On Fri, 2009-02-06 at 11:48 +0100, Gabriel Paubert wrote: > > I've not switched on anti-aliasing. I think it is important the geometry > > has hard edges, and anti-aliasing will just increase memory > > requirements / reduce rendering speed. (IIRC, it just renders bigger > > than required, then filters the result down). > > I believe that you are confusing anti-aliasing and multisampling. > Multi-sampling > indeed multiplies memory consumption. Simple anti-aliasing does not, if I > understand correctly http://www.opengl.org/sdk/docs/man/ and the blurb > in the description of glLineWidth.
AIUI, multisampling is about the only way to have arbitrary geometry anti-aliased in GL, due to the way the rendering pipeline works. There have to be defined rules about which pixel a given piece of geometry will hit, in order that the z-buffering and re-assembly of different primitives works without any seams. (See cairo rendering if you butt to rectangles which are off-pixel grid, and you'll notice a seam anyway, due to the non-global anti-aliasing). There are of course, other things in GL, such as mipmapping (a kind of texture anti-aliasing), which are useful for avoiding true aliasing on textures. I'm not so convinced applying the terminology "anti-aliasing" is correct for jaggies on object edges. Its certainly not what I'd call aliasing on a digital oscillocope.. that matches the texture-sampling case. The edges are more a quantization artefact, and the low-pass filter used for "anti-aliasing" is in fact more like a reconstruction filter. -- Peter Clifton Electrical Engineering Division, Engineering Department, University of Cambridge, 9, JJ Thomson Avenue, Cambridge CB3 0FA Tel: +44 (0)7729 980173 - (No signal in the lab!) _______________________________________________ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user