Zbrush to xNormal workflow

Alex1

Hey guys,

I'm currently making my first low poly asset and I've run into a bit of confusion regarding the workflow. Can someone please help me out?

I've modelled a high poly object, sculpted detail onto it (ZBrush) and produced a low poly version of it. I'm now ready to bake in xNormal, but I'm confused as to how to export the high resolution object - do I merge all of the subtools into one mesh, or export subtools individually?

The problem is, if I merge all my subtools, my polycount reaches 22 million and ZBrush crashes. So the obvious answer would be to either lower resolution or export individually. However, if exporting piece by piece, then the low poly has to be broken into those exact pieces too, right? But then, how about the baked out texture maps; it would mean having to combine each texture map multiple times piece by piece, correct?

Lastly, if baking subtools individually, don't you lose ambient occlusion information that you would otherwise gain from the intersection of two or more subtools (e.g. where the handle of a hammer makes contact with the hammer head)?

How do you guys do this? It would really help if someone could explain or share their workflow.

Thank you.

Bartalon

You can bake as many HP models as you want down to a single low poly mesh, as long as everything matches up in 3D space. It's not uncommon to have multiple HP mesh objects loaded into xNormal (my last project had 13).

Personally I tend to keep my LP mesh airtight (for current-gen stuff) when I can because it produces better bakes and wastes less UV space compared to intersecting geometry, and I usually don't need to explode my meshes.

Alex1

Ahh I didn't realise you can load multiple HR meshes per one low poly mesh. Thanks!

Sorry, but what do you mean by 'airtight' in that context?

Bartalon

Airtight means to just have one contiguous mesh where all the polygons are welded together. The alternative of intersecting polygons sometimes produces baking artifacts around where the polygons clip into each other, and any segments of polygons that clip inside the mesh becomes a waste of UV space.

It's not always better to have airtight meshes and it's not always bad to have intersecting geo in some places. It just depends on the model and any project constraints.

Alex1

Understood. Thanks for the help and explanation!