Sony recommends that large polygons be subdivided, often as part of the rendering pipeline rather than in advance, so as better to behave with affine texture mapping, per-primitive clipping and the texture and frame buffer caches.
However, it looks like the GPU takes only integral, 8-bit texture coordinates. So you can't subdivide arbitrarily without creating visible seams in your texturing.
E.g. if you have a polygon edge that runs along texture coordinates from 0 to 17, you can't split it a third of the way along without changing the overall texture distribution.
Is that true? If so, would Sony's response way back when just have been "it'll still look better than not subdividing, so don't fret about it"? Or do I have a failure of comprehension?
Subdivision and texture coordinates
I guess that's probably what Sony responded, yes. And also the reason why a number of games went for using one of the higher resolutions and a lot of gouraud shaded polygons rather than textured ones, like the Crash Bandicoot and Spyro series.
Who is online
Users browsing this forum: No registered users and 1 guest