Lance Birch wrote:
> 
> Now back to on topic stuff here... I was looking through your includes
> libraries etc etc and noticed something about particles systems.  I was
> thinking... have you thought about making a particle media system based on
> combining the particle systems you currently have with spheres of media for
> the particles?  I talked to Peter Popov about this a while ago and
> unfortunately haven't had the time to get the spcifications to him (sorry
> Peter), but MAX has a really good system based on this... My question is, do
> you think this would be too slow in POV-Ray?

to which Peter Popov replied:
> 
> It could be rendered really fast, in fact, if POV's media didn't mess
> up with intersecting container objects that much. I'll try to make a
> speed (and quality) comparison test using 1000 media in one container
> vs. 1000 separate containers... guess which one'll be faster. If the
> latter works, the particle system might as well be done. BTW this is
> one of the major items on my to-do list for this vacation (after July
> 15), as long as a .df3 creating utility and a fractal patch for POV
> (anyone working on any of those?)
> 
> And Chris, I have had some thoughts about this particle system that I
> haven't discussed with Lance (sorry, Lance). If you consider the
> project worthy we might as well share some ideas.

My plans for the POV-Ray Particle System certainly include media-related
features.  In fact, it is possible to create media-filled particles
using the Liquid Spray include file (as others have already done for
volcanoes, etc.), but while blob containers aren't such a problem they
do not allow for blending individual interiors/media.

One solution, then, might be to create a single blob container which is
filled with multiple media (the centres of the media being located at
the centres of the blob components).  This might improve sampling
results, as the container will be more closely fitted to the size of the
media, and of course you don't have the worry of excessive
max_trace_level's.  Another option might be to exploit the density
files  mentioned - it should be quite possible to output a *.df3 file
based on particle positions/densities, etc., and then (in the same
scene) read this in to create the media.

Of course, if either you, Lance, or you, Peter, (or indeed you,
as-yet-unnamed reader!) wish to regale me with particle system ideas I
am contactable at mailto:cco### [at] geocitiescom (although, No, I haven't
made up my mind about the Yahoo TOS...).  As I may have mentioned it's
not always possible for me to gain access to news.povray.org, so a reply
may be quicker forthcoming if the message is also directed more squarely
in my direction.

Post a reply to this message