In article <38aebcbe@news.povray.org>, "Lance Birch" <-> wrote:

> The fast displacement mapping looks's nice.

But is probably only possible due to the objects being tesselated to 
triangles...
However, I think it would be a useful feature to have the isosurface 
object optionally be reduced to a triangle mesh. You could even use a 
mesh version of the isosurface to bound one with the original solving 
methods...this might give the displacement features you want.(while the 
isosurface object doesn't need to be reduced to triangles to allow 
displacement, it might render faster that way)

Anything else?(other than network and multithreaded stuff...)
The only things I could find were NURBS and lens effects. And lens 
effects could be written for the post_process feature for MegaPOV(some 
things like blur are already included).

Where is some specific information on the texturing capabilities? All I 
could find was something saying there were over 40 patterns and blurred 
reflection/refraction...

-- 
Chris Huff
e-mail: chr### [at] yahoocom
Web page: http://chrishuff.dhs.org/

Post a reply to this message

Chris Huff <chr### [at] yahoocom> wrote:
: However, I think it would be a useful feature to have the isosurface 
: object optionally be reduced to a triangle mesh.

  You make it sound easy, but I think it isn't. How do you tesselate a
f(x,y,z) function?

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

You can spatially find points which lie on the surface by using several
methods.  The easiest is the spatial grid (in which you sample in rows and
columns down through the object at set intervals - you can build a shell of
the object by then joining all the points).

There are other methods for producing more accurate and efficient results
but I don't know how hard they are to implement.  Most of them use a simple
point test on the f(x,y,z) to determine if the point is inside or outside
the object.

--
Lance.
The Zone - http://come.to/the.zone

Post a reply to this message

In article <38b00462@news.povray.org>, Nieminen Juha 
<war### [at] sarakerttunencstutfi> wrote:

>   You make it sound easy, but I think it isn't. How do you tesselate a
> f(x,y,z) function?

I think there is an algorithm called "marching cubes" which might be 
sufficient. I believe it is commonly used in metaball 
modellers/renderers which don't do render directly like POV does.

I don't actually know anything about this "marching cubes" algorithm, 
but the name alone gives me some ideas...I don't think it would be an 
easy feature to implement in POV though. Maybe in 4.0...

-- 
Chris Huff
e-mail: chr### [at] yahoocom
Web page: http://chrishuff.dhs.org/

Post a reply to this message

On Sun, 20 Feb 2000 10:54:22 -0500, Chris Huff wrote:
>I think there is an algorithm called "marching cubes" which might be 
>sufficient. I believe it is commonly used in metaball 
>modellers/renderers which don't do render directly like POV does.
>
>I don't actually know anything about this "marching cubes" algorithm, 
>but the name alone gives me some ideas...I don't think it would be an 
>easy feature to implement in POV though. Maybe in 4.0...

Unfortunately, last time I checked that algorithm was patented.

Post a reply to this message

Ron Parker <ron### [at] povrayorg> wrote:
: Unfortunately, last time I checked that algorithm was patented.

  IMHO those who invent some algorithm and immediately patent it are fools.

  For example, let's take the fractal image compression algorithm. The author
patented it immediately after finding it. What happened? How many fractal
compression programs have you seen or used? How many programs has fractal
compressed image support? Just a handful. How much has the algorithm been
developed? Not very much, as far as I know. How much money has the author
received from licenses? I would guess that not very much.

  Let's take a free image format like PNG. It's very popular (perhaps the
most popular format in the internet after JPG) and gaining more popularity
every day. Most modern image processing programs support it. It's becoming
a de-facto standard in many places. Would this have happened if PNG was
patented? Certainly not.
  GIF was patented and began to lose popularity very fast.

  Patenting an algorithm will just make sure that it will not become popular
and that it will not be developed much further. It will also not get you
lots of money.

  I don't want to offend americans, but I think that they are fools because
they allow patenting of algorithms (or almost anything else). It's like
patenting hot water or a fork.
  Here you can have the copyright of a program, but you can't patent it.

-- 
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/

Post a reply to this message

"Nieminen Juha" <war### [at] sarakerttunencstutfi> wrote :
>
>   Patenting an algorithm will just make sure that it will not become
popular
> and that it will not be developed much further. It will also not get you
> lots of money.
>

    The first part of this statement may be correct, but I bet the rest is
false. Private industry and government uses things that we just don't have
access to, and probably don't really need. But they make money for the
people who need them and use them. If there is a manufacturing process that
uses a particular algorithm, patenting that will prevent someone else from
using it to run you out of business. So what if it's not "popular"? Popular
doesn't bring home any bacon unless you can convert it into cash. "Useful"
is much easier to convert into cash and if you want to make money by being
the -most- "useful", you don't -give- away your secrets.

    Making something useful and giving it away is a good tactic if your goal
is to make a name for yourself and then cash in on that name by getting a
job doing something else. But it is not the -only- good tactic.

Post a reply to this message

Ron Parker wrote:
> 
> Unfortunately, last time I checked that algorithm was patented.

AFAIK the marching triangles algorithm isn't.

Margus

Post a reply to this message

>   I don't want to offend americans, but I think that they are fools
because
> they allow patenting of algorithms (or almost anything else). It's like
> patenting hot water or a fork.
>   Here you can have the copyright of a program, but you can't patent it.

Yea, and look what happened to 128-bit encryption... *sigh*

Australians can rest assured that our "secure" internet transactions are
running on wonderfully weak 40-bit encryption.  Somehow I don't feel very
safe.

--
Lance.
The Zone - http://come.to/the.zone

Post a reply to this message

> What is it exactly about Lightflow that you like? I don't have Windows
> and there doesn't seem to be a PPC Linux version, so I can't run
> it(well, I could under SoftWindows, but...). I have looked at the web
> page a little, and just don't see what is so nice about it...

It sounds a bit childish: but the screw (the steel) on the web-page is
very good, I first asked me if it is a photo.
I want to make pictures with steel.
The other thing is: import of CAD-data (surfaces and volumes) via iges,
step, vda, sat,  ..., I know, this is very complex but you asked me ..
:-)

Christof

Post a reply to this message