ZBrushCentral

When is the best time to retopologize

I love Zbrush but I have a question for the community. I want to make all my characters game ready, however that arises a problem. When should I topologiZE


  1. before I put in the fine details, this way they look better because they move with the flow of the topology. Now this works but it quickly becomes a huge pain in the A$# for me, because as I work on it I come up with a bunch of new ideas that can make the design much more awesome. But then I have to jump back and forth fixing the topology to match the details. this gives me great quality but it takes way too much time.
  2. Topologize after you do the fine details, Now this becomes problematic because the details are just not moving with the topology, and look weird at close up. Now I can fix up the details on the new topology model, but that is time consuming as well.

Also if you guys have free time check out my website www.scorchingstudios.com tell me if you think I’m any good considering I picked up this program a year ago. I would love any feedback:).

It can depend on how you’re retopologizing and what you’re retopologizing for. If you start with a sphere or basic mesh and start sculpting from there, then at some point you may have to retopologize in order to create a better polygon density for sculpting in the finer details. Dynamesh and ZRemesh are methods of retopology that make quick work of this.

If you’re creating something that’s meant to be animated or used in another program, then chances are you’ll need to retopologize the mesh at the very end*. This could be something as simple as decimation master, or a custom edgeflow designed to support good deformation during animation. For the latter you generally only have to worry about the silhouette/form when creating your new mesh, as the fine surface detail will be recreated through various texture maps.

* This may be different if you’re working with a team, and depending on how they work. It is possible that you might have a premade basemesh for every character that you wouldn’t want to retopologize in order to speed up the process. Or you may need to lock in the main forms and proportions and retopologize then so that you can send a mesh off to the rigger, giving the animators something to work with while you spend time sculpting the fine details and texturing.

I think that Zremesher is really not that good, when it comes to hard surface models. So with hard surface its hard to get the proper edge flow, to get good textures when exporting. Also the perfect edge flow for a video game will not have the proper poly density to be able to evenly paint it in zbrush. So sometimes you have to add edges that you might remove later. Am i being too much of a perfectionist in this process, what do you guys think.

ZRemesher probably isn’t intended for every model, such as hard surface shapes. You either need to


  • have good edgeloops imported from an obj file modeled in another program
  • take advantage of the various creasing an edgelooping options that zbrush is continuing to improve in order to control harder edges
  • throw any notion of proper subdivision topology out the window and brute-force the shape using dynamesh and higher resolutions

Polypainting on the game-mesh would depend entire on what the mesh looks like. I’ve seen plenty of base meshes and final game meshes have more than adequate uniformity to allow for easy polypainting and sculpting, especially in the more cutting edge games that can have tens of thousands of polygons per character.

If for some reason you’re painting and sculpting on a mesh that isn’t designed to handle this, might as well retopologize form sculpting/polypainting and make your life easier. A quick dynamesh would be done in seconds and would give you all the density you could need. The basic answer to your original question is simply retopologize whenever you need to for sculpting, and retopologize whenever you need to for animation. Some models and projects are going to work great without ever needing to be retopologized, while a mesh that might have started its life as a sphere could be retopologized hundreds of times a day until the end result is in a game. Retopology is no longer the intimidating, time-consuming task that it was several years ago.

So I think the solution would be is to retopologize. To get sharp angles on my hard surface. Then dynamesh, paint and project the details from the dynamesh, to the good topo version in another program. Thanks for the great feedback