Any chance of this being possible

TIA

Pabs

Post a reply to this message

Yes.

Pabs wrote:
> 
> Any chance of this being possible
> 
> TIA
> 
> Pabs

Post a reply to this message

"Mr. Art" wrote:

> Yes.

How Mr Art.

Pabs

Post a reply to this message

This short bit of code shows how to use a 
planetary image as a displacement map on a sphere.
I had set it up as a macro to work with different
planet displacement maps, planet pigment maps, planet sizes,
and percentage displacements.
I know that this is displacement mapping only on a
sphere, but the same concept should work on other shapes.
I hope that this helps.

#macro PlanetMac(Pic1,Pic2,Rad,Prec)
	#local Pic= function{pigment {image_map {png Pic1 interpolate 2
map_type 1}}}
	#local R0=Rad;
	#local FUNC1=function {"sphere",<R0>}
	isosurface
		{
		function{FUNC1(x,y,z)-(Pic(x,y,z)*R0*Prec)}
		contained_by {sphere {0,R0*(1+Prec)}}
		method 2 max_gradient 10
		pigment	{image_map{png Pic2 interpolate 2 map_type 1}}
		finish	{ blinn 1 facets .2 reflection 1 reflect_metallic
conserve_energy}
		interior{ior 20}
		}
#end


Pabs wrote:
> 
> "Mr. Art" wrote:
> 
> > Yes.
> 
> How Mr Art.
> 
> Pabs

Post a reply to this message

"Mr. Art" wrote:

> This short bit of code shows how to use a
> planetary image as a displacement map on a sphere.
> I know that this is displacement mapping only on a
> sphere, but the same concept should work on other shapes.
> I hope that this helps.

I was thinking more along the lines of a patch which used isosurfaces to
implement displacement mapping

Post a reply to this message

Pabs wrote in message <38F3D600.A087CAF2@hotmail.com>...
>"Mr. Art" wrote:
>
>> This short bit of code shows how to use a
>> planetary image as a displacement map on a sphere.
>> I know that this is displacement mapping only on a
>> sphere, but the same concept should work on other shapes.
>> I hope that this helps.
>
>I was thinking more along the lines of a patch which used isosurfaces to
>implement displacement mapping


How about the following code:

#declare AnObject = union{
    sphere{<0,0,0>,1}
    torus{1.3,0.5}
}

#declare AnObjectFunc = pigment{
    object{
        AnObject
    }
    color rgb 0, color rgb 1
}

isosurface{
    function{AnObjectFunc+noise3d(x,y,z)}
    contained_by{box{min_extent(AnObject),max_extent(AnObject)}
    method 2
    max_gradient 100
    pigment{rgb 1}
}

I can't guarentee that this will work on the first try, but it will displace
the surface of any object using the noise3d function.  To use a different
object, just replace AnObject with the object you want.

Mark

Post a reply to this message

The syntax is all wrong for AnObjectFunc.


Mark Wagner wrote:

> How about the following code:
> 
> #declare AnObject = union{sphere{<0,0,0>,1}torus{1.3,0.5}}
> 
> #declare AnObjectFunc = pigment{object{AnObject} color rgb 0, color rgb 1}
> 
> isosurface{
>     function{AnObjectFunc+noise3d(x,y,z)}
>     contained_by{box{min_extent(AnObject),max_extent(AnObject)}
>     method 2
>     max_gradient 100
>     pigment{rgb 1}
> }
> 
> I can't guarentee that this will work on the first try, but it will displace
> the surface of any object using the noise3d function.  To use a different
> object, just replace AnObject with the object you want.
> 
> Mark

Post a reply to this message

> #declare AnObjectFunc = pigment{
>     object{
>         AnObject
>     }
>     color rgb 0, color rgb 1
> }

Are you sure this works - I don't think it will unless you are thinking about
the proximity pigment func

Post a reply to this message

In article <38f40526@news.povray.org>, "Mark Wagner" 
<mar### [at] gtenet> wrote:

> I can't guarentee that this will work on the first try, but it will 
> displace
> the surface of any object using the noise3d function.  To use a different
> object, just replace AnObject with the object you want.

I think you mean:
#declare AnObject = union {
    sphere {<0,0,0>, 1}
    torus {1.3, 0.5}
}

#declare AnObjectFunc = 
function {
    pigment {
        object {AnObject color rgb 0, color rgb 1}
    }
}

isosurface {
    function {AnObjectFunc(x,y,z)
        +noise3d(x,y,z)
    }
    method 2
    max_gradient 100
    contained_by {box {min_extent(AnObject),max_extent(AnObject)}}
    pigment {color rgb <1,1,1>}
}

And the object pattern turns out to not be very good for 
isosurfaces...it has two values separated by a sharp boundary, sort of 
an infinite amount of slope. A boundary can appear anywhere, so the 
isosurface algorithm has trouble finding them. And I don't think you 
could displace the surfaces from within the function... I have a pattern 
blurring feature I have been thinking of which might reduce this 
problem, but I am not sure...
The proximity pattern might be helpful for this, but it is slow. You 
might want to make a script that uses eval_pattern() to write the 
results of proximity to a 3D density file, and use that in the 
isosurface.
If your object is a blob, the blob pattern is useful. I have used it 
before with very good results.

Also, don't forget that when the function goes through a pigment, you 
can use warps like turbulence, black holes, etc to displace it. And with 
function patterns and pigment functions, you can put any function 
through a pigment. :-)

-- 
Christopher James Huff - Personal e-mail: chr### [at] yahoocom
TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
Personal Web page: http://chrishuff.dhs.org/
TAG Web page: http://tag.povray.org/

Post a reply to this message

> I was thinking more along the lines of a patch which used isosurfaces to
> implement displacement mapping

I was thinking about it & came up with the attached image
This is explained as follows
1 the ray is traced to the surface normally
2 from the point of intersection take a point d (the displacement value for
that poin in space) units along the normal for that point of intersection
this is the center of a 5*5 rectangle of points used to make a bicubic
surface (or something like that)
3 the other points are found by tracing back to the surface and then going
up the normals
4 the ray is reflected off the bicubic surface

There are a lot of problems with this but it is a start
1 It will be slow (as slow/slower as/than proximity pigment??)
2 at small angles of intersection the ray may go under the surface and miss
it
maybe the points could be got by shooting from a point along the ray instead

3 when he ray doesn't hit the original surface (but it should hit the
displaced surface) this means that edges of an object will be smooth as the
original object - maybe we could use a proximity function added to a
displacement function and convert it to an isosurface... or we could just
tesselate the whole primitive into bicubic patches...

Post a reply to this message

Attachments:
Download 'temp.gif' (4 KB)

Preview of image 'temp.gif'