Guide to the Inner Workings of the Mandelbrot-Set 1

[youhn]

No, but I did start to use "Find Minibrot" to skip some manual clicks. But depending on location, this does not always work. Sometimes the software thinks it find a center, and happy goes zooming further ... but it's not on the actual center. I will pay attention next time, to remember the type of locations where this happens.

[PieMan597]

I find that the auto zoomer fails when the spirals on screen have a higher iteration count than the "center" you want. I think that the  zoomer just zooms into the highest iteration.

[youhn]

Well, in that case I think the Mandelbrot-center-finder is not compatible with my current zooming method for Julia. I like high contrast between dense and sparse areas. Things like this:



This is just the half of the two-fold symmetry. The actual center is the small circle on the bottom-middle. With the Mandelbrot-finding function it zooms in on the leftmost dense spot.

[youhn]

These past few weeks the focus was on creating low level julias. Each time I more or less followed the following steps:

- Zoom into the Mset just a little to find a minibrot
- Zoom into a valley of choice to get interesting spirals (these lead to nice bounded shapes deeper on)
- Find a Julia without zooming too deep
- Start Julia morphing, each time on the new two-fold shape
- Never go to far away from the center for each new two-fold to morph

All raw fractals found and rendered with Kalles Fraktaler 2.7.9 at something like 7200px wide.

[quaz0r]

nice work 

[Kalles Fraktaler]

 

Did you use "Show inflection" at all?

[youhn]

No, I more like to plan some kind of method or principle and dive in using that in mind. I think show inflection can't show multiple morph steps ahead, so that's why I don't use it.

[Kalles Fraktaler]

The auto zoomer does try to find the center by mirroring parts of the view and select the position where it gets the lowest error when comparing iteration values per pixel.
However in dense areas there can be a huge difference between pixels, that destroys this method. Maybe comparing rgb values in the image would be more reliable also for dense areas

[jehovajah]

http://m.youtube.com/watch?v=9gk_8mQuerg
Nice introduction , but not sure he correctly ascribes the origination of Bhuddabrot .