Welcome to Fractal Forums

    Hi, I'm trying to make histogram coloring in my fractal generator, I've studied the descriptions on HDPZ.NET, and read over ievs "Newbie: How to map colors in the Mandelbrot set? " topic like a million times, to understand Duncans procedure, but it's still not working... I didn't rly understand the way he divided the iterations to multiple bands. And the way he updated the new colors.
If I'm using a 800x800 image, and as maximum iteration of 1000, i get the Histogram (the attached file).
I think it's correct how it is, but when i add them together as it should be, and make the percent tables, something isn't okay. The problem is with the numbers, so i think the procedure how it calculates the histogram is correct, and the values are correct)  at the fourth iteration I get the 0,793641115261387 value in the color table (79 percent)
, 5th =  0,181975744889239, 6th = 0,0755547660428449  etc. (that's also in the attached file, at the bottom), and then its dropping below 0,01... So it keeps skipping big parts from the color map... Any idea what's wrong  ????
Thx for the help!!!  :P

hmm, i do think that this distribution is what to expect, where do you store your FULL iteration values ?
those should be ~halve of the image when using it on the standard mandelbrot

Thanks for the answer!
 Full iteration values?  you mean iteration value for every single pixel? if yes, then see the attached file.
So, we both think the distribution is ok, then what am I missing? ???
I have a gradient editor what saves the gradient in a bmp file, it's length is 1000 pixels. So for every iteration I get the u  as  Histogram / number_ofpixels_outside_of_the_mandelbrot_set  * 1000, and this should give a value from 0 to 1000, and then it gets the color from the bmp, based on this. This works, however as I said in my first post, it simply skips the 99% of the color-range... =[

so, it has a logarithmic behaviour, you can smooth it out, or linearize it by taking the log ( or inverse ) of your calculated values, because
it is very likely that you hit the lower iterations very often, and some of the higher pop up on a low "chaoutical" base, because sometimes
you hit a deeper one, but only one time, and thus it gets a low histogram value

can you post the link to the histogram tutorial ?

http://www.fractalforums.com/programming/newbie-how-to-map-colors-in-the-mandelbrot-set/ (http://www.fractalforums.com/programming/newbie-how-to-map-colors-in-the-mandelbrot-set/) this is ievs post about the same topic, however I had hard time understanding Duncan Cs explanations at some critical parts.

http://www.hpdz.net/TechInfo_Colorizing.htm (http://www.hpdz.net/TechInfo_Colorizing.htm) This is a pretty good tutorial with diagrams and formulas, showing different techniques, why the easier ones not good enough, but I think it should be more detailed at the Histogram and Rank-order mapping. Also I'm not sure how to use the Fractional Counts (see bottom of that page), what values should I give to F and W?

If I take the log of that u, Im getting a bunch of negative numbers. I know, I know, I'm a real noob  :(, but I dont know how to interpret them.
Thanks for your patience  ^-^

It just doesnt help to smooth it out, cos no matter how I manipulate those numbers, the difference will be way too big between those iterations. I know I'm missing some critical part, there must be a trick to use that histogram, but i dont know what that is  :'(

can any1 help me pls  ???

I haven't checked this part of the forum in quite a while, so I just found your post. I'll take a stab at a better explanation over the next couple of days.


Duncan

  Hello Duncan,

  Thanks a lot for your support! I've already made up a solution myself, and sent my project to the competition's site, and theoretically today is going to be announced who are going to the finals. If you would like to see what have I done, I'd love to show it; and I may put up some questions if your time permits. If you have any messenger, it would be much easier to talk there.
  I'm waiting for your answer! have a nice day!

  Tamás

Tamás,

I enjoyed chatting with you. Why don't you post the link to your video. I bet the other members of the forum would enjoy seeing it.

Can you tell the group a little more about the contest you entered? I didn't get the specifics, and I'm sure the rest of the group would be interested in hearing about it.


Reagards,

Duncan C