I've been working with blobs for years to make characters.  Years ago I
came up with a workaround for the problem that occurs when you have two
cylinders intersect each other.  If the cylinders have been scaled
considerably, one ends up with two spikes sticking out of the "knee".

After thinking about Rune's complaint about the "twice translated blob
texture," I did some more playing around and began wondering if the
spiky knee feature is more bug than purely expected mathematical result.

Intuitively, the green and red features ought to be the same.  If they
were cylinders, they would be!
Comments?
[  This has existed since at least 3.02, uncle Ken!  8-D   ]

Post a reply to this message

Attachments:
Download 'blobshop03.jpg' (10 KB) Download 'blobshop03.pov.txt' (2 KB)

Preview of image 'blobshop03.jpg'

"Greg M. Johnson" wrote:
> If the cylinders have been scaled considerably, one
> ends up with two spikes sticking out of the "knee".

Of course. Cylinder blob components have rounded end caps, like if there
were spheres at the ends. When you scale a cylinder blob component, the end
cap gets scaled too, just like when you scale a sphere.

> Intuitively, the green and red features ought to be the same.

No.

> If they were cylinders, they would be!

You are confusing cylinder objects with cylinder blob components. Cylinder
blob components have rounded end caps (like if there were spheres in the
ends) while cylinder objects don't have that. You can't compare them.

If you want to avoid the problem you're describing, don't scale cylinder
blob components non-proportionally. Instead design them with the correct end
points in the first place, or at least the correct length between the end
points.

Rune
--
3D images and anims, include files, tutorials and more:
Rune's World:    http://rsj.mobilixnet.dk (updated Nov 5)
POV-Ray Users:   http://rsj.mobilixnet.dk/povrayusers/
POV-Ray Webring: http://webring.povray.co.uk

Post a reply to this message

Rune wrote:

> > Intuitively, the green and red features ought to be the same.
>
> No.
>
> > If they were cylinders, they would be!
>
> You are confusing cylinder objects with cylinder blob components. Cylinder
> blob components have rounded end caps (like if there were spheres in the
> ends) while cylinder objects don't have that. You can't compare them.

No, that's not what I'm doing. Of course, scaling along the 'y' axis gives
expected results, in changing the roundedness of the end cap.  Examples of the
correct, expected behavior in the code below and the attached image...

blob{
        threshold 0.4
        cylinder {<0,0,0>,<0,10,0> ,0.25,1    scale <1,0.1,1>}
        pigment{Red}
        finish{IsoFinish}
        translate <-1,0.0,0>
        }

blob{
        threshold 0.4
        cylinder {<0,0,0>,<0,.1,0>,   0.25,1 scale <1,10,1>}
        pigment{Green}
        finish{IsoFinish}
        translate <0,0.0,0>
        }


blob{
        threshold 0.4
        cylinder {<0,0,0>,<0,1,0>,   0.25,1 scale 1}
        pigment{Green}
        finish{IsoFinish}
        translate <1,0.0,0>
        }


> If you want to avoid the problem you're describing, don't scale cylinder
> blob components non-proportionally. Instead design them with the correct end
> points in the first place, or at least the correct length between the end
> points.

I scaled the x &  z dimensions of the blob component and get a wacky alteration
of the blob in the y!! This is not mathematically correct, IFMSS. Look at the
code in my original post here....

Post a reply to this message

Attachments:
Download 'blobshop03b.jpg' (4 KB)

Preview of image 'blobshop03b.jpg'

Forgot to add:
the reason of course for x & z scaling of a y axis blob component becomes apparent
when one is trying to model human muscles & bones in a somewhat realistic manner.
It's not all "cylindrical".

Post a reply to this message

"Greg M. Johnson" wrote: in message news:3C05481E.C2CEAFBA@aol.com...
> No, that's not what I'm doing.

You said that the red and green samples in your original post should look
the same, but they shouldn't, because you have scaled them differently.

> I scaled the x &  z dimensions of the blob component and get
> a wacky alteration of the blob in the y!! This is not
> mathematically correct, IFMSS. Look at the code in my original
> post here...

I did tell you that you should expect that when scaling the cylinder blob
component non-proportionally. Just imagine that the cap is a sphere. It
looks longer in the y dimension if you scale it longer in the y dimension,
but it also looks longer in the y dimension if you scale it smaller in the x
and z dimension.

It seems to me that you forget that you have not just scaled the red
cylinder blob component down in the x and z dimensions, you have also given
it a larger radius (4 times as big). That's why the rounded cap of the red
cylinder blob component is 4 times as long in the y dimension than that of
the green cylinder blob component.

I really don't see the problem. Could you explain in a more specific way
what exactly is the wrong behavior?

Rune
--
3D images and anims, include files, tutorials and more:
Rune's World:    http://rsj.mobilixnet.dk (updated Nov 5)
POV-Ray Users:   http://rsj.mobilixnet.dk/povrayusers/
POV-Ray Webring: http://webring.povray.co.uk

Post a reply to this message

"Rune" <run### [at] mobilixnetdk> wrote in message
news:3c05573f@news.povray.org...
> "Greg M. Johnson" wrote: in message news:3C05481E.C2CEAFBA@aol.com...
> > No, that's not what I'm doing.
>
> You said that the red and green samples in your original post should look
> the same, but they shouldn't, because you have scaled them differently.
>
> > I scaled the x &  z dimensions of the blob component and get
> > a wacky alteration of the blob in the y!! This is not
> > mathematically correct, IFMSS. Look at the code in my original
> > post here...
>
> I did tell you that you should expect that when scaling the cylinder blob
> component non-proportionally. Just imagine that the cap is a sphere. It
> looks longer in the y dimension if you scale it longer in the y dimension,
> but it also looks longer in the y dimension if you scale it smaller in the
x
> and z dimension.
>
> It seems to me that you forget that you have not just scaled the red
> cylinder blob component down in the x and z dimensions, you have also
given
> it a larger radius (4 times as big). That's why the rounded cap of the red
> cylinder blob component is 4 times as long in the y dimension than that of
> the green cylinder blob component.
>
> I really don't see the problem. Could you explain in a more specific way
> what exactly is the wrong behavior?

You're right.  That's what's going on, basically because of the original
sizes and the strength factor too I guess.  I thought this was about
comparing regular cylinder + sphere ends with blob ones.  Seeing the pov
script I too noticed how the y scaling (or sizing) where the spikes are
extend to a distance (more or less) where the blob surface would reach to
were it not scaled on just the two axes.

--
text{ttf"arial","bob h",.1,0pigment{rgb 9}translate<-1,-.2,3>}

Post a reply to this message