Retopology Under High Poly Mesh

Hi Guys,

I am doing some re-topology of a sculpted basemesh.

The issue I am having is that the retopo mesh seems to be underneath the high poly making it “skinnier” the original mesh. Can anyone take a look at the attached image and advise how to fix this? Is it as simple as upping the offset on the shrinkwrap modifier?


use the Offset option in the Shrink wrap modifier to make the mesh go a bit further away from the base mesh.

also, have the subsurface modifier before the shrinkwrap, that way you retain more details in the final mesh.

(and ofc mirror on the top)

Great thanks very much!

Putting subsurf at the bottom is causing me issues as this is a decimated model, it wants to snap to the bad topology underneath. Any way around this?

Is the shrinkwrap wrapping to the decimated model or the original? If it is wrapping to the original, then subsurf first with simple subsurf is what works best for me. The new topology will snap to the decimated mesh, add geometry, then snap to the original high detail mesh. You shouldn’t lose any detail.

Okay so just so I am clear on this. Retopo on the decimated mesh as usual. The import the ultra high poly version and snap to this to get perfect detail?

I usually keep the high poly version on a hidden layer (that way I can see in real time the details on the retopo, but it doesn’t slow down the viewport), keep the decimated model visible, but unselectable. While retopo, I snap to the decimated model, and use a shrinkwrap modifier to the high poly model. I also find that this method of retopo works best with simple subdivision instead of catmull-clark.

The issue with having the the subsurface modifier before the shrinkwrap is that I cannot apply the shrinkwrap without losing the form which I need to do to rig the mesh?

The purpose of the subdivision above the shrink wrap is so that you can retopo at a higher level and the subdivision takes care of adding the geometry needed for the details. When you are ready to rig, you apply the subdivision then the shrinkwrap. You will probably have another subdivision modifier because you will still probably want a subdivision on the final model as well.

Well I don’t like to use shirink wrap, I only use snap vertex feature, the workflow is optimized.
I have done a timelaps modeling and a retopo video showing this.

If you apply the subdivision you end up with a super dense mesh which slows down rigging massively. I have actually found that you dont need to apply the shrinkwrap to rig, Blender is clever enough to let you rig and move the mesh around whilst keeping the shrinkwrap details applied and then set it to subsurf at render time! So actually this solves my issue now :slight_smile: