vrml & volume-rendering - vrml

> i just read that vrml is/was[?] also used for volume-rendering...

Not really. Either it is a fake using stacks of surfaces or a non standard
browser extension.

> - which nodes will do this?

AFAIK in the X3D standard (X3D classical VRML encoding looks very similar to
VRML97), there are (plans to include ?) 3D textures...

> - which software is able to render a volumetrical vrml

http://www.bitmanagement.de/products...tact_vrml/6.2/

AFAIK this is a nonstandard extension (yet !)

> - is it possible to use both polygonal & volumetric rendering in one
> worldfile?

I do not know this bitmanagment contact version exactly, but i would wonder
if it would be not possible...

so long
MUFTI
--
Warning: If a process happens to return STILL_ACTIVE (259) as an error code,
applications that test for this value could end up in an infinite loop.
(Aus einer Micro$oft Dokumentation fuer GetExitCodeProcess)

Joerg Scheurich aka MUFTI schrieb:

>>i just read that vrml is/was[?] also used for volume-rendering...
>
>
> Not really. Either it is a fake using stacks of surfaces or a non standard
> browser extension.

hmmm I just read an article where - in my eyes just - a cylinder draped
with an imageTexture called a volumetric rendered object below a surface
(ElevationGrid)....the work was done with vrml....
so I was wondering about the terms they use...or the kind of software
either the abilities of vrml which AFAIK describing only polygonal-structurs


>
>
>>- which nodes will do this?
>
>
> AFAIK in the X3D standard (X3D classical VRML encoding looks very similar to
> VRML97), there are (plans to include ?) 3D textures...

[next] hmmm are 3d-textures a feature of the volumetric-rendering??? I
did never try them either read more about them all I realized is
that they would be nicer to drape [using coord-triples instead of e.g.
uv-mapped tuples] and better for interactiv behaviors [moving shadows,
reflections and so far...

>
>
>>- which software is able to render a volumetrical vrml
>
>
> http://www.bitmanagement.de/products...tact_vrml/6.2/
>
> AFAIK this is a nonstandard extension (yet !)
oh this looks very interesting, but I can't find any description how
this is made...hmm it seems to be that i have go deeper in that site
>
>
>>- is it possible to use both polygonal & volumetric rendering in one
>>worldfile?
>
>
> I do not know this bitmanagment contact version exactly, but i would wonder
> if it would be not possible...

best thanks!
I'll have a look
kindly regards
rolf
>
> so long
> MUFTI

Hi Rolf,

I think there are many different techniques for volume rendering.

E.g. raycasting, extraction of surfaces using marching cubes etc.

Depends on what you want.

With 3d graphics hardware one can render polygonal slices
through a volumetric texture and blend the layers together.
With a pixel shader one can compute a custom color value per pixel.
I made a simple example here
http://www.bitmanagement.com/develop...er/shader.html
see Volume Texture
I think in X3D amendement there is now a specification for a Volume texture
node, but it defines a Volume Texture as a big set of planar ImageTextures.


Greetings
Holger



"rolf gabler-mieck" <gabler.mieck@geographie.uni-kiel.de> schrieb im
Newsbeitrag news:eid5m0$ijb$1@inlux3.rz.uni-kiel.de...
> Joerg Scheurich aka MUFTI schrieb:
>
>>>i just read that vrml is/was[?] also used for volume-rendering...
>>
>>
>> Not really. Either it is a fake using stacks of surfaces or a non
>> standard
>> browser extension.
>
> hmmm I just read an article where - in my eyes just - a cylinder draped
> with an imageTexture called a volumetric rendered object below a surface
> (ElevationGrid)....the work was done with vrml....
> so I was wondering about the terms they use...or the kind of software
> either the abilities of vrml which AFAIK describing only
> polygonal-structurs
>
>
>>
>>
>>>- which nodes will do this?
>>
>>
>> AFAIK in the X3D standard (X3D classical VRML encoding looks very similar
>> to VRML97), there are (plans to include ?) 3D textures...
>
> [next] hmmm are 3d-textures a feature of the volumetric-rendering??? I did
> never try them either read more about them all I realized is
> that they would be nicer to drape [using coord-triples instead of e.g.
> uv-mapped tuples] and better for interactiv behaviors [moving shadows,
> reflections and so far...
>
>>
>>
>>>- which software is able to render a volumetrical vrml
>>
>>
>> http://www.bitmanagement.de/products...tact_vrml/6.2/
>>
>> AFAIK this is a nonstandard extension (yet !)
> oh this looks very interesting, but I can't find any description how this
> is made...hmm it seems to be that i have go deeper in that site
>>
>>
>>>- is it possible to use both polygonal & volumetric rendering in one
>>>worldfile?
>>
>>
>> I do not know this bitmanagment contact version exactly, but i would
>> wonder
>> if it would be not possible...
>
> best thanks!
> I'll have a look
> kindly regards
> rolf
>>
>> so long
>> MUFTI

Holger Grahn schrieb:
> Hi Rolf,
>
> I think there are many different techniques for volume rendering.
>
> E.g. raycasting, extraction of surfaces using marching cubes etc.
>
> Depends on what you want.
>
> With 3d graphics hardware one can render polygonal slices
> through a volumetric texture and blend the layers together.
> With a pixel shader one can compute a custom color value per pixel.
> I made a simple example here
> http://www.bitmanagement.com/develop...er/shader.html
> see Volume Texture
> I think in X3D amendement there is now a specification for a Volume texture
> node, but it defines a Volume Texture as a big set of planar ImageTextures.
>
>
> Greetings
> Holger
>
Hi Holger,
bets thanks! I have the task to check how I can realize a piece of earth
[dgm with depth-information] for a walkthrough like a worm...
through an interpolated fog-like-area
primary the goal is to implement this in java3d, but I stumbeld over an
article, which tells about vrml and volumerendering...
so it seem to be that we plan something like a tomographie for parts of
the earth....I got an idea to do this in the polygonal model...but it
seems that others prefer the volumetric model [at the moment I can't
really see any advantage...]
at the moment I am glad to have a look at your works! very imprsessive
and good!
best regards
rolf

rolf gabler-mieck wrote:

>at the moment I can't
>really see any advantage...

