I have two pieces of geometry (a jasmine flower and a Mugwort leaf) being driven by some emitter geo.

When the emitter geo is just a grid, everything is fine. However, every time I try to use more complex geo to drive the emission (e.g. the pigs head instead of a grid) I no longer see my scattered geometry rendered properly in Solaris/Karma (though it exists in the normal scene view).

This can be temporarily fixed for instance, by severing the opacity image texture in the Mugwort leaf shader. However, although the geo appears shaded again, the diffuse of the Mugwort (which is also driven by an image texture) is wrong. This makes me suspect that the complexity of the emitter geo is having a strange effect on UV positions in Solaris, but I see no logic as to why.


I have tried the following:

- Turning on "Thin walled" opacity
- Changing the geometry of the mugwort leaf (remesh, thickening, even turning it into a basic cube, same issue pattern)
- Trying an object with different UVs

- I also tried this in Redshift instead of Karma, same issue.

josha_eif00l
This makes me suspect that the complexity of the emitter geo is having a strange effect on UV positions in Solaris, but I see no logic as to why.

I don't think it's necessarily the complexity
can be just the fact that your new geo (pighead) has UVs that the scattered point inherit and then you may be copying them on instances within Copy SOP

so maybe try deleting UVs before scattering, or don'r inherit them or don't copy them onto instances

but of course it's just a guess without hipfile

tamte
so maybe try deleting UVs before scattering, or don'r inherit them or don't copy them onto instances

You're right. I just needed to use a delete attribute on the pigs head to clean it of UVs. Thanks!