael Bukin <bukinm AT inp DOT nsk DOT su> writes:
> This scheme is not symmetrical.  It will be more symmetrical
> if you take central pixel from each box as a value for target pixel.

Right. The heart of the Allegro routine for big->little goes like so:

	 else {                           /* little -> big scaling */
	    x2 = sx >> 16;
	    COMPILER_LODSB();
	    for (x=0; x<dest_width; x++) {
	       if (masked) {
		  COMPILER_MASKED_STOSB();
	       } 
	       else {
		  COMPILER_STOSB();
	       }
         sx += sxd;
	       if ((sx >> 16) > x2) {
		  COMPILER_LODSB();
		  x2++;
	       }
	    }
	 }

...but to be symmetrical it should start off with an half-increment...

  sx += sxd/2;

...which requires turning the loop around a bit so that the
test for compensation...

	       if ((sx >> 16) > x2) {
		  COMPILER_LODSB();
		  x2++;
	       }

...comes first. The 15, 16, 24, and 32 are parallel to this AFAICS.

        Tom