On 6/23/09 11:28 AM, "Tom Browder" <tom.browder@...> wrote:
> On Tue, Jun 23, 2009 at 10:03, Erik Greenwald<egreenwald@...> wrote:
> ...
>> Both 'gravotronic' and 'bitronic' were used as goofy placeholder names, with
>> the proc-db intended to make animation sets. 'bitronic' being two control
>> points moving away to show the interlinking surface and gravitronic being a
>> cluster controlled by a gravity simulation to create a 'jiggly' ball.
>>
>> If you'd like to put effort into that proc-db, I'd appreciate it. I had
>> other tasks come up that took priority. Let me know if you have questions
>> about anything metaball related
>
> Hi, Erik, yes, can you briefly explain the parameters used to create a
> metaball (is it something like a bezier curve in 3d)? I've used my
> builder script to play around with some and see that they are
> certainly not all non-linear.
I apologize for the delay in responding, been coping with a broken wrist.
The "metaball" primitive contains a method ID, threshold value, and an
unordered set of control points. Each control point contains a 3d location,
a "field strength", and a possibly a "blobbiness" value (called "goo" in
rt_metaball_add_point).
There are three method ID's defined
* "metaball", which is the Tokyo Metaball approximation of the Blinn
Blobby Surface. This method is not implemented yet.
* "blob", the Blinn method from SIGGRAPH '82 (I think). I can dig up the
journal and give you the bib info if needed.
* "iso", which is a simple \Sum_{i}\frac{f_{i}}{d^{2}} computation like
you'd see for computing gravitational magnitude or point charge in a
basic physics course.
The surface of the primitive exists where the summation of the points
contribution is equal to the threshold, with the general fldstr/distance^2
pattern.
The blobbiness value is only used in the blob method, and modifies the
gusseting effect.
In general, you should start with a threshold of 1 and a blobbiness/goo of 1
This primitive is odd in that it effectively defines values for all points
in space and pretends there is a surface at the contour. It's also in an
experimental/developmental stage and suspect to fundamental changes in its
data structure.
If I had free time, I'd be working on the proc-db to generate animations of
different values changing to provide a visual demonstration, but I don't :(
Hopefully that answers your questions? :)
> Thanks.
>
> Regards,
>
> -Tom
-Erik
|