fractalrebel
|
|
« Reply #30 on: January 04, 2013, 02:29:16 AM » |
|
Some of the images look really promising. I hope to see if I can get it working in UltraFractal using my 3D Fractal Raytrace module. Here is a Mandelbox example using the same module:
|
|
|
Logged
|
|
|
|
LMarkoya
Strange Attractor
Posts: 282
|
|
« Reply #31 on: January 04, 2013, 07:10:58 PM » |
|
SOme of the later images are really beautiful.....nice work all
|
|
|
Logged
|
|
|
|
fractalrebel
|
|
« Reply #32 on: January 07, 2013, 12:55:21 AM » |
|
I have been playing around with the really cool Mandelex discovered by Hiato. I have implemented it as 3D using 3D Fractal Raytrace in Ultrafractal. Here is an image that I call Slums.
|
|
|
|
cKleinhuis
|
|
« Reply #33 on: January 07, 2013, 01:12:10 AM » |
|
for examining slices of 3d formulas ultrafractal is perfectly suited, just implement z coordinate as param nice pic, and congrats to hiato for this!
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
DarkBeam
Global Moderator
Fractal Senior
Posts: 2512
Fragments of the fractal -like the tip of it
|
|
« Reply #34 on: January 07, 2013, 10:35:44 AM » |
|
Hey ... In this case he used it as a raytracer. The problem is that uf raytraces veeery slowly. MB3d is like 20x or 50x faster
|
|
|
Logged
|
No sweat, guardian of wisdom!
|
|
|
cKleinhuis
|
|
« Reply #35 on: January 07, 2013, 12:25:49 PM » |
|
yes, i meant just for nicely rendering the slices, as 3d renderer it is practically impracticable
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
DarkBeam
Global Moderator
Fractal Senior
Posts: 2512
Fragments of the fractal -like the tip of it
|
|
« Reply #36 on: January 07, 2013, 01:28:30 PM » |
|
|
|
|
Logged
|
No sweat, guardian of wisdom!
|
|
|
fractalrebel
|
|
« Reply #37 on: January 07, 2013, 06:43:30 PM » |
|
Lets clarify things here. I am using a full 3D raytracer written for Ultrafractal, not slices or any such thing. The algorithm is based upon an article by Hart, Sandin and Kauffman for estimating the distance from a point to a fractal surface. The raytracer includes camera positioning, multiple lights, shadows, transparency, surface normals, rotations, etc. Distance estimates from the camera to the surface can be made with a number of potential types which require the derivative of the fractal function. Two methods which don't require an explicit derivative are delta-DE and brute force. The brute force method is VERY SLOW and I rarely use it. It can slice a 3D object, but that was a later add-on. It may not be totally full featured but it gets the job done. Its main drawback is that it is slow .
|
|
|
Logged
|
|
|
|
fractalrebel
|
|
« Reply #38 on: January 07, 2013, 06:58:53 PM » |
|
DarkBeam: I ignored them, using the functions shown within the "formula" function. I am not too sure of the implementation of nonLinear pull based upon what is inside the "formula" function, so I am still experimenting. Obviously I am using a 3D version, which is a plugin for 3D Fractal Raytrace in UF.
|
|
|
Logged
|
|
|
|
fractalrebel
|
|
« Reply #39 on: January 07, 2013, 08:54:41 PM » |
|
There have been questions raised in this thread about whether it is possible to do raytracing with Ultrafractal. Since the 3D Mandelex requires a raytracer and I am experimenting using a Ultrafractal raytracer, I don't think it is off topic to answer in this thread. Here is an example that uses reflection, refraction with a user-defined refractive index and shadows. This was rendered completely and totally within Ultrafractal.
|
|
|
Logged
|
|
|
|
David Makin
|
|
« Reply #40 on: January 07, 2013, 09:29:02 PM » |
|
Hey ... In this case he used it as a raytracer. The problem is that uf raytraces veeery slowly. MB3d is like 20x or 50x faster Erm ? Really ? My Wip3D formula renders a degree 8 Mandelbulb (complete unzoomed one in a 928*626 window) in just 6.7 secs - this is with smooth iteration colouring standard gourad+Phong lighting including 2 light sources - camera without shadows or specular, plus positional with specular and hard shadows - including "star"-fix and positional ambient soft-shadowing. Or alternatively my escape-time 3D IFS formula ray-traces a Menger sponge with similar lighting and at the same resolution in 0.86secs. Or the Wip3D formula again at same res. and lighting etc. renders the original Lowe Mandelbox in 7.23secs. The image below, exactly as seen here took just 11.69secs: http://www.fractalforums.com/index.php?action=gallery;sa=view;id=13133Colouring used is (2D) triangle inequality average.
|
|
« Last Edit: January 07, 2013, 09:30:34 PM by David Makin »
|
Logged
|
|
|
|
cKleinhuis
|
|
« Reply #41 on: January 07, 2013, 09:35:42 PM » |
|
ok, people, i meant ultrafractal is best suited for just rendering the slices as plain 2d, because it is a very sophisticated animation/editing/exploring tool with an incredible precision, the hacks with 3d rendering are possible, i know, but i am as well aware that the program is not really nice for doing such, the biggest problem is that something that is done in the init/global part can not be succsesive be seen, most of the 3d renderers have black screen for ages and then sometime the result is there, but this makes it very hard to conveniently edit something.... i would love to see more images of the mandelX, this thread seems to become hijacked
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
David Makin
|
|
« Reply #42 on: January 07, 2013, 09:47:57 PM » |
|
ok, people, i meant ultrafractal is best suited for just rendering the slices as plain 2d, because it is a very sophisticated animation/editing/exploring tool with an incredible precision, the hacks with 3d rendering are possible, i know, but i am as well aware that the program is not really nice for doing such, the biggest problem is that something that is done in the init/global part can not be succsesive be seen, most of the 3d renderers have black screen for ages and then sometime the result is there, but this makes it very hard to conveniently edit something.... i would love to see more images of the mandelX, this thread seems to become hijacked Apologies, but to say that UF is 20* slower (or worse) than MB3D or MBer is just plain wrong Also MMFWip3D and my 3D IFS do not use the global section so you *do* get nice progressive renders. Am going to implement the Mandelex in my Wip3D formula shortly - plus release an update that makes it easy for anyone to add any test 3D/4D formula very quickly - in fact for testing it should be the quickest way of testing a new formula for anyone.
|
|
|
Logged
|
|
|
|
cKleinhuis
|
|
« Reply #43 on: January 07, 2013, 10:06:21 PM » |
|
bring it on are your formulas available via library update !? why dont you use the library ?!
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
David Makin
|
|
« Reply #44 on: January 08, 2013, 04:31:47 AM » |
|
bring it on are your formulas available via library update !? why dont you use the library ?! Because I usually only add "finished" formulas to the UF formula database - i.e. those I don't plan on making any major changes to and the Wip3D formula is far, far from "finished"
|
|
|
Logged
|
|
|
|
|