The advantage of ray-casting VR is an authentic data representation;
specifically, the authentic iso-surface representation is getting
increasingly cheaper by ray-casting means then by myriads of tiny
polygons. Plus interactive threshold manipulation with no-preprocessing
is hardly an option for polygonal presentation. Additionally; the
medium-high opacity transfer function setting allows enriching
iso-surface representation since ray "interacts with undersurface"
what provides scattering effect - different density distribution along
ray path (within thin film) acquired different RGB output; as the
result the iso-surface colorization gives visual clue about underling
"tissue". Definitely, "polygonal way" can be used to get the
same visual result but it is the same as usage of sewing machine as
stapler.



> Holger Grahn schrieb:
> > Hi Rolf,
> >
> > I think there are many different techniques for volume rendering.
> >
> > E.g. raycasting, extraction of surfaces using marching cubes etc.
> >
> > Depends on what you want.
> >
> > With 3d graphics hardware one can render polygonal slices
> > through a volumetric texture and blend the layers together.
> > With a pixel shader one can compute a custom color value per pixel.
> > I made a simple example here
> > http://www.bitmanagement.com/develop...er/shader.html
> > see Volume Texture
> > I think in X3D amendement there is now a specification for a Volume texture
> > node, but it defines a Volume Texture as a big set of planar ImageTextures.
> >
> >
> > Greetings
> > Holger
> >
> Hi Holger,
> bets thanks! I have the task to check how I can realize a piece of earth
> [dgm with depth-information] for a walkthrough like a worm...
> through an interpolated fog-like-area
> primary the goal is to implement this in java3d, but I stumbeld over an
> article, which tells about vrml and volumerendering...
> so it seem to be that we plan something like a tomographie for parts of
> the earth....I got an idea to do this in the polygonal model...but it
> seems that others prefer the volumetric model [at the moment I can't
> really see any advantage...]
> at the moment I am glad to have a look at your works! very imprsessive
> and good!
> best regards
> rolf

stefanbanev@yahoo.com schrieb:
> rolf gabler-mieck wrote:
>
>
>>at the moment I can't
>>really see any advantage...
>
>
> The advantage of ray-casting VR is an authentic data representation;
> specifically, the authentic iso-surface representation is getting
> increasingly cheaper by ray-casting means then by myriads of tiny
> polygons. Plus interactive threshold manipulation with no-preprocessing
> is hardly an option for polygonal presentation. Additionally; the
> medium-high opacity transfer function setting allows enriching
> iso-surface representation since ray "interacts with undersurface"
> what provides scattering effect - different density distribution along
> ray path (within thin film) acquired different RGB output; as the
> result the iso-surface colorization gives visual clue about underling
> "tissue". Definitely, "polygonal way" can be used to get the
> same visual result but it is the same as usage of sewing machine as
> stapler.
>
>
>
>
[] hi stefan,
thanks for you extended answer, nice for have a look...
it wasn't a goal to make myriads of polygones...just those which could
be seen. subearth isn't transpareny. so no deeper look will be necesary
- in my eyes -
the thing is that I have to display also geoobjects like trees &
buildings and so far at the surface [which making no sense in
volumetric-matter]
my idea was more to have the surface as hightfield and many datapoints with
depthinformations, which has to be interpolatet with an non-linear
algorythm [idw or e.g krieging]
when any user click a path on the surface that path should be digged and
shown, so the amound of polygones is depending on the resolution of the
data...the interpolation along that path should produce a rasterimage
which will be draped at those digged-walls.

an other idea was to set many[...] cubes [one for each sure-resoultion ]
with an instance of fog depending on the interpolated colorvalue,
but I never try to set more than 10 fog nodes

A view through the earth will produce just an strange pattern which will
be not interpretable...

all work will be done using java&java3d...
but thats the point why posting here my java-app reads vrml files and I
realized
the possibility of vrml to do something using volume-rendering in any node.
do you have any further hints or links to get inform about the
advantages of volumerendering

so the tomography wasn't a correct comparement,
best regrads
rolf

[]

>>>
>>
>>Hi Holger,
>>bets thanks! I have the task to check how I can realize a piece of earth
>>[dgm with depth-information] for a walkthrough like a worm...
>>through an interpolated fog-like-area
>>primary the goal is to implement this in java3d, but I stumbeld over an
>>article, which tells about vrml and volumerendering...
>>so it seem to be that we plan something like a tomographie for parts of
>>the earth....I got an idea to do this in the polygonal model...but it
>>seems that others prefer the volumetric model [at the moment I can't
>>really see any advantage...]
>>at the moment I am glad to have a look at your works! very imprsessive
>>and good!
>>best regards
>>rolf
>
>