[Paraview] Paraview 3 parallel rendering artefacts (with Points, Wireframes and Volume representations)

[Moreland, Kenneth]

The artifacts for the points and wireframe are caused by a technical
detail.  They happen because the points and wireframe are actually taken
from the surface, and the surface extraction used in this instance is
simplified to extract the surface local to the current process.  Thus,
those planes are the process boundaries.  You will also see this with
surfaces rendered with opacity less than 1.  You can get around this
problem by running the Extract Surface, which will use ghost cells to do
the right thing.

The artifacts with volume rendering look like a real bug.  I have
submitted a bug report about it.

http://www.paraview.org/Bug/bug.php?op=show&bugid=5315&pos=

Thanks for bringing this to our attention.

-Ken

> -----Original Message-----
> From: paraview-bounces+kmorel=sandia.gov at paraview.org
[mailto:paraview-
> bounces+kmorel=sandia.gov at paraview.org] On Behalf Of Marc Baaden
> Sent: Saturday, July 07, 2007 12:20 PM
> To: paraview at paraview.org
> Subject: [Paraview] Paraview 3 parallel rendering artefacts (with
Points,
> Wireframes and Volume representations)
> 
> 
> (NB: this is a follow-up on a previous thread with several issues,
> focusing on one of them )
> 
> kmorel at sandia.gov said:
> >>> - there seem to be some artefacts for eg Volume Rendering or
Points
> >>> representation (presumably at the boundaries of the rendered
> >>> sub-images). Can these artefacts be avoided or corrected for?
> 
> >> It would be helpful to see the artifacts.
> 
> Ok, here are snapshots:
> http://www.shaman.ibpc.fr/pv3_arefact1.png
> http://www.shaman.ibpc.fr/pv3_arefact2.png
> http://www.shaman.ibpc.fr/pv3_arefact3.png
> http://www.shaman.ibpc.fr/pv3_arefact4.png
> 
> There are some kind of "ghost" planes that only appear when I use
> parallel rendering. The same data on a single node has none such
> planes. Furthermore the number of planes seems to depend on the number
> of processors: with two processors there is one plane dividing the
data
> into two parts, with 4 processors there are two planes dividing into
> four parts.
> I have seen this clearly with points, wireframe and volume
> representations. The artefacts depend on the viewing angle. One can
> make them disappear from some angle or make them more apparent
choosing
> the right direction to look from.
> Can it be avoided or corrected for?
> 
> Thanks in advance,
>   Marc Baaden
> 
> NB: the snapshots were made using paraview 3.0.2 on MacOSX running a
> server on 4 processors
> --
>  Dr. Marc Baaden  - Institut de Biologie Physico-Chimique, Paris
>  mailto:baaden at smplinux.de      -      http://www.baaden.ibpc.fr
>  FAX: +33 15841 5026  -  Tel: +33 15841 5176  ou  +33 609 843217
> 
> 
> _______________________________________________
> ParaView mailing list
> ParaView at paraview.org
> http://www.paraview.org/mailman/listinfo/paraview