Recent Posts

Pages: [1] 2 3 ... 10
Image Threads / Re: Other Polynomial and Rational Maps
« Last post by pauldelbrot on Today at 04:27:52 PM »
Herman Ring

A Herman Ring. Found in the Supernova formula and rendered using DE. The linked black rings are the Herman ring together with its preimages. White is the Julia set. The exterior black region and all simply-connected black regions are parts of A(∞) and A(0). 0 is in the interior of, and near the center of, the lower of the largest two simply-connected black regions.
Downloads / Re: Infinity Fractal v1.0.13
« Last post by gerson on Today at 03:58:57 PM »
@ mrrudewords that problem occurs when one use the zoom function. Zoom the image and after try to change the size of the zoom square using mousewheel in sequence, without doing other action. If you try to do other thing it let the square be changed again (just turn to other program, try set new power or iteration or coloring mode...).
I suggest you let:
"control s" to "save as" option -> save image and parameter at the same time with same new name
and control shift p to save only parameter
and control shift s to save only image (don"t need to ask to save parameter here and it suggest the actual name as all software does).
This let more simple to user understand.
I noticed that sometimes it didn's suggest new name with control s and sometimes it create a image without extension on program directory.
The julia function has strange behavior. I choose a zoom image in mandelbrot and set to julia mode. It set julia box to -0,576374514116629,-0,631165823016141. So I try to change the value (for example to -0,5763745141166,-0,631165823016141) just delete two last digits from real part and try to change bailout so program advice that will full recalculate and give a black image.
I think the julia value is in conflict betwen sqare soom position and julia box value so it crash.
Suggest to let user change coordinates. So it would have 3 box, x, y and scale (that would be used to x and y). Is necessary to take care to don't be conflict betwen this box and zoom square too.
One option to clean code is a run/stop (control R/control Q) botton or to set every box proprieties to run image when mouse out (I dont know the right term) or "enter" occurs.
Image Threads / Re: Twofold
« Last post by pauldelbrot on Today at 02:37:19 PM »
Downloads / Re: Infinity Fractal v1.0.13
« Last post by mrrudewords on Today at 10:26:51 AM »
@Sabine62 @gerson Unfortunately I am not seeing this issue myself, which makes it rather difficult to fix. Are you able to identify when it happens, i.e. if it ocurrs after you do something specific? If you can give me some steps to try to replicate the issue I should be able to sort it out.

Note: I've noticed a few new bugs I've managed to put in:

If you click the colouring gradient length slider while it's already working, it seems to double up on recolouring and really thrashes the CPU. I am fixing this by disabling all controls while the app is busy.

Also, there's a strange issue where it tries to save images with a name that doesn't follow the format, so I'm looking into that too.
@marcm200 Logo-turtles! Yes! Got it for my daughter in those days... good gracious, I feel so old now (for a sec or two!)  :)) :)) :))

Thanks for also posting the 3D-view of the fitness-conditions, I always find it easier to understand something a bit better when I can visualize it  :thumbs:

Downloads / Re: Infinity Fractal v1.0.13
« Last post by Sabine62 on Today at 09:26:26 AM »
@gerson Glad I am not the only one with that issue  :D Your tip is great, until now I had only found that if I change resolution it would work again.
Thank you! :)  :thumbs:
@gerson: That's a good idea. I could save the random generator seed (time stamp) to recreate the string of numbers to use in other formulas. But I hope there's not a dependence of the outcome to the actual sequence of "random" numbers, that would be a downfall to implementing genetic algorithms in my system.

Designing some other fitness functions I found three that gave quite a distinct and consistent result. The images show the best individuals of 15 different runs of the program with 10^5 mating events and a population size of 800 individuals. Black is interior, the lighter the gray the faster the escape happens.

The upper left depicts a 3D view of the fitness function (the bar height of the 11x11 grid
represents the iteration count the fitness functions desires at that point).

The W-function finds mostly valleys, the slope-function finds smaller bulbs on the border of larger ones and the EcoRI-function finds highly branching structures.

What I learnt so far, is, that having big plateaus in the fitness function (e.g. a constant function) finds almost exclusively interior, therefore new grids will have a large difference between the smallest and the greatest grid height.

Hi Matt!

I think I have the source for all, 1.0.19 - 1.0.31, what do you need? I'll put it on my website and pm the link  :thumbs:
Animation of Mandelbrot foam (see
w \leftarrow q w/z\\
z \leftarrow z^2 + w^2 +c
with q rotating around a circle of radius 1/2.
Video should work for Chrome users, others may need to download it.

Click for video

Meet & Greet / Re: I'm back!!
« Last post by chronologicaldot on Today at 02:38:49 AM »
@WAUthethird -  I did manage to find the server, yes, but thank you for sharing the info. :) I didn't know there were two. I guess I found the active one because people were using it.
Pages: [1] 2 3 ... 10