gg92507
Forums Newbie
Posts: 9
|
|
« on: March 28, 2013, 09:39:40 AM » |
|
Forum Readers-- Here is a 10 100 zoom of the Mandelbox. Mandelbox comes from applying a folding function repeatedly to (any) initial object. To illustrate this, we start with the Earth, and fold it up into a Mandelbox. Then we zoom in by a factor of 10 100. I think this is the biggest zoom so far of the Mandelbox. Please correct me if I am wrong.
http://www.youtube.com/v/rBCseNGI-UY&rel=1&fs=1&hd=1Go to 1:40 to skip right to the zoom. It took two weeks to render this on my Geforce GTS 450 and GTX 650m. Thanks for watching. --George
|
|
« Last Edit: March 28, 2013, 10:07:40 AM by cKleinhuis, Reason: modified youtube link to be directly displayed in post! »
|
Logged
|
|
|
|
cKleinhuis
|
|
« Reply #1 on: March 28, 2013, 10:11:51 AM » |
|
wow, well done with the earth-morph how did you do it ? using your own software ? the beginning is slow, but afterwards the camera is going a liiiitle too fast, nice deep zoom into the mandelbox scale -1.5
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
gg92507
Forums Newbie
Posts: 9
|
|
« Reply #2 on: March 28, 2013, 10:39:42 AM » |
|
Thanks. I agree--the beginning is slow. And the zoom is too fast. But a Googol is a long way to zoom--if I don't zoom quickly, it'll take too long.
|
|
|
Logged
|
|
|
|
Dinkydau
|
|
« Reply #3 on: March 28, 2013, 03:56:33 PM » |
|
The video doesn't play smooth for me and the audio is way out of sync (ends too early).
|
|
|
Logged
|
|
|
|
Jesse
Download Section
Fractal Schemer
Posts: 1013
|
|
« Reply #4 on: March 29, 2013, 12:32:03 AM » |
|
Very awesome, George!
I have not seen a kind of arbitrary precision on a GPU yet, so the only way i know to get something like this would be ultrafractal, but if this needed two weeks on GPU, UF might take forever...
Would be cool to know what libraries or whatever you used to make this work.
|
|
|
Logged
|
|
|
|
gg92507
Forums Newbie
Posts: 9
|
|
« Reply #5 on: March 29, 2013, 04:48:00 AM » |
|
Jesse--
I haven't seen ultrafractal--I'll have to look into it. I have two GPUs so it would take twice as long with just one. (Or half as long if I had one fast GPU.) Maybe ultrafractal is faster.
I use Java and OpenCL. It uses 416 bit precision at the end.
--George
|
|
|
Logged
|
|
|
|
schizo
Iterator
Posts: 156
|
|
« Reply #6 on: March 29, 2013, 08:49:53 PM » |
|
Wow. That was deeeeep. Really impressive rendering and a very original idea to fold the earth.
|
|
|
Logged
|
|
|
|
marius
Fractal Lover
Posts: 206
|
|
« Reply #7 on: March 30, 2013, 12:21:03 AM » |
|
I use Java and OpenCL. It uses 416 bit precision at the end.
What range of iteration counts were in play at the end?
|
|
|
Logged
|
|
|
|
gg92507
Forums Newbie
Posts: 9
|
|
« Reply #8 on: March 30, 2013, 12:50:55 AM » |
|
Marius--It only needed about 230 iterations at the end. It isn't much of a zoom by mandelbrot standards. I suppose if I zoomed somewhere else, it might require a different number of iterations. I have no idea where would be interesting, so I picked a random location.
|
|
|
Logged
|
|
|
|
KRAFTWERK
|
|
« Reply #9 on: April 11, 2013, 08:35:04 AM » |
|
Wow, didn't see this before, impressive zoom... do I spot some disconnected parts flying by??? Looks like a Scale -1.5 box With min R 0.5?
|
|
|
Logged
|
|
|
|
gg92507
Forums Newbie
Posts: 9
|
|
« Reply #10 on: April 12, 2013, 07:18:49 AM » |
|
Kraftwerk--You are right about the scale and min R and disconnected parts. Should it to be connected? Maybe there is an error in the program.
|
|
|
Logged
|
|
|
|
cKleinhuis
|
|
« Reply #11 on: April 12, 2013, 01:18:31 PM » |
|
@gg the rendering looked fine to me, no obvious error in the formula
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
|