[vtkusers] vtkOpenGLGPUVolumeRayCastMapper - issue with texture streaming

[pegga]

Dear All,

I am facing a problem when trying to render "Big" data using
vtkOpenGLGPUVolumeRayCastMapper.
My data are usually quite large and as such, there is no way it fits GPU
memory.

My solution for now

1) Use vtkImageResample to down sample the data so that it fits GPU memory.
This mode will
be used for interactivity (fast rendering). There is a problem with
GetReductionRatio method
of vtkOpenGLGPUVolumeRayCastMapper which is not implemented in latest VTK
releases (since OpenGL2 it seems), but this is not a big problem.

2) Use SetPartitions method on vtkOpenGLGPUVolumeRayCastMapper. Using this
method, one can render image data block by block. By choosing correct number
of blocks in all dimensions, I can render whole data set in native
resolution. This is especially useful when zoomed in so that I can see all
the details.
Composite blend seems to work well, but Maximum intensity blend is not
working correctly. The problem is there are visible blocks in resulting
image. See uploaded images for details.
https://ibb.co/e9C3QG (visible blocking)
https://ibb.co/mC2Reb (rendered without SetPartitions)

What can I do to fix the problem? Is there some setting parameter I am
missing of is it necessary to fix a bug in mapper itself? Could someone give
me a hit where to start with fixing this issue?

Best Regards,
Petr



--
Sent from: http://vtk.1045678.n5.nabble.com/VTK-Users-f1224199.html