This seems like a strange one. I’m hoping someone with more dataprep experience can enlighten me as to what might be going on.
We have created a blueprint for the visual dataprep system that goes through imported datasmith materials from 3ds max. It attempts to identify the type of material based on naming conventions, parameters/textures used, then it will reparent the material instance to one of our optimized master materials and try to carry over the parameter settings and texture assignments.
That portion seems to be working correctly when executed. When the files are committed all materials and textures are showing up and assigned correctly.
We also have an optional portion that renames the material instances to conform to our unreal naming standards by adding a prefix or suffix to the filename. EX: putting _MI at the end of all material instances. If this is used its done at the end of the blueprint after everything else is completed and its using the ‘rename asset’ node from the editor utility library.
If this is used and committed, the materials are present, with the adjusted names, the correct parent material and assigned to the correct objects.
The strange part is when the rename portion is used, any textures that were carrying over from the datasmith material are not being copied over to the content folder during the commit action. the result is the material will initially appear correct, with the texture of the correct name assigned, but when you navigate to the texture folder it displays, its not actually there. Additionally the next time you re-open the project, that texture assignment will be null since it doesn’t actually exist.
I’ve been going through the script and I’m not seeing any obvious correlation between renaming materials and their texture references. Any insight or ideas would be helpful.
