Nieminen Juha wrote:
> 
>   Couldn't these macros be used with the PCM format?
>   You could convert an ordinary mesh to PCM format with the triangle
> compressor program, then read it to arrays with Colefax's PCM macro and
> then use these macros.
>   Have you considered asking Colefax if you two could work together in
> a "superset" of mesh handling macros?

  I had similar thoughts when I read John's reply. It would be highly
advantagous if you could use his tools with existing mesh files instead
of only those you can generate yourself. For example many low polygon
count models are available for free on the net but are of poor quality
because of the low polygon count. The perfect solution of course if
subdivision of the existing triangles to help smooth curves and moderate
the number of seemingly flat faces. Sounds like a good suggestion to me.

-- 
Ken Tyler

See my 850+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html

Post a reply to this message

On Sat, 28 Aug 1999 17:08:52 -0400, John VanSickle
<van### [at] erolscom> wrote:

>After much pondering, I have developed a suite of macros for working
>with triangle meshes.  One of these macros will take a mesh of triangles
>and calculate the smoothing normals required.  Another will perturb the
>vertices of the mesh, making it rougher-looking.
>
>However, the macro of interest is the divider, which takes a rough mesh
>of large triangles and replaces it with a mesh of smaller triangles that
>give a more curved appearance.  This process can be repeated, yielding
>a fine mesh of triangles that simulate a large variety of curved
>surfaces.  Its primary purpose is to allow a user to create a smooth
>organic shape, containing many triangles, by defining a rougher mesh of
>far fewer triangles, and refining that in an automated way.
>
>The Subdivision Surface Suite (SSS) can be found at:
>
>  http://users.erols.com/vansickl/sss.htm
>
This looks MOST useful.

I'll take a look at it as soon as poss :)


Cheers,

Cliff Bowman
Why not pay my 3D Dr Who site a visit at http://www.who3d.cwc.net/

Post a reply to this message

Nieminen Juha wrote:
>   Couldn't these macros be used with the PCM format?
>   You could convert an ordinary mesh to PCM format with the triangle
> compressor program, then read it to arrays with Colefax's PCM macro and
> then use these macros.
>   Have you considered asking Colefax if you two could work together in
> a "superset" of mesh handling macros?

and Ken <tyl### [at] pacbellnet> replied:
>   I had similar thoughts when I read John's reply. It would be highly
> advantagous if you could use his tools with existing mesh files instead
> of only those you can generate yourself. For example many low polygon
> count models are available for free on the net but are of poor quality
> because of the low polygon count. The perfect solution of course if
> subdivision of the existing triangles to help smooth curves and moderate
> the number of seemingly flat faces. Sounds like a good suggestion to me.

For what it's worth, I can only add my support to such an idea: indeed, Warp
(Juha/Mika/whatever?!) and I had planned the PCM format to be expandable (as
the addition of bicubic patches shows).  Subdivision surfaces really
shouldn't be too difficult to implement with the PCM format (haven't we
heard that one before!), and although I haven't tried John's file as yet I
am, of course, happy to look into it....

Post a reply to this message

Chris Colefax wrote:
> 
> For what it's worth, I can only add my support to such an idea:
> indeed, Warp (Juha/Mika/whatever?!) and I had planned the PCM format
> to be expandable (as the addition of bicubic patches shows).
> Subdivision surfaces really shouldn't be too difficult to implement
> with the PCM format (haven't we heard that one before!), and although
> I haven't tried John's file as yet I am, of course, happy to look into
> it....

The Subdivision Surface Suite uses triangles meshes to generate finer
triangles meshes, and consequently the type one PCM file works just
fine.  I've written a macro to read the PCM type one files into the
arrays that the SSS uses, and the results can be seen in a post on
povray.binaries.images.  I had the SSS read in the torus from torus.pcm,
and do two subdivisions on it; the original torus, the once-divided
torus, and the twice-divided torus, are all depicted, in both flat and
smoothed forms.

The PCM format seems to have been written from a different approach than
what I took when designing the SSS.  Specifically, the SSS was designed
with expectation that the smoothing normals would need to be
re-calculated after each subdivision; hence, it doesn't even bother to
maintain the data internally, but generates it when needed
(specifically, when the mesh object is declared).  For this reason, my
PCM reader ignores the normal data in the file and tosses all of the
triangles into one mesh; after the subdivider has a go at the mesh, the
normal data all has to be re-calculated anyway.  This has a side benefit
that a lot of people have asked for:  The SSS can read in a mesh without
normals, and generate them.

I have written a macro to write the SSS mesh as a PCM1 file with flat
triangles, and when I get some breathing room I'll write another macro
to generate smooth triangle data in the PCM1 format.

I'm also planning the ability to add basic geometric shapes to the mesh,
such as the tetrahedron, octahedron, and icosahedron (all made of
triangles), and also allow the user to add Bezier patches of the first
through the fourth degree (even allowing the u-splines and v-splines to
be of differing degrees in the same patch); however, they would be
converted to triangle meshes, and not be carried internally as Bezier
patches.

There still remains the issue of individually-texturing the triangles.
I'm also planning to add UV support when it becomes part of the official
POV-Ray release.

Regards,
John
-- 
ICQ: 46085459

Post a reply to this message

On Tue, 31 Aug 1999 22:21:36 -0400, John VanSickle
<van### [at] erolscom> wrote:

[snip]
>
>There still remains the issue of individually-texturing the triangles.
>I'm also planning to add UV support when it becomes part of the official
>POV-Ray release.
>
Have we been given any indication of how different the UV syntax is
likely to be in the official version?

Anyone?


Cheers,

Cliff Bowman
Why not pay my 3D Dr Who site a visit at http://www.who3d.cwc.net/

Post a reply to this message