HOME | DD

Description
Apo 2.08JK

Hi to all!

On the picture above the interaction of JuliaRandomN with several Blurs is demonstrated. On each level of itterations N randomly changed: 4, 2, 3, ...and one from 4 Blurs is selected. Here is short story of how it's possible.

I'm not professional programmer. So I don't know a deep structure of Apo program. But it was very interesting to investigate IFS process. So I've found some compromise.

The standard Apo procedure for rendering IFS fractal is the next:
1)Apo chooses Transform according to the rules of chaos game;
2)sends FTx, FTy to plugin which is responsible for that Transform;
3)plugin calculates FPx, FPy and returns their values to Apo;
and so on...

It is possible to reorganize this process as following:
1)We introduce static variables in plugin. They will play the role of FTx, FTy. Their values are kept inside plugin.
2)We determine the rules of Transform choice inside plugin (all Transforms are determined as inside subroutines).
3)plugin calculates FPx, FPy,FPc (for color) and returns their values to Apo.

So we have absolute control on IFS!!!

In SphTiling3.c, SphTiling4.c,SphTiling5.c you can find the realization of this idea for simple IFS without coloring. In JuliaRandomN.c you can see how it's possible to organize more complicated IFS. The important instrument is a function DN1(m), which generates the sequence of random numbers and gives the possibility for its exact repetition any time we need.

In Truchet.c you can find what can be done through direct coloring in simple Sierpinski.The Truchet coloring scheme is superimposed on Sierpinski carpet.You can choose coloring scheme with the help of several constants. The usage of Z=fmod(A*Z+B,1) (instead of simple linear Z=A*Z+B) was introduced by Jed Kelsey. It gives more possibilities for deep coloring, but it is the source of dusty noise on the other side.

So it is short introduction in my experiments with controlled IFS in Apo.

In download file you can find samples with source code and .dll.

All the best.
eralex61.

Description
Apo 2.08JK

Hi to all!

On the picture above the interaction of JuliaRandomN with several Blurs is demonstrated. On each level of itterations N randomly changed: 4, 2, 3, ...and one from 4 Blurs is selected. Here is short story of how it's possible.

I'm not professional programmer. So I don't know a deep structure of Apo program. But it was very interesting to investigate IFS process. So I've found some compromise.

The standard Apo procedure for rendering IFS fractal is the next:
1)Apo chooses Transform according to the rules of chaos game;
2)sends FTx, FTy to plugin which is responsible for that Transform;
3)plugin calculates FPx, FPy and returns their values to Apo;
and so on...

It is possible to reorganize this process as following:
1)We introduce static variables in plugin. They will play the role of FTx, FTy. Their values are kept inside plugin.
2)We determine the rules of Transform choice inside plugin (all Transforms are determined as inside subroutines).
3)plugin calculates FPx, FPy,FPc (for color) and returns their values to Apo.

So we have absolute control on IFS!!!

In SphTiling3.c, SphTiling4.c,SphTiling5.c you can find the realization of this idea for simple IFS without coloring. In JuliaRandomN.c you can see how it's possible to organize more complicated IFS. The important instrument is a function DN1(m), which generates the sequence of random numbers and gives the possibility for its exact repetition any time we need.

In Truchet.c you can find what can be done through direct coloring in simple Sierpinski.The Truchet coloring scheme is superimposed on Sierpinski carpet.You can choose coloring scheme with the help of several constants. The usage of Z=fmod(A*Z+B,1) (instead of simple linear Z=A*Z+B) was introduced by Jed Kelsey. It gives more possibilities for deep coloring, but it is the source of dusty noise on the other side.

So it is short introduction in my experiments with controlled IFS in Apo.

In download file you can find samples with source code and .dll.

All the best.
eralex61.