> >I suspect a proper implementation of this would go a long way toward
> >making the B/W vector emulations look better. It doesn't address the
> >trails you leave, but I suspect carefull use of alpha buffers could
> >help (but 8 bits of alpha won't allow much precision in the decay).
> >Hmmm, dump the last N frames into separate buffers and blend according
> >to predefined formula... Hmmm tons of video RAM, but it could be an
> >option for people with SGIs.
>
> Maybe. Or just a single frame buffer whose contents "decay" to black
> over time with a simple pixel op (e.g. start with the last frame's image in
> the frame buffer, reduce it by doing (say) a 30% blend with black, then draw
> the current frame - this would cause lines to fade out over the subsequent 3
> frames). This would seem to more closely model the behavior of real hardware,
> without requiring multiple frame buffers, and I think most 3D cards will do
> a single blend operation quite quickly).
A fixed percentage each frame was my first thought, but that gives a nice
exponential decay and I'm not so sure that's the right characteristic...
Also, to do it right you should have more than 8 bits per pixel. Remember
you want *seconds* of persistence on Asteroids - that's a lot of frames to
persist through :-) The main thing I wanted to get someone working on was
texture mapping the "glow" around the vectors. Now that I think of it, that
could also be used to "defocus" the beam for those fuzzy Star Wars vectors.
There I go thinkin' again...
BTW, finding an old high persistence monitor is not my idea of the
correct solution. I want 1600x1200 with the texture mapping AND decay, that
should look *better* than the real thing - except for the brightness of the
bullets... Just can't win can you?
Hey, this is the most vectory stuff I've posted in a while :-)
-- ___ __ _ _ _ | \ / \ | | | || | phkahler@oakland.edu Engineer/Programmer | _/| || || |_| || |__ " What makes someone care so much? |_| |_||_| \___/ |____) for things another man can just ignore. " -S.H.Received on Tue Oct 5 08:04:38 1999
This archive was generated by hypermail 2.1.8 : Fri Aug 01 2003 - 00:32:44 EDT