Title: quat cubic4D with abs() frag Post by: mclarekin on August 29, 2017, 06:06:36 AM @sabine. Not sure if I have the DE totally correct :hmh:
InitialW usually a small number like 0.2 Title: Re: quat cubic4D with abs() frag Post by: Sabine on August 29, 2017, 04:26:15 PM :o :o :o
Will check tonight! Title: Re: quat cubic4D with abs() frag Post by: Sabine on August 29, 2017, 09:47:48 PM @mclarekin First impression: fantastic! :beer: :beer: :beer:
Will need to test some more and DE might be not 100%, but while what I tried to do was an utter failure :embarrass:, this produces exactly what I want. Now I must try to understand what you have done there:) Why for instance scale 1222 must be 1,2,2,2... And why what you have done does what cubicquat in mb3d does, while its formula is quite different... Below testrun with testversion of new raytracer which C-JR is building. Title: Re: quat cubic4D with abs() frag Post by: mclarekin on August 30, 2017, 01:09:58 AM scale 1222 is just the same maths rearranged. The constant "2.0 * " has been taken out of the formula and used as default parameter.
If I put them back, then it would be scale1111 as the vec4 parameter. It just the way I have always coded a quat, (although in Mandelbulber2 it is scale 1220 so that it is 3D until you start adding 4D parameters) Quote gradient = (vec3(escapeLength(p+xDir*EPS), escapeLength(p+yDir*EPS), escapeLength(p+zDir*EPS))-r)/EPS; Yep, that looks like DeltaDE type thing. With analytic versus deltaDE , generally I use deltaDE with formulas where analytic won't work. But sometimes it is worth trying both as It can be parameter specific ie some parts of the fractal are best with analytic and other parts are best with deltaDE. Title: Re: quat cubic4D with abs() frag Post by: Sabine on August 31, 2017, 09:15:43 AM :embarrass: Sometimes answers can be so simple... :embarrass: Thank you very much, mclarekin! :beer: :beer: :beer: |