[Paraview] Offset between grid and points on spherical projection

[Moreland, Kenneth]

I don't know why the lighting is different, but here some guesses you can try.

The default resolution for the sphere is quite coarse and results in large facets. The lighting on the large flat facets of the sphere might be quite different than the smaller facets of the grid. Try increasing the resolution of the sphere.

The sphere source comes with normals on the points whereas your grid might not. That could result in different lighting equations. You could try running the generate surface normals on your grid (assuming it is poly data). That's assuming that your grid is not a clipped sphere, which could result in inconsistent normals as the edge.

It could be that the sphere is using specular highlighting whereas the colored mesh is not. Try turning off specular highlighting on the sphere (under the display properties).

-Ken

From: Fernando Paolo <fspaolo at gmail.com<mailto:fspaolo at gmail.com>>
Date: Thursday, June 26, 2014 2:29 PM
To: Kenneth Moreland <kmorel at sandia.gov<mailto:kmorel at sandia.gov>>
Cc: "paraview at paraview.org<mailto:paraview at paraview.org>" <paraview at paraview.org<mailto:paraview at paraview.org>>
Subject: [EXTERNAL] Re: [Paraview] Offset between grid and points on spherical projection

Ok, that seems to work for the points. Thanks so much, Ken.

Not sure if this is also related to a similar issue. But when I add a 'Source -> Sphere' of radius = 1 (same radius as the grid) to the plot, there is different "illumination" between the surface of the sphere and the grid. This "effect" also changes with angle of view. In the attached figure you can clearly see how the illumination changes across the boundary between the surface of the sphere (lighter) and the grid (darker). This two surfaces should blend.

Once again, thank you!
-fernando


[Inline image 1]



On Thu, Jun 26, 2014 at 12:45 PM, Moreland, Kenneth <kmorel at sandia.gov<mailto:kmorel at sandia.gov>> wrote:
This looks like it might be caused by the coincident topology resolution feature in ParaView's rendering. The problem when trying to draw surfaces and points or lines on that surface as you are doing is that you get z-buffer fighting so that the points/lines don't just show up on top like you want.

ParaView implements a couple of tricks to prevent this from happening, but they all sometimes cause shifts in the data so that they do not line up correctly. My suggestion is to first turn off coincident topology resolution and see if that fixes this problem (although it might introduce a new one). Go to settings and then Render View -> General to get to the coincident topology resolution options.

-Ken

From: Fernando Paolo <fspaolo at gmail.com<mailto:fspaolo at gmail.com>>
Date: Thursday, June 26, 2014 1:36 PM
To: "paraview at paraview.org<mailto:paraview at paraview.org>" <paraview at paraview.org<mailto:paraview at paraview.org>>
Subject: [EXTERNAL] [Paraview] Offset between grid and points on spherical projection

Hello,

I have a 2d grid projected on top of a spherical surface. On the same surface, and on top of the grid, I am also projecting a set of points delimiting the boundaries of the quantity in question. The problem is that both the grid and the points when plotted together are displayed with an offset in between them (in the "r" direction from the spherical projection). Also interesting is the fact that when I change the angle of view the "offset" also changes (see attached figures).

The coordinates of both the grid and the points are simply converted from spherical lon/lat/radius=1 to cartesian xyz.

The points are projected using an UNSTRUCTURED GRID format (polydata ascii file). And the grid is projected using the XDMF 3DSMesh, with the following template:

   <Grid Name="Mesh" GridType="Uniform">
     <Time Value="%f" />
     <Topology TopologyType="3DSMesh" NumberOfElements="{0} {1} {2}"/>
     <Geometry GeometryType="XYZ">
       <DataItem Name="Coordinates" Dimensions="{3} {4}" NumberType="Float" Precision="4" Format="HDF">
         PATH_TO_XYZ
       </DataItem>
     </Geometry>

     <Attribute Name="DATA_NAME" AttributeType="Scalar" Center="Cell">
       <DataItem Dimensions="{5} {6} {7}" NumberType="Float" Precision="4" Format="HDF">
         PATH_TO_DATA
       </DataItem>
     </Attribute>
   </Grid>

If both datasets have a radius = 1, why is there any offset in between them? Since this offset also changes with the angle of view, it seems that both datasets are not being projected on top of the same spherical surface (note: I've tried setting different radius for both datasets). How can I fix this?

I appreciate any help!
-fernando


[Inline image 1]

[Inline image 2]


--
Fernando Paolo
Institute of Geophysics & Planetary Physics
Scripps Institution of Oceanography
University of California, San Diego

web: fspaolo.net<http://fspaolo.net>



--
Fernando Paolo
Institute of Geophysics & Planetary Physics
Scripps Institution of Oceanography
University of California, San Diego

web: fspaolo.net<http://fspaolo.net>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20140626/d7e50dc7/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen shot 2014-06-26 at 11.54.19 AM.png
Type: image/png
Size: 26982 bytes
Desc: Screen shot 2014-06-26 at 11.54.19 AM.png
URL: <http://public.kitware.com/pipermail/paraview/attachments/20140626/d7e50dc7/attachment-0003.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen shot 2014-06-26 at 11.50.28 AM.png
Type: image/png
Size: 24539 bytes
Desc: Screen shot 2014-06-26 at 11.50.28 AM.png
URL: <http://public.kitware.com/pipermail/paraview/attachments/20140626/d7e50dc7/attachment-0004.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen shot 2014-06-26 at 1.16.10 PM.png
Type: image/png
Size: 30674 bytes
Desc: Screen shot 2014-06-26 at 1.16.10 PM.png
URL: <http://public.kitware.com/pipermail/paraview/attachments/20140626/d7e50dc7/attachment-0005.png>