I can't seem to apply a material to a group in lops - That at the sop level had the group created after importing bgeo files with a file cache.

Although I can apply a material in lops to the bgeos themselves.

However, if I stash those bgeos and create a group after, I can apply material to the group in lops.

Hmm, the material primitive and material binding is authored correctly in both cases.

It looks like a problem related to the time-dependent nature of the imported geometry. On the sopimport1 LOP, if you change Import Data > Author Time Samples to Never, then it works.

It looks like a problem related to the time-dependent nature of the imported geometry.

Yes it does because in playing around with it I found I could reproduce the issue with the file import node.

There is the option to either import as a sequence or choose a single individual file/frame.

I found it doesn't matter what source of bgeo's I use, it only depended if I choose a sequence or single file/frame.

I guess your solution is good, if that option was what it was intended for only? Otherwise there might be an issue if setting it to never is not an option for other case scenarios? I don't know since I don't fully understand by the description if this is the only case it was meant for.

Prior to your posting I had submitted a bug with a more simplified hip of the above one I posted. There I have stash node removed and whatever choice of bgeo one uses to see the effect, one can ‘toggle’ between a sequence or single file/frame on the import node to see the issue.

I included a simple set up just to generate some animated spheres to export and import for those curious.

rafal
Hmm, the material primitive and material binding is authored correctly in both cases.

It looks like a problem related to the time-dependent nature of the imported geometry. On the sopimport1 LOP, if you change Import Data > Author Time Samples to Never, then it works.

Is this related to the change made recently to allow authoring geometry subset indices timesamples, or are the materials assigned directly to mesh/transform prims?

The material is assigned correctly to the “Loop_A” geo subset. It's just that the material is not displayed correctly. So this is hydra or render delegate issue.

I reported this issue to Pixar, along with a pull request to fix it: https://github.com/PixarAnimationStudios/USD/issues/1058 [github.com]
We should be able to incorporate this fix into Houdini's USD tool at some point, but I'm not sure exactly what the timing will be.