[GRASS5] Voxel progress

[benducke at compuserve.de]

Helena,

The data included in the slovakia3d set works fine,
the 3d region is in this case preset to 0.0-4000.0.
However, if I change the 3d settings, I run into
the same problem:

g.region b=3000.0
r3.mapcalc test=1.0

loading the new volume "test" into nviz places it at the
wrong z-location (0.0 - 1000.0 instead of 3000.0 - 4000.0).
Slices visualisation gives me the expected result, although
at the wrong location.

If I reset the region's z-range to 0.0 - 4000.0,
nviz shows "test" extents to be 0.0 - 4000.0.
If I slice through the data, however it appears at
the RIGHT position (3000.0 - 4000.0). The rest of
the z-range is displayed as NULL (btw., it would be
nice to have transparent NULL voxels).

I hope this will help to track down the problem.

Best wishes,

Benjamin


----- Originalnachricht -----
Von: Helena Mitasova <hmitaso at unity.ncsu.edu>
Datum: Samstag, 11. März 2006 2:56 pm
Betreff: Re: [GRASS5] Voxel progress

> 
> 
> >
> > HOWEVER, I fail to visualise such a layer model in NVIZ.
> > The first problem is that NVIZ seems to be unable to correctly
> > position volumes that have a bottom > 0.0.
> 
> Ben,
> 
> have you tried whether the current nviz volumes work with the  
> Slovakia data set?
> http://mpa.itc.it/grasstutor/data_menu2nd.phtml
> 
> When the volume support was written the 3D region definition was  
> separate from the
> 2D region - later they were merged and it is possible that the nviz 
> 
> volume support has not been
> adjusted properly. Also, this capability has not been tested much 
> so  
> there could be bugs.
> So let me know whether you see the same problems with the Slovakia  
> data set and we
> can try to track it down.
> 
> Helena
> >
> > You can test this even with a very primitive volume
> > (use the 3d region settings as shown above):
> >
> > r3.mapcalc test=1.0
> >
> > If you load this into NVIZ (and set the bottom of the 3d region > 
> 
> > 0.0),
> > it will not display the extents of the volume as it should.
> > If you start rotating the view around a bit and increase
> > height enough, the extents will be temporarily visible and you 
> will  
> > see
> > that the whole model has been
> > displaced from its original z-range 9.0 to 12.0 and is displayed
> > at 0.0 to 3.0!
> >
> > Obviously, it is not possible to display slices or isosurfaces in 
> 
> > such a
> > model, as the NVIZ extent and the real data extent do not match.
> > So this seems to be a bug in NVIZ(?).
> >
> > But here is what really puzzles me: if I adjust 3d region 
> settings and
> > set the bottom to 0.0, NVIZ shows the correct extent, but has the 
> same> problems!
> > I.e. I only get a display of the extent as wireframe for as long as
> > rotate the view, change the height or whatever.
> >
> > Everything disappears after that and I cannot display slices or  
> > isosurfaces.
> >
> > is it possible that NVIZ has trouble with small 3d regions or 
> high 3d
> > resolution?
> > Am I doing something totally silly?
> >
> > Finally, some points that I could not figure out from any  
> > documentation
> > and am still a bit unsure about. Maybe someone could reaffirm my
> > conclusions about these:
> >
> > 1. GRASS Z coordinates increase towards the top of the region
> > 2. r3.in.ascii expects the bottom slice first in the ASCII input 
> file>
> > I would so much like to get this working. It would be tremendously
> > useful for archaeological applications of GRASS GIS!
> >
> > Best,
> >
> > Benjamin
> >
> > _______________________________________________
> > grass5 mailing list
> > grass5 at grass.itc.it
> > http://grass.itc.it/mailman/listinfo/grass5
> 
>