vtkOpenVDBReader should support opening AMR or HTG instead of only image data
Currently, the vtkOpenVDBReader reads multi resolution grid as a single image data, resulting and huge reading time and large memory usage.
Example with a sample file from https://www.openvdb.org/download/: https://artifacts.aswf.io/io/aswf/openvdb/models/buddha.vdb/1.0.0/buddha.vdb-1.0.0.zip
- Reading time: 59s
- Max memory usage: 1.6G
When reading it with vdb_view, a tool from OpenVDB, we get:
- Reading time: 0.5s
- Max memory usage: <1G
A solution would probably be to support reading the file a AMR grid, which is what vdb_view is doing and seems to be how one is supposed to read such a file:
FYI @acbauer