Toroidal Symmetry IFS animation

[doncasteel8587]

If the animation doesn't show up, please view the deviation here: http://www.deviantart.com/deviation/43928839/

You'll have to click the download icon to see the animated gif.

[bradorpoints]

That is really impressive.  Have you considered writing a plug-in for Maya, Lightwave - or the big players?  Something like this could add new life to the current particle emitters of the beforementioned applications.

[doncasteel8587]

Have you considered writing a plug-in for Maya, Lightwave - or the big players? 

I'd love to work with Maya/MentalRay, but my experiance with mel is that I can't get to the things I'd need to get to, i.e. the actual raytracing. Or at least I'd need some help from an expert.

It is my ultimate goal to someday hook up with them and get it to work.

[lycium]

Or at least I'd need some help from an expert.

an expert ray tracer or expert maya/lightwave user?

edit:

thinking about it a little more, integrating it into lightwave/maya/mental ray etc is probably impossible due to the high level of system integration that would be necessary (most likely at the source code level). for example, if you have other objects inside the density volume that would need to be taken into account somehow, and you'd need to add photon mapping hooks etc. on the other hand, what you're basically dealing with is a grid of densities, so if you're willing to forgo your own rendering methods i'm pretty sure there are tools for importing volume densities into those applications - this brings up issues like how animation will be handled, and there i think is where the major scripting will lie.

of course, this discussion will probably be rendered moot if you follow the gpu rendering line of attack i suggested on deviantart

another option would be to write a basic photon mapping renderer that handles inhomogeneous participating media (as the technical term goes). i've got jensen's book on the subject (http://graphics.ucsd.edu/~henrik/papers/book/) and know it well, so i could help you out there so long as the project isn't commercial. something like this isn't too difficult, maybe a week of work: http://graphics.ucsd.edu/~henrik/images/parti.html

[doncasteel8587]

I've often thought about someday adding global illumination and caustics, but I think that is still a long ways off.

I think as time permits getting into the GPU is the first path to pursue, which leads me to the question, can I program and run Nividia cg applications without a Nivida graphics card? It will be a while until I'll be able to afford one. I was just wondering if I can start learning the language before having the card?

[lycium]

global illumination isn't really such a big deal, but i agree, one step at a time...

as for getting an nvidia gpu, i bet you can find some manner of gf6 for like $50, perhaps less. even something poopy is fine so long as it's a geforce6 or newer, they're all equally programmable; a pit stop to develop your applications before you realise the 10x performance increase a $150 card will provide. unfortunately, you really need some manner of shader model 3 (geforce 6 or newer) hardware to get into this kind of rendering, since it needs to be able to do loops etc.

really, geforce6 level hardware is totally abundant, almost ubiquitous. in 1st world america i have a hard time believing there isn't one within 100m of where you are right now! i've mentioned before that i have a spare geforce 6600 non-gt (about as poopy as it gets, but still 8-way parallel and highly programmable), which i could offer but again you'd pay more on shipping and customs than you'd pay to get one locally. i just checked ebay (my first time) and there are 256mb models selling for $20! skip dinner one night maybe?

edit: hell, i didn't eat for 4 days after having all my wisdom teeth removed recently