Data type
Currently, the framework only supports data stored as raw binary. The precision can be either float or double.
For the best performance during visualization, it is recommended to use data in float (unless for some reason you really need float).