Hi all,

On the subject of feature requests..is there an official channel for sugestions?

Anyway, what I'd really, really like to see is this:
Instead of a pigment and color_map only affecting the surface of the object, I'd like
to
see the possibility that the texture could physically *eat away* the object wherever a
specific keyword is used as a color_map entry ie erode. As I understand it, procedural
textures exist in 3d space anyway, but are only seen where the surface of the object
is,
so perhaps this feature could be implemented easily?

----------
Andy

Post a reply to this message

In article <36e80c83.0@news.povray.org> , "Andrew Cocker" 
<and### [at] acockerfreeservecouk> wrote:

> Anyway, what I'd really, really like to see is this:
> Instead of a pigment and color_map only affecting the surface of the object,
I'd like to
> see the possibility that the texture could physically *eat away* the object
wherever a
> specific keyword is used as a color_map entry ie erode. As I understand it,
procedural
> textures exist in 3d space anyway, but are only seen where the surface of the
object is,
> so perhaps this feature could be implemented easily?

Did you try a color map and transparency? I am not completely sure, but it
should work...

    Thorsten

Post a reply to this message

Thorsten Froehlich wrote in message <36e81444.0@news.povray.org>...
<snip>

>Did you try a color map and transparency? I am not completely sure, but it
>should work...
>
>    Thorsten

All that does is make parts of the surface clear..the entire inside of the object
remains
completely empty. I want the object to appear completely solid, but with color_map
entries
cut into it. I dont want to use media to do this, as I want to be able to texture the
object as any other ( metal finishes etc )
Anyway, using your suggestion of a Clear map entry, you can't really use highlights as
they appear on the clear section also. I suppose you could use a texture_map to get
round
this problem, but it's not what I was originally suggesting.
I know that Bryce3D has such 3d procedural texturing...not sure how it's implemented
though.

---------
Andy

Post a reply to this message

Andrew Cocker wrote:
> 
> Hi all,
> 
> On the subject of feature requests..is there an official channel for sugestions?
> 
> Anyway, what I'd really, really like to see is this:
> Instead of a pigment and color_map only affecting the surface of the object, I'd
like to
> see the possibility that the texture could physically *eat away* the object wherever
a
> specific keyword is used as a color_map entry ie erode. As I understand it,
procedural
> textures exist in 3d space anyway, but are only seen where the surface of the object
is,
> so perhaps this feature could be implemented easily?
> 
> ----------
> Andy

I think this is mostly what displacement mapping does. It's actually
quite a resource hog AFAIK. I think BMRT can do it.

I might be wrong but that's just human :)

Post a reply to this message

Bryce 3D doesn't really do anything POV-Ray can't do (better).  I assume
you are talking about the volume materials in Bryce.  There's one option
to do specular shading on a volume, which POV can't do, but it's so
incredibly slow that it's hardly useful....however, it may be possible
to do even this using a proper scattering model with media.

There's also a lot of problems with volumes in Bryce.  They tend to have
a lot of artifacts like dark edges.

-Mike

Andrew Cocker wrote:
> 
> Thorsten Froehlich wrote in message <36e81444.0@news.povray.org>...
> <snip>
> 
> >Did you try a color map and transparency? I am not completely sure, but it
> >should work...
> >
> >    Thorsten
> 
> All that does is make parts of the surface clear..the entire inside of the object
remains
> completely empty. I want the object to appear completely solid, but with color_map
entries
> cut into it. I dont want to use media to do this, as I want to be able to texture
the
> object as any other ( metal finishes etc )
> Anyway, using your suggestion of a Clear map entry, you can't really use highlights
as
> they appear on the clear section also. I suppose you could use a texture_map to get
round
> this problem, but it's not what I was originally suggesting.
> I know that Bryce3D has such 3d procedural texturing...not sure how it's implemented
> though.
> 
> ---------
> Andy

Post a reply to this message

No, displacement mapping is the same as a bump map only it actually bends
the object's surface (the Isosurface patch can do this).

In respect to having the texture erode the object however, this is quite
complex.

Basically what it would mean is the the renderer would have to (when it hit
the texture with the erode keyword on whatever colour you've chosen) keep
tracing the path until it hit an area of a different colour (determine the
colour of each point through the object).  The problem with this is, while
it being very slow, it also has to have a way to determine if it's actually
hit the colour boundary... for instance, you could have an erode fade to a
solid colour, in which case the renderer would have to do a huge amount of
work!!!  Also, it then has to determine a surface normal for this newly
found point of colour which would be nearly impossible!

So basically, it's not possible by using a texture.  The only real way I
could see is to use the Isosurface patch to recreate the colour pattern as
an object and then difference this from the object which you want to erode.

