cKleinhuis


« on: December 17, 2009, 03:13:58 AM » 

i have entered the rather simple burning ship formula it is the burning buld fractal at 2 exponent ! front and rear view rendered with pixelbender skript on gpu


« Last Edit: December 17, 2009, 08:45:24 PM by Trifox »

Logged


divide and conquer  iterate and rule  chaos is No random!



kram1032


« Reply #1 on: December 17, 2009, 02:07:04 PM » 

looks way more interesting than the 2Dversion



Logged




matsoljare


« Reply #2 on: December 17, 2009, 05:50:26 PM » 

The 2D burning ship is very interesting, when you start to zoom in on it...



Logged




bugman


« Reply #3 on: December 17, 2009, 06:14:17 PM » 

Very neat! I'm not clear what do you mean by abs(z)? For the complex case it should be: x_{new}+iy_{new} = x²y² + 2ixy  z_{c}


« Last Edit: December 17, 2009, 06:17:09 PM by bugman »

Logged




kram1032


« Reply #4 on: December 17, 2009, 06:15:33 PM » 

abs z is clear. it's just the radius with phi=0° and theta=0° just compare polar coordinate complex abs



Logged




gaston3d
Guest


« Reply #5 on: December 17, 2009, 06:17:18 PM » 

what algebra is it? i've got different results with quaternions: z[n+1] = (abs(z.a) + abs(z.b)*i + abs(z.c)*j + abs(z.d)*k)^2 + h



Logged




bugman


« Reply #6 on: December 17, 2009, 06:19:50 PM » 

abs z is clear. it's just the radius with phi=0° and theta=0° just compare polar coordinate complex abs If so, then I don't think this is the correct formula for the Burning Ship, but it's still interesting.


« Last Edit: December 17, 2009, 06:35:34 PM by bugman »

Logged




kram1032


« Reply #7 on: December 17, 2009, 06:46:08 PM » 

ah, wait, yeah, forgot... abs(real)+abs(imag) for complex... well... convert from spherical to cartesian and then abs(x) abs(y) abs(z)



Logged




cKleinhuis


« Reply #8 on: December 17, 2009, 08:44:46 PM » 

and it is the same for more dimensions sorry for confusion i have corrected the above notation @gaston3d the used algebra is polar coordinate triplex, d.white & p.nylanders version, as defined on: http://www.fractalforums.com/theory/triplexalgebra/i will go and try to find more simple base fractals, was experimenting with barnsley. until i understand how to modify the derivations for new functions i will wait for trying newtonian, or diverging fractals it is very enjoying seeing that thing morph in realtime on my gts250 gpu, i really hope there will be a numbers library with arbitrary precision for gpus soon to work with greater precision


« Last Edit: December 17, 2009, 08:49:54 PM by Trifox »

Logged


divide and conquer  iterate and rule  chaos is No random!



BradC


« Reply #9 on: December 17, 2009, 08:47:32 PM » 

Do you mean (x, y, z) = (x, y, z) ?



Logged




cKleinhuis


« Reply #10 on: December 17, 2009, 08:50:11 PM » 

yes



Logged


divide and conquer  iterate and rule  chaos is No random!



gaston3d
Guest


« Reply #11 on: December 17, 2009, 09:10:16 PM » 

i think formula z[n+1] = z[n]^2 + c is still improper and is equivalent to z[n+1] = z[n]^2 + c
one of definitions of absolute value is: abs(a+bi+...) = sqrt(a^2+b^2+...), then (abs(z))^2 = z^2



Logged




gaston3d
Guest


« Reply #12 on: December 17, 2009, 09:28:42 PM » 

... i really hope there will be a numbers library with arbitrary precision for gpus soon to work with greater precision ...
i am not into gpu programing, but read somewhere that 64 bit double precision were introduced in shader model 5 (directx11) and is supported by nvidia gt200 chipset



Logged




bugman


« Reply #13 on: December 17, 2009, 09:48:56 PM » 

Do you mean (x, y, z) = (x, y, z) ?
You might also want to try {x, y, z}, as that seems to give interesting results as well, and the cross section in the xy plane still contains the 2D Burning Ship fractal.



Logged




cKleinhuis


« Reply #14 on: December 17, 2009, 10:15:07 PM » 

Do you mean (x, y, z) = (x, y, z) ?
You might also want to try {x, y, z}, as that seems to give interesting results as well, and the cross section in the xy plane still contains the 2D Burning Ship fractal. yes, that is what i did, i think it is nice to have possible 3d analogons of existing fractals using the triplex algebra, i am right now into trying several formulas, which do not need additional functions ( sin,cos ... ) btw. having the original 2d complex numbers fractal in one plane is a must ...



Logged


divide and conquer  iterate and rule  chaos is No random!



