Hi,
I would like to reduce polygon count only for areas that do not need detailed sculpting. I have e.g. a head and I do not need the back side of the scull to have high resolution since It is a part without any details. Is it possible to reduce the polygons only in this regions?
Hello @Mirko
The answer is “Yes, but…”
There are a number of ways to quickly do this. Currently, the Tessemate feature is very useful here. Check out that video. You can also use ZRemesher locally by hiding all but the mesh portion you want to remesh, and enabling “Freeze Border”.
If all you need to do is reduce polycount while keeping detail intact, say for a 3d print, then Decimation Master is the way to go. DM will remove points where they are not contributing to detail, while keeping the detail intact.
However, any method that only re-wires a portion of a mesh independently of the rest is going to produce some pretty ugly topology that may not work well for sculpting, or many other processes. They should be viewed as quick and dirty, or temporary procedures. This may be suitable in the short term depending on your need, or for 3d print projects that don’t have many topological concerns, but for a quality mesh for more advanced uses, you will need to eventually re-mesh the entire thing with more deliberate topology.
If you’re still going to be remeshing frequently, for instance, then the entire mesh will just be done over anyway, and you will lose that local reduction.
The best way to do this, when you’re at a stage where you will no longer be frequently changing the mesh’s topology, is to ZRemesh, or manually retopologize a new low poly mesh with clean topological flow and evenly distributed polygons, mostly quads, as close to square shaped as possible. This is the best possible form for getting the best results with ultra high detail sculpting, painting, and texture and displacement map creation in Zbrush.
When creating this topology, create it in such a way that more polygons are delivered to high detail areas, and away from areas that dont need it. When subdivided, the areas with more polygons will be able to hold finer detail.
Well I have to confess that my major concerns where about the memory that is used for a high poly count tool. I am working on a very old MacBook where it is nearly impossible to sculp on a tool with say 1 million polygons. But of course the lower the resolution the less fun it is to sculpt. So I need some kind of mean between working without delays and having a sculptable tool. Could you recommend a poly count to be able to sculpt with fun ?
Maybe a solution could also be to devide Front and back side of the skull into two subtools?
Hmm.
One of the strengths of ZBrush’s unique architecture is that it performs remarkably well on fairly modest hardware–hardware that wouldn’t be able to crunch that many polygons in programs that are more hardware dependent. 1 Million polygons is not significant in Zbrush terms. Unless we’re talking about a laptop more than 10 years old, I suspect that technique is more of a factor than hardware here.
Click to read further--collapsed for readability.
A common pitfall is to work too long at a single level of subdivision–the result of using tools like Dynamesh beyond their intended range of usefulness.
The “classic” Zbrush sculpting workflow is to work on a tool with multiple levels of subdivision and a low poly base. A mesh in this form is still required to get the best results during high detail sculpting, posing, and texture creation/export. Zbrush has a lot of features to improve performance while working like this that you lose with only a single level of subdivision. For instance, if a lower level of subdivision is available, Zbrush will display this while navigating or rotating the model, keeping performance high by reducing what your hardware has to display.
In recent years, many newer tools, like Dynamesh, have been introduced in Zbrush to help users create meshes. Many of these tools are built around the ability to rapidly change the mesh’s topology, which limits their use to a mesh with no subdivision levels. These tools are designed to be used at low to medium levels of detail to help establish the basic form. Once this form is established, depending on your output needs (working for print has different needs than working for digital render in an external program), the idea is that one would still generally transition to a tool with a low poly base and multiple levels of subdivision for sculpting fine detail, and preparation for export.
A common (but not the only) way to make this transition is to ZRemesh the model to create a new , low poly version of the mesh you just made with Dynamesh or one of the other single subD tools, with clean, well distributed topology. This mesh can then be subdivided into multiple subdivision levels which will allow you to achieve high rez detail sculpting, and the fine control of being able to work on the mesh at a variety of different resolutions.
Generally when you reduce polygon count, you will also lose detail. If the mesh you are ZRemeshing contains finer detail than can be captured with the resulting low poly mesh, you can Project the Detail from the original onto it, once it has been sufficiently subdivided.
In addition, make sure to split any portions of mesh that don’t need to be able to be sculpted on at once into separate subtools. So for instance, on a character mesh, different parts of the main body might be separated into Polygroups to make them easier to work with, but kept as part of the same subtool, because you would want to be able to sculpt on the body as one piece. A backpack, or a hat, or a gun, or a jacket, should all be a separate subtool.
Finally, use of features like Sculptris Pro can allow you to sculpt high detail areas into a low detail mesh, but for many outputs, you are still going to want to clean up that topology eventually. Also keep in mind what I said above about trying to sculpt too much detail at a single level of subdivision.
I should not tell you that my MacBook is from 2010 
So let me recapitulate if I got you right because I have some difficulties in understanding the technical terms in english.
I usually start with a low subD level mesh and then when it comes to detailed sculpting I increase the level. When I think I am finish then I reduce the level (e.g. via ZRemesh) and project the higher mesh on the lower mesh (in short). But what if then I forgot to sculpt a detail and need to sculpt again on that mesh?
Or if I created a human body, converted it to low polys and then want to use that body as a template for another body. What would be the approach to change e.g. the face of that old body for my new body?
And is it possible to switch between subdivision levels. Lets say for the face I switch to a high level and for the back side of the skull I switch to a low level?
I am not talking about projecting the meshes but only switching between the levels.
I also don’t understand why Polygroups helps. Because if e.g. I subdivide one polygroup and the other one not (assuming that it is possible) then the total mesh kind of gets raggedly isn’t it?
Sorry to heat you with that much questions. But since you guys are so experienced It is a good chance to get some insight for me.
Thanks to your posted Tessemate video @Spyndel. I think I got the answer now that I watched the video. So you can add tessellation where it is needed. Like for the face I can add more tessellation then for the back of the scull.
But let me ask you one thing. This tessellation seems to add triangles always or at least when you need more details. Isn’t it bad to have a mesh with triangles instead of quads? Or is that not an issue because you can always remesh to quads at anytime?
Triangles aren’t ideal for the sculpting process–quads produce smoother results–however the geometry produced by Sculptris mode is generally adequate for sculpting low to medium detail subjects, which is what it’s intended for.
Like I said above though–these tools are best used at low to medium levels of detail to create a mesh. While you’re creating that mesh you have a lot of freedom with the topology. But for many purposes, you want to be eventually be working towards a goal of having clean, quadratic topology. This will produce the best results when you’re ready to start sculpting very fine detail, or UV map your mesh and create textures for it.
Otherwise , you have a whole lot of questions about some fundamental 3d issues. The answers to these questions vary with what you want to actually do with your mesh. If you’re working for print the answers will be different than if you’re working for a static digital render, which will be different again from working for animation, or a real time game.
Be sure to start in the Pixologic Classroom and learn the fundamentals of the entire program, rather than fixating on a single feature. Without this knowledge everything you try to do in the program will be more difficult. Also figure out what you want to do with your work in Zbrush, as this will begin to answer some of your questions.
Thanks a lot @Spyndel. I think it is just that there are sometimes too many ways to do it and you always need to decide what best fits. As an example while I was sculpting a mouth I changed from clay to inflating to move or move elastic branch. But then i realized that the dam standard or pinch brush could fit more. But now I think that crease brush might also be better to solve some issues. Also in case of remeshing there is that zremesher, the decimation master and the tesselation thing where you have to choose between what best fits my needs. Anyway… of course I am watching those classroom videos as well as some YouTube videos etc. But here i can ask much more precise what is my issue and that is why I love this opportunity.