Clarification on facesets

[Pete Segal]

From what I've read it looks like an individual mesh can have multiple orthogonal face sets. If I understand this correctly, I think this is a good thing, as I can have one group of face sets associated with materials, for example, and and another associated with component or face id's.

I can see how this would work when I was reading files that I've exported from my application, but is there a convention for naming a face set for exchange between applications?

For example, the face set has a name, in which I might typically store a material name for that face set - allowing me to have multiple materials for different faces within a mesh. AFAICT there is no mechanism for identifying the type of face set, so i wouldn't be able to tell what the face set is intended for.

From this I conclude that all we're doing with face sets is creating a mechanism for associating multiple labels with faces, without ascribing any context to those labels.

Do I understand this correctly?

Thanks,

Pete

[Steve LaVietes]

Yes. A faceset definition is a name and a set of faces. While these
are often used (at our facility, at least) for things like associating
materials and textures, AbcGeom itself doesn't currently aim provide
further context.

One of the reasons that a faceset is a child object rather than just a
property of a mesh is for future growth. Theoretical conventions for
describing application-specific or facility-specific contexts (i.e.
AbcPrman, AbcModo) could be applied to either meshes or facesets.

-stevel

[Brian Leair]

Hi Pete,

That's a good description of what face sets currently do. Right now a SubD or a PolyMesh has a single flat list of all face sets defined for it.

I had internally debated the idea of "groups of related face sets" with Joe, but because it's a bit open ended we couldn't come up with a concrete proposal 
For now my thought or hope would be all of us here might document and use a naming convention to express the context. E.g. face sets used for materials could use names of the from "material_<user visible name>", or another example - face set names prefixed with  "rigging_partition_<user visible name>", or "shading_group_<user visible name>". The naming convention could then at least provide a hint... it would still be up to each application to do something "fancy" with the face sets, and right now for Maya each face set will just create an attribute on the object listing its the face numbers. I don't think the maya writer takes anything from your scene as expressing a face set, other than this attribute.   So, the current support is basic, but then one could write a set of small mel scripts to take those face numbers and create a shading group, or an exclusive partition, or whatever was needed. This seemed like a reasonable first stab at a fairly open ended concept.

Also, it seems likely that each studio will want to capture different data for what a face set expresses. As Steve said, we made the face sets objects so they have room to expand in the future.

cheers,
-brian

--
You received this message because you are subscribed to the Google
Groups "alembic-discussion" group.
To post to this group, send email to alembic-d...@googlegroups.com
To unsubscribe from this group, send email to
alembic-discuss...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/alembic-discussion?hl=en
 
For RSS or Atom feeds related to Alembic, see:
 
http://groups.google.com/group/alembic-dev/feeds
 
http://groups.google.com/group/alembic-discussion/feeds

[Pete Segal]

Thanks Brian,

 

I agree that it is important to establish a convention. I think it’s safe to assume that materials would be a common face grouping, and I would propose that we all agree to preface all material face sets with “material_”.

 

While many applications may have other types of tags they could attach to faces, we should at least be able to find consensus on materials.

 

Thanks,

Pete

[Steve LaVietes]

I'd be reluctant to establish any formal (within AbcGeom) conventions
based on object naming.

It's also not clear to me that we'd want to establish anything
material-related at the AbcGeom level as the definition of a
"material" varies greatly across applications and pipelines.

In my mind, the inclusion of facesets within AbcGeom is to allow faces
to be "addressable." What happens once you get there is beyond the
scope of AbcGeom.

I understand the interest in a standard categorization of facesets in
this case. If we wanted to consider more formally defining this as
meta-data or a property, I'd want to consider the expectations placed
on each client application when reading or writing the data.

-stevel