[vtkusers] MPI Offscreen Volume Rendering

[Kevin H. Hobbs]

On Tue, 2006-05-16 at 11:22 -0400, Kevin H. Hobbs wrote:
> On Fri, 2006-05-12 at 23:00 +0200, John Biddiscombe wrote:
> > Now ideally the Reader would be able to accpet a "piece" description
> > which was a bounding box in space, the master controller would query
> > the file, and get the bounding box of everything, you could subdivide
> > space into regions, read regions into each process as pieces and then
> > render non overlapping using back to front sort on the top level actor
> > objects. All would be fine.
> > 
> 
> Yes, that's exactly what I want. Unfortunatly it's not easy. The tiling
> is driving me batty.

OK, as far as I can tell the renderer on each process renders exactly
the same view. This isn't that surprising as it's exactly what the
documentation says the composite render manager does. Tiling is not done
as far as I've been able to figure out. Tiling would require each camera
to have a slightly different view. I copied a bit of code out of the
vtkRenderLargeImage class, but after closer inspection of that class I
see that it wasn't even close to enough to correctly tile the images. It
looks like it shifts the actors, and when I try to copy that code I
either get nothing in the render window or exactly the same thing in
every render window.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: MPIVolumeRenderMovie.cxx
Type: text/x-c++src
Size: 8724 bytes
Desc: not available
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20060516/668f61ec/attachment.cxx>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20060516/668f61ec/attachment.pgp>