ZBrushCentral

Problem with diff mesh - work around

Diff Mesh Tweak

I’ve never had a problem with Diff Mesh before. I haven’t used it a whole lot but I’m trying to more fully integrate Zbrush into my sculpting/modeling pipeline rather than waste time exporting assets to tweak in other software. That said I was surprised to find this little bug nugget lurking in the feature.

I have Zb 4r4 and the Diff Mesh feature has always worked for me, like I said I use it sparingly but I’m trying to rely on Zb a little bit more so I’m using features of the program more. Yesterday I threw a Diff Mesh extrusion onto a trench coat I’ve started working on and it completed as I expected it to. I ran out of time to work on it before work and walked out the door without saving my work (oops on my part) so I had to redo the 3-4 steps I managed to throw at the project after I got back 8 hours later.

Problem was that the output mesh wasn’t behaving like it should. The generated mesh looked like I expected it to but when I tried to subdivide it things started getting weird, that�s when I started trouble shooting. At first I noticed a few problems that seemed to contribute to the issue.

First the mesh that outputed seemed to have a �sticky mask� on it I couldn’t clear. I thought that was the problem and started writing up a trouble shooting ticket for it until I dug a little deeper. When ever I tried to subdivide the mesh it acted like a local subdivision. It would subdivide the mesh but it wouldn’t actually smooth it (operation produced the low frequency faceting artifacts) and ate my polygroups. The mesh density was where it needed to be but I couldn’t step threw subdivision levels because they weren’t in the geometry options.

I started digging a little deeper by doing mesh integrity tests. These all passed to my surprise. I understand though that most of the integrity tests look for ordering issues which understandably all passed as the mesh in question was generated in Zbrush (from Zspheres to the Diff Mesh operation) but in the context of this bug it was a misnomer.

Knowing something was wrong with the topology I started poking with the Modify Topology features and made sure to use the WeldPoitns and MergeTris buttons. I’m not exactly sure which button fixed the problem but the generated mesh has some issues with arbitrary holes or zero surface area triangles that needed to be merged.

Once I preformed these steps I could subdivide the Diff Mesh as expected and have access to my subdivision levels though there is still a problem with the mesh that I can’t reconstruct subdivision levels (attempted this fix at various subdivision levels wondering if the thin wall nature of the output mesh was to blame for the ordering errors) as suggested by my earlier comment about triangles the reconstruct operation fails and mentions triangles on the mesh in the log output.

While I do have a functional mesh there shouldn’t be any triangle on it as its A- generated from Zspheres and B- should solve into a quadrangular mesh. Other than the existence of holes in the mesh that generate triangles I’m not sure how it would produce triangles and seems like an issue that needs to be addressed at the coding end of the equation.

I’d be happy to throw you guys my base mesh from the Zsphere so you can see if you can recreate the problem on your end, since this issue has produced a work around I don’t see much of a reason to post screenshots unless I want to make a youtube video demonstrating it.

Just wanted to post this as a recommended work around should anyone else run into this problem in the future and need a little help in fixing it.

Good luck to all!

The difference mesh feature is more or less a holdover from the days before we had the Extract feature. Extract also has a number of settings that allow you to get higher quality, and requires fewer steps (and less guess work) compared to going the difference mesh route.

Is there a particular reason you’re not using Extract instead?