Archive: 3d fractal


5th September 2005 10:14 UTC

3d fractal
i made this when i was just messing around with an earlier preset..changed quite a lot and then i just decided to try something random and it drew a freaking serpinski triangle!! for all those who don't know what a serpinski triangle is:
http://www.primepuzzles.net/images/S...i_triangle.png
that's it...anyway...i need your help on the maths side of things...right now it runs a decent speed, but ia m sure i have filled it with useless code and other crap...so give me your opinion/help! and of course tell me what you think! i know it's not that flashy to look at ...but at the moment it's still just a single scope so... :P


5th September 2005 10:36 UTC

In my opinion: Use the triangle ape for something thats going to be optimized, i tryed using the superscope and it jumped up about 3-6 fps, i don't really recommend using the texer2 because it's slow, and it would even faster using the triangle ape (mainly because it is lots of triangles) and may get 20+ fps...
:)


5th September 2005 10:44 UTC

if i knew how to use the triangle ape...maybe..but thanks for your input :D i'll see what i can do..if anyone wants to try and optimize it..go for it!


5th September 2005 10:56 UTC

x1,y1
x2,y2
x3,y3
= your six main vars for positions of the triangle try reading the last part of PAK-9's AVS Programming Guide; it has a chapter on triangle ape.


5th September 2005 11:25 UTC

cheers..still doesn't answer much but i'll take a look at his guide


5th September 2005 13:10 UTC

x1=asin(cos(i*$PI*2))*0.7;
y1=asin(sin(i*$PI*120))*0.7;

That is a nasty rawk h4x for making a solid sheet. Whats more it takes 4 (count em) trig functions per point ( = 40,000 trig functions just to make ur solid sheet) which will slow it down a lot. You should also precalculate your rotation sines and cosines since they dont change per point (thats another 120,000 trig functions you can avoid)

If you want to make a solid sheet the most logical method is to define a row, then define your sheet as a series of those rows. I've included a preset as an example but it needs AVSTrans


(P.S This isnt a 3D fractal its a 2D fractal projected onto a plane, a 3D sierpinski sleve would be a tetrahedron filled with recursively smaller tetrahedrons)


5th September 2005 13:51 UTC

thanks for your help with the maths pak


5th September 2005 14:01 UTC

Forgot the file, dur :igor: