I came across Mark’s (?) post from a while back about using a “pretend” resolution much higher than your actual number of pixels and antialiasing by only lighting up the “fractional” pixels by a fractional amount. It’s a pretty cool technique.
Am I right in thinking this would work just as well in 2 dimensions (with a pixel matrix) as it does in one (on a linear strip)? For my current project I’m developing a generalised “sprite” animation system for a biggish 2D display (32 x 18) and I was considering using matrix transforms as a basis for movement and rotation—but rotation wouldn’t look so good without antialiasing. This seems like it could fix some of my problems.
Has anyone used matrix maths to do something like this before? Does it work well?