Welcome to Fractal Forums

Found a Little inconsistency within import function of v2.07. In 1.21 the fractal constant factor is meaningless in Julia mode. The params for Julia constant is taken 1:1 as written. In v2 additionally the constant factor is applied to Julia constants. So if Julia_mode 1; is set in 1.21, reset constant factor to 1 in v2 to Import propperly. Nothing really serious - just to mention.
Don't ask why my params sometimes have a constant factor different than 1 - it is like that.

Buddhi changed the way the constant multiplier is applied. I can not remember the reasoning behind it but there were two ways of looking at it.

One way is this :
We create a fractal algorithm which we call a M-set.  We obtain a Julia constant/point from that set. Applying a constant multiplier to this point has effectively changed the Julia coordinates. Therefore we are not creating a Julia from our chosen point but somewhere on the path of the z vector.  So clicking on Julia point 1.0,1.0, 1.0 then applying uniform constant multiplier of 0.7,  is identical to just clicking on 0.7, 0.7, 0.7. So that is the argument for not applying the constant multiplier to the Julia Constant

BTW. Locking resolution to DE level may finally be implemented in V2.09 :)

yeah, yeah, I know. I use this correlation (which is the same as in the two dimensional Mandelbrot and Julia set) since the beginning. I applied it in 1.x via measurement tool and external spread-sheets.
But this is about importing properly, not about math basics...

If you could confirm I have the  basic maths correct,   then I could  argue for the the multiplication of the Julia constant to be removed from the code, then a conversion would not be required.

I may lack experience , but to me the old way of applying the Julia constant seems more correct. :)

Not quite sure, we're talking the same stuff.
Indeed clicking 1.0,1.0,1.0 while 0.7 is active constant factor means 0.7,0.7,0.7. But ONLY in v2.x!!!
in 1.x it remains 1.0,1.0,1.0 - that's the difference. And it leads to the wrong Import behaviour.

The parameter in initial post, shows nothing useful when imported to v2. Resetting the constant factor to 1.0,1.0,1.0 shows the wanted result - try it out. So importing Julia params, should reset constant factor by default and everything should be ok... I hope I made it clear.

Sorry my English is below average, but I am trying to say there are two solutions

1) write a conversion for the import

2) change  Mandelbulber code back to how it was in V2.06 and previous versions. (i.e. no conversion required.)


If my basic maths is correct then I would say 2) is the correct solution.  I am looking for confirmation  that I am correct.

Which do you think is the best solution??

I will implement conversion of Julia constant to provide compatibility with v1.21. So if in old settings will be constant multiplier 0.5 and Julia constant 2.0 2.0 1.0 then Julia constant will be converted to 4.0 2.0 2.0. Then the result will be correct.
I did this change in constant multiplier because I have found that to have correct transition between normal mode and Julia mode there is also neeed to multiply Julia constant by contant multiplier. Then shape of the fractal stays the same during transition.

@Buddhi. Thanks. :)

Yes I remember now, that was the benefit that I had forgotten,  the infinite possibilities in  morphing/transitioning,  with the use of 6 variables.

Still not sure, that this is the solution, but you are the experts.
In general, the new behaviour with applying constant factors also to julia sets is useful, because - especially in mandelboxes - you can grab a point in constant multiplied m-set and get a geometrically similar form in j-set. BTW this is also the way M3d works.
In v1.x you needed to multiply a measured point with the constant factor manually, to get a j-set similar to your m-set section. I did this noumerous times, while not resetting constant multiplier to 1. It was meaningless in v1.x!
My initial example file contains the following lines

In v1 it was meant to be calculated with exactly the constants written (I already applied the constant factor by hand!)

v2 applies the constant factor again so a jset is calculated according to
not the one I want...

Imho the "correct" way - if such exists - would be ignoring a set constant factor, when importing jset params from 1.x
I did it like that with the given example: After import i set consant factor to 1,1,1 - not touching j-constants - and voila you get the image like in v1!

Maybe you are right. As you wrote better solution would be if 1.x settings have Julia mode activated, then it will ignore constant factor (by resting it to [1,1,1]) and keep Julia constant the same. Both solutions will give correct image, but this would be less confusing for the user. Thank you for this tip.
Now I'm starting to implement it. In 2.08 it will work in correct way.

That sounds great, Buddhi. As you say, the resulting images might be the same, but as an art focussed fractalist, I will feel familiar with that behaviour, without considering math context too much. Not the worst, I guess, as I might be one of the most loyal users of mandelbulber. It's allways a pleasure, to help improving my favorite fractal program.
Thanks a lot, both of you!  :beer:

In the morning with a clear head,  :)

@ Buddhi
Maybe the best solution is implement a separate multiplier for Julia Constant. This will make compatibility with all versions except V2.07.

Julia constant and addCpixel are two individual  constants therefore a common multiplier is not appropriate

@ buddhi

Moving away from the v1.21 import issue

 z =  fn(z) +  (cpixel *  constantMultiplier)  +  (JuliaC *  NewVec3Multiplier),  is how i set up most formulas in OpenCL V1.21 and this allows for the six parameter morph between normal and Julia Mode.

@Taurus

Your Mandelbulber gallery art is the  "showcase"  for  the program.  It is the reason that I (and I assume many others) were attracted to Mandelbulber.  I would say you are "one of the most loyal and influential users of Mandelbulber. ;D