Re: Java3D 1.3 Objects vanished AND resampling Irregular2DSet

> Hi Desiree,
> 
> If you have horizontal layers and look at them from the side,
> it will be hard to see actual information even if you make
> the layers thick - you will just see a sort of horzontal bar,
> sort of like looking at a table edge.

A table edge would be flat. I am interested in a picture where I can see
the run of the profiles. Sorry, I am not good in describing. I want that
it looks similar to the last picture (at the very bottom) in my homepage:

http://www-ipf.bau-verm.uni-karlsruhe.de/Personen/hilbring/welcome.html

It would be nice to fill up the gaps between the "lines", but I guess that
is not easy, so I was just thinking of thicking the "lines".

> 
> There are a bunch of tricks you can play with VisAD to get
> different views of data. Given a terrain surface ((x, y) -> z)
> you can map z to both ZAxis and IsoContour and ask for a
> contour line at a single z level, with a ConstantMap to
> LineWidth for thick lines. 

That would give me horizontal contour lines, right? Can I map that to
a combination of (x,y) to get a vertical image?


If you have a terrain layer plus
> multiple hydrogeological layers, you might use ConstantMaps
> to ZAxis to stack them up vertically, and then look at them
> from an oblique angle.

Sorry, I do not understand this idea.

> 
> Or you might have a number of FlatFields with MathTypes:
> 
>   ((x, y) -> (zN, layerN))
> 
> where each layer is on a terrain surface offset in the z
> direction by a constant amount from the original terrain.
> You could have ScalarMaps of each zN to ZAxis, all with
> indentical calls to setRange(), and ScalarMaps of each zN
> to SelectRange, where the ranges for each zN are offset
> by the amount of its offset from the original terrain. This
> would produce a stack of terrain slices, one for each hydro
> layer.

Please, can you clarify that for me I have troubles visualizing that in my
head.

Thank you very much for your help.

Desiree


> 
> Good luck,
> Bill
> 
> Desiree Hilbring wrote:
> > 
> > Hi Bill,
> > 
> > okay maybe I should tell your first my original intention.
> > I have 1 terrain layer and several hydrogeological layer in my 3D-view.
> > They are all created with VisAD, grabbed and included in my own Java3D
> > scenegraph. That is working fine. What I want know, is to define
> > horizontal slices through that scene, so that I get a kind of 2D-View,
> > where I can see the profile of the hydrological layers.
> > It is easy in Java3D to clip the 3D scene with the ModelClip function. But
> > unforetunately looking at the shapes directly from the side, makes them
> > almost invisible. So I thought using Volume rendering I would get a real
> > profile of the slice.
> > 
> > >
> > > You can construct an Irregular3DSet from your Irregular2DSet using
> > > the appropriate Irregular3DSet constructor signature, but it would
> > > have manifold dimension = 2 (because it is just a 2-D surface
> > > embedded in 3-D). You cannot resample from a 3-D Set with manifold
> > > dimension = 2.
> > 
> > That is, why I am wondering how or if it is possible to add the thickness
> > values to my Irregular2DSet.
> > My original data gives me the information of discrete points in 3D.
> > For example
> > 
> > Point 1 has x,y,z coordinates and the first depth layer ends at -1m,
> > the second depth layer ends at -2m ...
> > Point 2 has other x,y,z coordinates and the first layer end at -1.5m, the
> > second depth layer at -2.1m ...
> > ...
> > 
> > But volume rendering is probably not really what you
> > > want here, to get a good visual appearence. Volume rendering creates
> > > a sort of transparent fog, used when the user wants to get an overall
> > > view of volume-filling data.
> > 
> > Would that work for the slices?
> > 
> > But terrain data lies on a surface
> > > rather than filling a volume, so a crisp terrain surface is better.
> > >
> > > To give your terrain "a bit of thickness", perhaps create a second
> > > FlatField with the same domain Set as the first (your Irregular2DSet),
> > > but with z values less than the z values of the first FlatField by
> > > some constant. If you render both FlatFields, it will give two
> > > terrains with the same shape, a small distance apart.
> > 
> > But from the side, I would just see two times almost nothing than ont
> > times almost nothing. (My water layers are lying that way.)
> > 
> > >
> > > If you are really ambitous you can fill in the gap between the
> > > surfaces by a thin strip around the perimeter, as a Gridded3DSet
> > > with manifold dimension = 2. Its grid dimensions would be:
> > >
> > >   2 x number_of_perimeter_points
> > >
> > > The tricky part would be finding the perimeter points, in order, of
> > > your original Irregular2DSet.
> > 
> > I guess, that does not help me for the slices. I would get a kind of
> > unfilled polygon, right?
> > 
> > >   FlatField terrain_new = (FlatField) terrain.resample(...);
> > 
> > okay sorry, sometimes, I am just blind.
> > 
> > Thanks Desiree
> > >
> > > Cheers,
> > > Bill
> > > ----------------------------------------------------------
> > > Bill Hibbard, SSEC, 1225 W. Dayton St., Madison, WI  53706
> > > hibbard@xxxxxxxxxxxxxxxxx  608-263-4427  fax: 608-263-6738
> > > http://www.ssec.wisc.edu/~billh/vis.html
> > >
> 
> -- 
> ----------------------------------------------------------
> Bill Hibbard, SSEC, 1225 W. Dayton St., Madison, WI  53706
> hibbard@xxxxxxxxxxxxxxxxx  608-263-4427  fax: 608-263-6738
> http://www.ssec.wisc.edu/~billh/vis.html
>