On Sun, 2008-07-06 at 18:54 -0400, Jim Muth wrote:
Once again I have gone to a very large magnitude to find the image, which is why the mathtolerance entry is included in the parameter file. These high-order minibrots grow more interest- ing the deeper they lie. It's too bad that arbitrary precision does not work with type=formula. It's probably due to the DOS memory limit. (When, if ever, we have a S.O.T.A. Windows ver- sion of Fractint, I assume this will be changed.)
I could hard code your DivideBrot5 formula and add arbitrary precision support. The problem with arbitrary precision and most formulae is the use of the function variables. We currently don't have all the functions coded in arbitrary precision. Jonathan
frm:DivideBrot5 { ; Jim Muth z=0, c=pixel, a=real(p1)-2, b=imag(p1)+0.00000000000000000001: z=sqr(z)/(z^(-a)+b)+c |z| < 1000000 }