View Single Post
Old 11-24-2004, 11:39 AM   #33
pruss
Evangelist
pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.pruss ought to be getting tired of karma fortunes by now.
 
Posts: 461
Karma: 819417
Join Date: Nov 2004
Quote:
Originally Posted by Maceo
I think sub-pixel rendering is quite different from anti-aliasing. Unfortunately, you cannot show the effects on a CRT screen, where both techniques look similar (yes, I find this kinda confusing, too
Subpixel rendering would be a big nuisance to implement on PDAs given rotatable or flippable screens. One would need to have pre-generated bitmaps of fonts for all four possible orientations, or else have a truetype renderer on the PDA.

Mobipocket has subpixel fonts but doesn't check for Clie NX flip status. This means that on a Clie NX, in one flip position you get a display that's better than grayscale antialiasing and in another you get a display that's worse than grayscale antialiasing.

Grayscale antialiasing doesn't have any of these problems, and doesn't also have the problem that subpixel rendering might infringe on Microsoft patents. It doesn't look quite as good perhaps (though if one doesn't like the slight colored fringes that ClearType puts up, one will dispute this) but it is a lot easier to implement. And it does look good--see Plucker 1.8.
pruss is offline