Ricky Reusser wrote:
> Well, it's been a while so I hope this attachment works.  It's just an mpeg
> attached with KNode so we'll see.
> 
> Here's the beginning of an attempt at that Visual Simulation of Smoke paper
> (graphics.stanford.edu/~fedkiw/papers/stanford2001-01.pdf [I hope I got
> that right - I had to type it in from OS X -- if not, google it!]).  Each
> pixel represents one voxel, so that makes the simulation 100 x 160.  It has
> vorticity confinement, the semi-lagrangian scheme and a conjugate gradient
> solver.  That last part really isn't as bad as it looks.  Frankly, I only
> vaguely understand the concept, but the algorithm isn't hard to implement.
> 
> I still have to do the cubic interpolation, and (of course) add another
> dimension, but so far it at least looks somewhat like smoke.  Oh and I
> suppose incomplete cholesky preconditioning would be a plus too.  The mpeg
> compression made it darker than it should be so the eddies lost a lot of
> detail, but it isn't too bad.  Once I get things working better, I'll post
> the c++ code, but I warn you in advance that it's pretty ugly coming from
> someone who's just proud they figured out how to work objects.
> 
> It's sad this is how I spend my summer.  Just a parting thought.
> 
>  - Ricky

wow, i thought i'd left my cigarette in the monitor
it needs a red glow and some ash

it would be superb in a cafe scene

and i was hoping it was in pov sdl, then i read the text

now rather sad

stephen

Post a reply to this message