Well like most when modeling, i use a subdivions modifier, to smooth things out, often i subdivicde twice.
Now i am wondering, since my model could use a bit more detail at some areas; (and its allready textured etc) i might apply one sudivision, edit model and only do a 1 level remaining subivsion, or apply 2, when only using proportional mesh adjust editing…
My concerns are, that this might become problematic when applying an armature later.
For example, correcting automatic weightpaints, might be complexer ? … or not.
Adding extra bones later might become complexer ?.. or not ?
Maybe there are speed benefits of lower poly mesh since subdiv is applied after armature deform ?
Maybe subdiv after a armature, is the best option to go for to keep things smooth looking ?
Maybe some of you have experience with this, and like to share some advice on this topic.
You should look into the multiresolution modifier. It lets you subdivide and sculpt on the higher resolution, while still retaining your low poly mesh. It can be used almost identically to subsurf. Most likely you would have the armature modifer at the top of the stack, then the multires modifier below it.
Tp use the multires, add a multires modifer, click the subdivide button. Make all the levels (preview, sculpt and render) 1, go into sculpt mode, start with a clay brush, in symetry you probably want x on. sculpt in some of the bigger details, when you want more resolution, click subdivide again, change all levels to 2, sculpt some finer detail…
It would really help if you could explain what you are really doing here…
I understand you need a high poly mesh as a final result, but what are you doing with the mesh? If you are just doing static poses for final renders, then a high poly mesh might be ok. It would be sluggish to pose, but once it’s posed, you render the result. On the other hand, if this is an animation project, where you’re scrubbing the timeline and looking for real time playback in the 3d view, you’d really want a low poly copy to work with so it can be posed quickly. Then when it’s render time, use a high poly version for the final renders.
Could you model in the areas that need fine detail? Instead of using a modifier with a high level of subdivision, (making the whole model vertex heavy), couldn’t you just model in more detail in the areas where it’s needed?
Additionally, you can create a high poly mesh, then a low poly mesh and bake normal/bump maps to the low poly mesh. This allows you to see detail of the high poly mesh on the low poly mesh.
So on to weight painting, higher density model might not auto weight as well as a low poly. Either way, auto weighting only gets you so close to what is needed. Auto weighting does a fair job of weighting a mesh, but it always needs further weight painting touch ups.
