|
|
clipka wrote:
> CShake <cshake+pov### [at] gmailcom> wrote:
>> Here's a test of a sphere_sweep version of a deformed mesh cap.
>
> Looking good! Now all you need to do is get some "cushion deformation" into the
> mesh (if you know what I mean; looking head-on, the wires should appear to
> "bend outward" where they're closer to the mike's central axis), so that they
> hit the "equator" at a shallower angle, and then you'll have the real thing.
>
>
>> Render time: 1hr29min. (dual core, 3.7b31) Radiosity on, no focal blur.
>
> Hum, let's see what focal blur makes of it...
>
>
9hr36min. Radiosity and some subtle focal blur. I'm happy with how it
turned out. It does have that 'cushion deformation', I went and took
some reference photos of a real mic for reference.
I'd be willing to share the math if someone is interested, though I'd
need to clean it up a bit.
The general idea I did was start with a square grid (one axis at a
time), do some contractions based on the cosine of the i and j positions
(the matrix iterating variables, x and z really), then drop it onto a
sphere by saying 'point.y=sqrt(abs(r^2-x^2-z^2))', then tuck any points
outside the radius down underneath the equator and CSG difference them
out. The intermeshing was by taking each point and pushing it in or out
in the sphere normal direction by the wire diameter, alternating each
time. A fun side effect of this method is that if I have the number of
wires even, then they all bend into each other instead of missing.
Post a reply to this message
Attachments:
Download 'microphone.png' (283 KB)
Preview of image 'microphone.png'
|
|