Level set use case and Field3D, containers complexity

[Kirill Lykov]

Dear Field3D users,

I'm very new in the Field3D so have a few newbie questions:

1) I'm looking for a solution to solve the following use case: create in a CAD or 3D animation software a geometry, save it as a levelset, open this file in my application and do some fluid-related stuff inside this geometry.

Can I solve this use case using Field3D? I'm I right that Field3D data format is supported by Maya from the box? If yes, do you know where I can read about it a bit more?

2) What is the complexity for random and sequential access for different Field3D containers? I looked through the pdf documents available at Field3D website but could not find an evaluation for that.

Best regards,

Kirill

[Sosh Mirsepassi]

1) I'm looking for a solution to solve the following use case: create in a CAD or 3D animation software a geometry, save it as a levelset, open this file in my application and do some fluid-related stuff inside this geometry.

Can I solve this use case using Field3D? I'm I right that Field3D data format is supported by Maya from the box? If yes, do you know where I can read about it a bit more?

Field3D is a file format which can store your levelset, however it doesn't have the algorithm to convert your geometry to levelset.  You need to calculate iso values on your own.

There is a Field3D exporter out of Maya plugin in the contribution section of the repository but in order for you to use the Field3D in Maya you need to write you own MPxFieldNode plugin that reads the f3d files.  

 

2) What is the complexity for random and sequential access for different Field3D containers? I looked through the pdf documents available at Field3D website but could not find an evaluation for that.

 

There are two basic fields types in Field3D: dense and sparse
Dense fields are just 3D containers of voxel data (e.g. 100x40x30) but saved internally as in a linear array. 
SparseField are by default in blocks of 8x8x8 data containers are more suitable for data's that are sparse in nature and don't fill up the entire grid.
Both of these containers have sequential and random access functions.