[Paraview] x3d geometry export garbled?

[Cook, Rich]

I have discovered that this is connected with the Miranda reader.  I believe the actual culprit is the "Cell Data to Point Data" filter.  In that step, somehow the x3d exporter fumbles somewhere.  Just a theory.  Without any developer feedback, I'm going to just make stuff up.  :-)
When I export the geometry from my scene and then re-import it, then save the scene to x3d, the colors are properly preserved and blender can deal with it.
I now have a path forward.  It's ugly, but if I can put it into a script once then fuggetaboutit it's done.  That's a pretty big "if" but I think
Thanks for listening to my podcast folks!  I'm here each week.
-- Rich

Summary:

1.  First create the desired contours in Paraview using parallel rendering.  How to:
    A.  Open the plot.mir file in ParaView after connecting in parallel.
    B.  Enable the variables you want to have available to contour and color by; click Apply
    C.  Since paraview views Miranda data as being zonal, to create a contour of Miranda data, you have to first apply the "Cell Data to Point Data" filter.   Enable "Pass Cell Data" to allow the next step to color by nodal values, which looks better than zonal, given the contour is nodal.   click Apply
    D.  Apply a Contour filter by clicking the onion shell in the toolbar or with Filters->Contour.  Click "Compute Scalars".  Click Apply.  You can now change the Coloring to be by another variable.

2.  Choose File --> "Save Geometry".  This will save all timesteps, so be patient.

3.  Import the saved geometry again and then export each timestep as an x3d file.  With many timesteps, you might want to script the tedious part.



On Nov 19, 2013, at 11:45 AM, "Cook, Rich" <cook47 at llnl.gov<mailto:cook47 at llnl.gov>> wrote:

I have looked closer at the image.  What first looks like the tiles are out of order is not the case.  What I believe is actually happening is that a different color map is being applied to each tile.  So the problem is in the "colors" section of the x3d file.  You can see this by looking at the patterns across the "tile" boundaries in the attached image.  The patterns are continuous but there is a discontinuity in the colors being applied.  For example, the green square in the image surrounds 9 tiles which share such a pattern in the image, kind of shaped like an ameoba.
I suspect that the part of the paraview code which assigns colors is not savvy about the idea that there are 16 processors and so is using a local coloring scheme.

<BlenderScreenSnapz001_copy-2.jpg>

On Nov 18, 2013, at 4:50 PM, Richard Cook <cook47 at llnl.gov<mailto:cook47 at llnl.gov>> wrote:

Hi,
I have been trying to get paraview to play nice with blender.  Specifically, I'd like to export an x3d file which blender will draw nicely.
I was able to do so from a dataset when reading the data locally, but to do so requires some awful contortions.  I have to export the data first from VisIt as a set of VTK files, then read those in with paraview, then export as x3d, so blender imports them as a set of objects, and it is difficult to color all those pieces in blender.  So I'd like to have a more direct way to do this.
What happens in parallel is that if I "Export Scene" of a contour that looks like this in paraview:
<ParaViewScreenSnapz003 copy.jpg>


The resulting x3d object displays like this in blender:


<BlenderScreenSnapz001 copy.jpg>

I believe that this means that Paraview is exporting the data out of order.  But there may be ambiguity in some specification.  Can someone help with this?
Thanks!

--
✐Richard Cook
✇ Lawrence Livermore National Laboratory
Bldg-453 Rm-4024, Mail Stop L-557
7000 East Avenue,  Livermore, CA, 94550, USA
☎ (office) (925) 423-9605
☎ (fax) (925) 423-6961
---
Information Management & Graphics Grp., Services & Development Div., Integrated Computing & Communications Dept.
(opinions expressed herein are mine and not those of LLNL)




--
✐Richard Cook
✇ Lawrence Livermore National Laboratory
Bldg-453 Rm-4024, Mail Stop L-557
7000 East Avenue,  Livermore, CA, 94550, USA
☎ (office) (925) 423-9605
☎ (fax) (925) 423-6961
---
Information Management & Graphics Grp., Services & Development Div., Integrated Computing & Communications Dept.
(opinions expressed herein are mine and not those of LLNL)




--
✐Richard Cook
✇ Lawrence Livermore National Laboratory
Bldg-453 Rm-4024, Mail Stop L-557
7000 East Avenue,  Livermore, CA, 94550, USA
☎ (office) (925) 423-9605
☎ (fax) (925) 423-6961
---
Information Management & Graphics Grp., Services & Development Div., Integrated Computing & Communications Dept.
(opinions expressed herein are mine and not those of LLNL)



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20131120/52e8ef5f/attachment-0001.htm>