Tor Olav Kristensen wrote:

> Here's a little peek at what I'm working on these days.
>
> The image shows a bivariate NURBS surface together with
> it's control points. A few of the control points (yellow
> spheres) are not visible.
>
> The two "straight" lines visible upon the surface of the
> bend is caused by my mesh2 macro, which was not made for
> surfaces with this kind of topology. (The normal is not
> calculated correctly along these lines, but I'll probably
> fix that soon.)
>
> Tor Olav
>

Although I do find NURBS interesting, I'm curious:
are they easy to visualize? I mean, do you have some
sort of external modeller for that, or are you using pure
POV and some other methods, like Wireframe etc?

Cause when I use POV, I need to either visualize things
very easily, or I need a lot of patience and trace a lot
of 360 degree views etc...

--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html
Email: Tim### [at] gmxde

Post a reply to this message

Tim Nikias wrote:
> 
>...
> Although I do find NURBS interesting, I'm curious:
> are they easy to visualize? I mean, do you have some
> sort of external modeller for that, or are you using pure
> POV and some other methods, like Wireframe etc?
> 
> Cause when I use POV, I need to either visualize things
> very easily, or I need a lot of patience and trace a lot
> of 360 degree views etc...

Tim,
I have barely used any modellers, so I have very little
experience with them. I like to write POV scripts instead.

When I'm working with NURBS, I first try to sketch on
paper the general shape that I want to make. (It also
happens that I am just curious about what shape a certain
macro will output.)

And then I try to figure out where I'll have to put the
control points (that will pull the surface towards them).

Then I call my NURBS and mesh macros and feed them with
the control points and the output from some macros that
generates different knot vectors and weights for the
control points.

When I see the output from this, I might adjust the
control points and their weights. Sometimes I also switch
to other knot macros (in order to have the surface to
"approach" the control points in a different "way".)


Tor Olav

Post a reply to this message

If you really want to use NURBS, you'll have to use a modeller. That's way
easier. I guess it's worth the effort to build a patch for Moray or write a
Java application for this. But who's got the time?  ;-)

I absolutely like this stuff.

--
Apache
POV-Ray Cloth experiments: http://geitenkaas.dns2go.com/experiments/
Email: apa### [at] yahoocom
ICQ: 146690431

Post a reply to this message

That's a very nice shape. It has something "useful" to it. Have you been
able to reduce the calculation time of your macros? The shape gives me an
idea: Maybe it's easy to make interesting lathes with nurbs. Something to
try out!


Regards,
Hugo

Post a reply to this message

Tor Olav Kristensen <tor### [at] hotmailcom> wrote:
> Here's a little peek at what I'm working on these days.
>
> The image shows a bivariate NURBS surface together with
> it's control points. A few of the control points (yellow
> spheres) are not visible.
>
> The two "straight" lines visible upon the surface of the
> bend is caused by my mesh2 macro, which was not made for
> surfaces with this kind of topology. (The normal is not
> calculated correctly along these lines, but I'll probably
> fix that soon.)

now if you could get this to directly read the output from rhino, save
converting to a mesh a exporting etc etc


--
Rick

Kitty5 WebDesign - http://Kitty5.com
POV-Ray News & Resources - http://Povray.co.uk
TEL : +44 (01270) 501101 - FAX : +44 (01270) 251105 - ICQ : 15776037

PGP Public Key
http://pgpkeys.mit.edu:11371/pks/lookup?op=get&search=0x231E1CEA

Post a reply to this message

I bought an expensive book on nurbs mathematics recently, I haven't had
time to study it yet.  It would be nice if I didn't have to work.  I
learned so much when I was unemployed.  
-- 
Dan Johnson 

http://www.geocities.com/zapob

Post a reply to this message

"Rick [Kitty5]" wrote:
> 
> Tor Olav Kristensen <tor### [at] hotmailcom> wrote:
> > Here's a little peek at what I'm working on these days.
> >
> > The image shows a bivariate NURBS surface together with
> > it's control points. A few of the control points (yellow
> > spheres) are not visible.
> >
> > The two "straight" lines visible upon the surface of the
> > bend is caused by my mesh2 macro, which was not made for
> > surfaces with this kind of topology. (The normal is not
> > calculated correctly along these lines, but I'll probably
> > fix that soon.)
> 
> now if you could get this to directly read the output from rhino, save
> converting to a mesh a exporting etc etc

Rick,

If Rhino outputs the control data for its NURBS into binary files, then
those files would be difficult (or impossible ?) to read with the POV SDL.
(In that case I think a file converter program would be needed.)


I have looked briefly at the proposed NURBS extension to VRML-97.
(http://www.blaxxun.com/developer/contact/3d/nurbs/spec/nurbsproposal.html)

- And I think that maybe I could adopt my macros to adhere as close
as possible to such a standard.

There's one great worry though: Trimming of NURBS. (To allow for
complex surface topologies.) - I fear that POV's SDL will be to slow for
this task.


Btw.:
I (and others) have already made macros that will export the meshes
(e.g. mesh2 or STL) produced by these parametric surface functions.


Tor Olav

Post a reply to this message

Hugo wrote:
> 
> That's a very nice shape. It has something "useful" to it. Have you been
> able to reduce the calculation time of your macros?

The problem with my current NURBS macros is that they build HUGE functions.
(I.e.: These functions are slow to evaluate.)

In order to make some NURBS macros that builds smaller functions, I need a
set of macros that will will do calculations with matrices and polynomials.

I'm about to finish such a set of such support macros. It is a lot of work
to figure out how to make these macros efficient, easy to use and to decide
how to organize the data that "wanders" between them.

(I suspect that the time to build the functions will be longer. But since
the functions will be smaller and faster to evaluate; I expect that the
total parsing time will be shorter.)


> The shape gives me an idea: Maybe it's easy to make interesting lathes
> with nurbs. Something to try out!

I have have done some successful experiments lately with B-splines that
follows circular arcs. The next thing I will try to do is to make splines
that follows elliptical arcs (and other conic sections) and then use them
in NURBS surfaces.

So "advanced" lathes, extrusions and cross sections that follows splines
shouldn't thereafter be very difficult to do.


Tor Olav

Post a reply to this message

Dan Johnson wrote:
> 
> I bought an expensive book on nurbs mathematics recently, I haven't had
> time to study it yet.  It would be nice if I didn't have to work.  I
> learned so much when I was unemployed.

Yes, such books are expensive.

I too bought a similar book a while ago: "Geometric Modeling with Splines"
(For about 700 NOK). It touches many themes, contains a lot of math, but
it is a little bit "heavy" to read.


Tor Olav

Post a reply to this message

Apache wrote:
> 
> If you really want to use NURBS, you'll have to use a modeller. That's way
> easier.

Yes, using a modeller would be the best solutions for many problems.

But I think that there are geometric problems that NURBS macros in POV
can "solve" just as well as any modeller can.


> I guess it's worth the effort to build a patch for Moray or write a
> Java application for this. But who's got the time?  ;-)
> 
> I absolutely like this stuff.

Lately I have been thinking that it is about time that I learn myself
a new programming language. And I'm very curious about Java.

So I thought that some NURBS programming could inspire me to read and
experiment some with Java. Maybe I could try to make a simple NURBS
modeller that would run in a browser window and spit out mesh2 files.

I found an online .pdf book about Java programming: "Thinking in Java"
by Bruce Eckel, that I have started reading. I like the way he writes.

Here it is:
http://www.mindview.net/Books

Now I too just need some more time !


Tor Olav

Post a reply to this message