[GRASS-dev] fixes before 6.2.1 [was: ps.map: scale wrong and
benjamin.ducke at ufg.uni-kiel.de
Mon Nov 20 12:57:31 EST 2006
OK, I have done some experimenting and this is what it looks like for
It indeed seems to be a problem with small region settings.
Basically, anything with a cell resolution < 1.0 in any axis does not
seem to display (i.e. results gets totally displaced as illustrated
in the screenshot I sent earlier).
IN ADDITION, the number of rows, cols and depths must at least be
10! If any of those fall below, the result gets more or less strongly
Now, I am not sure if this is a problem with some sort of scaling or
translation, maybe something that also happens with larger cells and
region settings but goes unnoticed in those cases, because the offsets
shrinks to a very small value ???
I am not sure if that also makes sense in relation to Michael's report.
Helena Mitasova wrote:
> can you do a small test to confirm that the problem is indeed with
> small numbers by multiplying the coordinates in your area by 100 (so you
> get res=5) and see whether nviz volumes still don't work?
> The volumes support for nviz was implemented when the 3D region
> was handled separately from 2D region, after everything was merged into
> a single file it may not be reading the region correctly when the
> numbers are small,
> thanks Helena
> Benjamin Ducke wrote:
>> Just a few observations on nviz volume support, hope it provides some
>> useful informations.
>> For me, the slovakia3d volume has always worked with NVIZ. Also, when I
>> create a new volume in that location, I can vizualize it with NVIZ w/o
>> problems. E.g.:
>> r3.mapcalc tmp=1.0
>> works fine, also if I change the 3D region settings, e.g
>> g.region b=2000.0
>> r3.mapcalc tmp2=2.0
>> still displays OK.
>> HOWEVER, for my vizualization tasks I use very small, local coordinate
>> ranges (archaeological site stratigraphies of just a few cubic meters)
>> with very small 3d cell sizes.
>> E.g., I have a 3D region set to this:
>> projection: 0 (x,y)
>> zone: 0
>> north: 61
>> south: 58
>> west: 14
>> east: 17
>> top: 12.00000000
>> bottom: 9.00000000
>> nsres: 0.05
>> nsres3: 0.05
>> ewres: 0.05
>> ewres3: 0.05
>> tbres: 0.05
>> rows: 60
>> rows3: 60
>> cols: 60
>> cols3: 60
>> depths: 60
>> cells: 3600
>> 3dcells: 216000
>> Now, when I create a simple volume like above, It does not display in
>> nviz, I can neither create iso surfaces nor slices.
>> When I zoom out (increasing the "Height" slider) then NVIZ * sometimes *
>> shows the outlines of the volume and I can see, that it has been totally
>> displaced, below my 3d working region (see attached image)!
>> Apparently, there are some scale effects here and nviz cannot handle
>> volumes at these small region and resolution settings.
>> Another remaining problem (though not a bug) is that NULL values
>> in volumes cannot be set transparent independently of other 3d cell
>> values (not so sure if this is possible for 2d rasters in nviz).
>> Well, good luck fixing these and the other issues with nviz.
>> I think it would be tremendously important to finally have working
>> 3d raster vizualization in GRASS. I for one would be extremely
>> grateful for it.
>> Markus Neteler wrote:
>>> Maciej Sieczka wrote on 11/16/2006 07:07 PM:
>>>> Hamish wrote:
>>>>> So ps.map is pretty much broken in 6.2.0. Bummer.
>>>>> If we can figure out why NVIZ volumes are broken and fix the following
>>>>> bug, I'd vote for a quick 6.2.1 release.
>>>>> NVIZ volume problem (test with slovakia3d dataset) - sometimes it
>>>>> * fails from command line; ok if the volume added within NVIZ (bug
>>>>> It would be nice to fix this one too:
>>>> Fixing the gis.m, so that it preserved the aspect ratio for
>>>> single-click zooming in the restricted mode, is also a critical issue.
>>> There are some fixes which didn't make it yet into gis.m of 6.2. There
>>> was offlist communication between Michael, Hamish and me on that
>>> but I completely lost track which change should be submitted there.
>>> grass-dev mailing list
>>> grass-dev at grass.itc.it
>> grass-dev mailing list
>> grass-dev at grass.itc.it
Benjamin Ducke, M.A.
Institut für Ur- und Frühgeschichte
(Inst. of Prehistoric and Historic Archaeology)
Christian-Albrechts-Universität zu Kiel
D 24098 Kiel
Tel.: ++49 (0)431 880-3378 / -3379
Fax : ++49 (0)431 880-7300
More information about the grass-dev