> Hi !

> Okay, antoher question for special-effects: Are there any Morphing-tools
> available ?

	It will be interesting to 1) see one and 2) actually try to use
it. A 3D association of points as to which points morph into
which? 2D is "interesting" enough with graphic assistance. 

	But if you just want to show off, topologists have
transformations mapped.

-- 
http://www.giwersworld.org/artiii/

Oh my God! They've rendered Kenny!

How to profit from the end of civilization as know it available 
here soon.

Post a reply to this message

Margus Ramst wrote:

> Well, for standard primitives I'm afraid even Chris isn't the key.

	The old morphing a cube into a sphere trick. 

-- 
http://www.giwersworld.org/artiii/

Oh my God! They've rendered Kenny!

How to profit from the end of civilization as know it available 
here soon.

Post a reply to this message

On 29 Nov 1999 11:52:23 -0500, Nieminen Juha
<war### [at] punarastascstutfi> wrote:

>  Morphing should be theoretically possible with povray 3.1 (altough it may
>be easier with uvpov).
>  The idea behind morphing is that you specify a (2D) triangle mesh in the
>first image and an equivalent mesh (same number of triangles at same relative
>locations) in the other image and them interpolate between the meshes. The
>triangles have the UV-textures defined by the images and these textures are
>interpolated too.
>  It shouldn't be difficult to set up a tool (include file?) for doing this
>with povray. It's like the triangle mesh morphing plus uv-mapping with
>the textures interpolated.

That's exactly what I had done (thanks to the UV mapping macro) but
it's all gone now. Anyway, I abandoned the problem because I ran into
a great problem with adaptive mesh subdivision (more precisely, POV's
arrays being static). Without it, great deformations were a pain,
especially near the image edges, and parsing was slow enough with a
20000 triangles mesh, so I gave up on non-adaptive subdivision. It
doesn't matter anymore either way.


Peter Popov
pet### [at] usanet
ICQ: 15002700

Post a reply to this message

To morph primitives you can use isosurfaces with something like this :
P = P1*(1-clock) + clock*P2
where P1 and P2 are the respective equations of the start and end
shapes.
Just an idea, but it worked to make and column having a round base
morphing smoothly to a square top (with y instead of clock).
Of course this does not solve the problem of morphing CSG compounds or
meshes.

G.

Matt Giwer wrote:

> Margus Ramst wrote:
>
> > Well, for standard primitives I'm afraid even Chris isn't the key.
>
>         The old morphing a cube into a sphere trick.
>
> --
> http://www.giwersworld.org/artiii/
>
> Oh my God! They've rendered Kenny!
>
> How to profit from the end of civilization as know it available
> here soon.

Post a reply to this message

Morphing meshes is relatively easy (I stress _relatively_ here),
especially if the number of vertices is the same / similar.

Margus

Gilles Tran wrote:
> 
> Of course this does not solve the problem of morphing CSG compounds or
> meshes.
>

Post a reply to this message

Nieminen Juha wrote:
> 
>   Morphing should be theoretically possible with povray 3.1 (altough it may
> be easier with uvpov).

But it really isn't practical in script/macro form. It would be too
slow, especially because you rarely have meshes with the same number of
vertices and faces. I think POV script is just too clumsy and slow to do
testellation (even non-adaptive) in an effective manner. It may cope OK
with a few thousand vertices, but an average mesh is ten times that.

Margus

Post a reply to this message

Margus Ramst wrote:

> But it really isn't practical in script/macro form. It would be too
> slow, especially because you rarely have meshes with the same number of
> vertices and faces. I think POV script is just too clumsy and slow to do
> testellation (even non-adaptive) in an effective manner. It may cope OK
> with a few thousand vertices, but an average mesh is ten times that.
>
> Margus

There's something I've still have to try which is morphing between two Poser
meshes, for instance two meshes of the same character but with different poses.
The number of triangles would be slightly different but I guess they would be
more or less at the same "place" (foot, head etc.). Even if it wouldn't be a real
smooth morph, the effect of having little triangles flying around and  taking
shape could be cool.
G.

Post a reply to this message

I was just thinking that an unusual type of morphing would be just moving
the csg primitives to their new location so as to turn one object into
another. Say your computer into a refrigerator...

Disk drive button turns into the handle, small cylinders that rounded the
corners a little grow into larger ones. Scale changes, texture changes,
mousepad turns into a tile floor...

In some changes it might be more TransFormer(tm) than Morph, but that might
be cool.

Not that I have the time or patience to do that...

Post a reply to this message

>Just an idea, but it worked to make and column having a round base
>morphing smoothly to a square top (with y instead of clock).


Ooohh, I've been looking for a way to do this. Could you please write the
entire formula for doing this? TIA

Post a reply to this message

TonyB wrote:
> 
> >Just an idea, but it worked to make and column having a round base
> >morphing smoothly to a square top (with y instead of clock).
> 
> Ooohh, I've been looking for a way to do this. Could you please write the
> entire formula for doing this? TIA

	My solution:
y*(x^2+z^2-1)+(1-y)*(max(abs(x), abs(z))-1) threshold 0

	Gilles':
(abs(x)+abs(z))*(1-y)+(x^2+z^2)*y threshold 1

	(From the french ng where this question appeared a few weeks ago)
		Jerome

-- 
*******************************

* they'll tell you what can't * mailto:ber### [at] inamecom
* be done and why...          * http://www.enst.fr/~jberger
* Then do it.                 *
*******************************

Post a reply to this message