CompuServe Messages

#sxp variables

    13-Sep-95 21:32:48
Sb: #190539-#sxp variables
Fm: Martin Doudoroff/KUB 74664,2144
To: Larry Minton 74010,3076
>> Could you provide me with a little description of variable SXPDPX? "Surface delta x" is a bit vague <g>. << Weeeeeelll, you're asking a lot here, you know? Actually — as I'm sure you'll be perturbed to hear — we don't entirely understand, ourselves. Here's what the IPAS Toolkit says about them: "The variation of position used for anti-aliasing. These values represent the width, height, and depth of the 3D bounding box of the region that the current rendering pixel represents. The bounding box is centered at the point specified by the parameter[s SXPX, SXPY, SXPZ]." p.38 Anti-aliasing in the context of a scanline renderer is essentially a question of SUB-PIXEL rendering. You need to end up with a single value for the pixel, but for the pattern to be smooth across the surface (hence, not pixelated), you must account for any potential shading variation across the area the pixel represents: an average. The values presented in these variables should give you a range of 3D values which your SXP script must calculate in order to arrive at this average. For a coded example, see Tom Hudson's STRIPE.S3D. It's kinda scary, but I've heard that if you knead your cerebrum slowly and evenly over long stretches of time, many things can be comprehended. Martin