--
Lance.


---
For the latest 3D Studio MAX plug-ins, images and much more, go to:
The Zone - http://come.to/the.zone

Post a reply to this message

On Fri, 12 Mar 1999 15:47:53 +1000, Lance Birch <lan### [at] usanet> wrote:

>So basically, it's not possible by using a texture.  The only real way I
>could see is to use the Isosurface patch to recreate the colour pattern as
>an object and then difference this from the object which you want to erode.

If I remembered to include it, the next superpatch will include an isosurface
function called pigment() that allows you to use any pigment as the scalar
field.

Post a reply to this message

Lance Birch wrote in message <36e8aaf3.0@news.povray.org>...
<snip>
>Basically what it would mean is the the renderer would have to (when it hit
>the texture with the erode keyword on whatever colour you've chosen) keep
>tracing the path until it hit an area of a different colour (determine the
>colour of each point through the object).  The problem with this is, while
>it being very slow, it also has to have a way to determine if it's actually
>hit the colour boundary... for instance, you could have an erode fade to a
>solid colour, in which case the renderer would have to do a huge amount of
>work!!!

I take your point, though I had envisaged it only working in a color_map where clean
steps
are used, not fades.

>Also, it then has to determine a surface normal for this newly
>found point of colour which would be nearly impossible!

Would that still be the case if fading was not used, just clean steps of colour?

<snip>
>Lance.


I'm not at all conversant with the inner workings of POV, so whilst my suggestion made
perfect sense to me, as you point out, I'm probably asking the impossible. I just
thought
it was so simple
ie

pigment {bozo
color_map {
[0 Erode]
[0.5 Erode] // clean step
[0.5 rgb 1]
[1 rgb 1]
}
}

    When shooting rays at the surface, where Erode is found, the ray enters the object
until it hits a color_map entry other than Erode ( procedural textures occupy 3d space
I
think ). This point then becomes the surface of the object. Surely the surface normal
of
this point could be worked out by sampling the position of the surrounding points ( no
idea personally of the math of what I'm suggesting here, but it makes sense to me ;~)
).
If this was workable, I don't see it being neccessarily as slow as media, depending on
the
scale of the pigment used.

I think that such a feature would be fantastic, so forgive me for pursuing the point.

---------
Andy

Post a reply to this message

Ron Parker wrote in message <36e914ac.0@news.povray.org>...
>On Fri, 12 Mar 1999 15:47:53 +1000, Lance Birch <lan### [at] usanet> wrote:
>
>>So basically, it's not possible by using a texture.  The only real way I
>>could see is to use the Isosurface patch to recreate the colour pattern as
>>an object and then difference this from the object which you want to erode.
>
>If I remembered to include it, the next superpatch will include an isosurface
>function called pigment() that allows you to use any pigment as the scalar
>field.
>

Forgive my ignorance Ron, but I don't understand what this means. Does it mean that
you
will be able to implement what I'm asking for?

---------
Andy

Post a reply to this message

On Fri, 12 Mar 1999 16:32:56 -0000, Andrew Cocker 
	<and### [at] acockerfreeservecouk> wrote:
>
>Ron Parker wrote in message <36e914ac.0@news.povray.org>...
>>On Fri, 12 Mar 1999 15:47:53 +1000, Lance Birch <lan### [at] usanet> wrote:
>>
>>>So basically, it's not possible by using a texture.  The only real way I
>>>could see is to use the Isosurface patch to recreate the colour pattern as
>>>an object and then difference this from the object which you want to erode.
>>
>>If I remembered to include it, the next superpatch will include an isosurface
>>function called pigment() that allows you to use any pigment as the scalar
>>field.
>>
>
>Forgive my ignorance Ron, but I don't understand what this means. Does it mean that
you
>will be able to implement what I'm asking for?

In a sense, yes.  Actually, it means that _you_ will be able to implement 
what you're asking for.  It will only work for CSG-able objects, though, 
so not on meshes or bicubics.
  
What you would do is make an object that is the isosurface of the pigment 
field at a given threshold.  Parts of the pigment below that threshold
would be solid, and other parts would be hollow (or vice versa).  You could 
then use color maps or CSG (depends on what the patch supports; I've forgotten 
and don't have the code in front of me) to make the required intervals hollow
or solid as needed.  CSG intersect the resulting isosurface(s) with your
object, and the result will be what you're looking for.  You will want to 
let the color_map for the solid parts "bleed" into the hollow parts a little,
to avoid unfortunate results similar to the coincident surfaces problem.

It won't be fast, but it might be faster than media.  It won't be available 
for a while, though; I'm still testing stuff.

Post a reply to this message