Reply To: Hochelf’s Blender Progress

[Shawn Blanch]

This is how the “World of SubD” works according to my understanding (it may or may not be correct so take it with a grain of salt). It’s really magical when you think more deeply about it.

I like to think of “Subdivide Collapse” as a way to keep generating new faces as you sculpt. It subdivides the long edges on a face and collapses the short edges (creating and destroying geometry as you sculpt, though it feels more like you are just creating more so than destroying). This works well when adding hills and valleys to the sculpt or using the grab or snake hook brushes.
But it works poorly when trying to get good pinching and creasing on something because the creases end up joining back into one piece when they get close enough (no longer creasing).

“Collapse Edges” removes the short edges and leaves the long edges intact. Collapsing is probably best to show rather than tell.
This is a short video of edges collapsing manually but the process is happening automagically during sculpt mode when you set Dynotopo to Collapse Edges.
Collapse Edges works well when you have an area of high poly (edge/face) density that you want to bring back to match the rest of the model. For example: if you were sculpting a High Quality Human: the face would probably have a very high poly density at some point and the rest of a body would have less poly density. Faces need to be detailed to be believable. Once you have finished getting the fine details refined and believable you could use a brush with low strength and turn on Collapse Edges to lower the poly density.

“Subdivide Edges” does what it says. It splits the long edges of a face up more and more where needed to prevent stretching (effectively creating more faces and geometry but only where needed). I like using this when using the crease and pinch brushes where I want lips and other naturally creasing areas to look more realistic. The other methods don’t respond well to heavy pinching.

“Relative detail” changes the size of the dynamic topology (creation and deletion of edges/verts)based on how close to the object the viewport is (the closer you are the more detail you will see). This can be good and bad. If you zoom out and sculpt over an area that has a ton of detail using Subdivide Collapse w/ Relative Detail, the high quality details and small faces will get replaced with less detail and larger faces. The opposite is also true when zooming in.

“Constant detail” uses the same detail no matter how far or close you are to the object being sculpted.

Keep in mind that some brushes do not execute the Dynotopo. The smooth brush is one of them (I think that was a wise design choice).

Hopefully this explanation helps a little.