Hi all!
As described here there are three possible ways to implement modular character
Currently in my project I use FSkeletalMeshMerge to combine some part of character to new mesh using c++ code form that page. But recetly I’ve figured out that meshe merge doen’t support morph vertices. In short , If I try to add morh targets manually, unread does LOD reconfuration and fails because BaseMesh->GetImportedModel()->LODModels is empty. So I have several questions) :
Is it possible somehow create Imported LODModels using generated mesh?
I’ve tryed to do it using incoming meshes array but with no luck, it’s too unstable and seems like I created vertices array in wrong order. Furtermore I had exception “A FRenderResource was deleted without being released first!” and have no clue why.
What are dowsides using Master Pose except rendering time x N (where N parts of character)?
Which way is simpler to hanle collisions with character and cases such as headshots
Physics MasterPose:No Copy Pose: AnimDynamics or RigidBody vs MeshMerge:yes?
What’s actually means Physics is not supported for MasterPose?
Maybe there is something what I’ve missed. Currently I’m about to refactoring my app to MasterPose, because morph vertices is mot critical but nice to have feature. Or will I shoot myself in the foot?
Thanks