Many models and meshes for animations / games have hundreds of thousands of polygons contained within themselves. Are these large-polygon-count meshes generated through something like subdivision modeling, or does subdivision modeling simply refer to how the model is eventually rendered?
Can we do this the other way to generate low poly variants of complex models?
Is the leftmost picture representative of anything that ever gets actually showed when creating a model? Do modelers start by making something that looks like that? It seems difficult to model a character like that because it doesn't look much like what you want it to look like.
Does subdivision here provide some functionalities like interpolation or smooth?