On 03/26/2015 02:06 AM, Cousin Ricky wrote:
> On 03/25/2015 12:47 PM, Sven Littkowski wrote:
>> actually, using blobs is not difficult, still, I have a case where I
>> self simply don't succeed to find the right values. If you like, try
>> yourself on it and help me with a solution.
>>
>> Basically, I have a more or less cylindrical body, and wanted a smooth
>> transformation into a cylinder across, at its end. My own solution
>> doesn't look good, after days of trying.
>
> My historical tendency when faced with intractable blobs has been to
> give up on the blobs and use isosurfaces.

Here are some isosurface functions that duplicate your forms.

--------------------[BEGIN CODE]--------------------
#include "functions.inc"
#include "roundedge.inc"

// the cylinder
#declare fn_Cylinder_part = function
{ min
   ( RE_fn_Wheel (y, max (0, abs(x) - 12.5), z, 450, 50),
     f_sphere (x / 0.215, y, z, 475) * 0.215
   )
}

// the blob
#declare fn_Cylinder1 = function // equivalent to a blob component
{ pow
   ( 1 - pow (min (f_r (max (0, abs(x) -50 ), y, z), 2500) / 2500, 2),
     2
   ) * 0.7
}
#declare fn_Cylinder7 = function // equivalent to a blob component
{ pow
   ( 1 - pow (min (f_r (max (0, abs(x) - 2000), y, z), 300) / 300, 2),
     2
   ) * 1.5
}
#declare fn_Cylinder8 = function // equivalent to a blob component
{ pow
   ( 1 - pow (min (f_r (max (0, abs(x) - 2500), y, z), 300) / 300, 2),
     2
   ) * 1.5
}
#declare fn_Blob_part = function
{ (   0.75
     - fn_Cylinder1 (x * 4, y, z - 50)
     - fn_Cylinder7 (x * 4, y, z - 500)
     - fn_Cylinder8 (x * 4, y, z - 650)
   )
}
--------------------[END CODE]--------------------

The blob section uses only the 3 components, namely those that interact 
with the cylinder, less the components that (I perceive) were used to 
model the smooth transition.

Unfortunately, I can't seem to get them to blob well together.

Post a reply to this message