Thank you Richard,
I will play with both although I actually like Zbrush more from the tutorials etc.
Yes, I’ve heard that painting is way better in Mudbox but so far I don’t know why as I haven’t gotten to that point yet.
Would you know a good painting tutorial for mudbox that would actually demonstrate why is it better over Zbrush, please?
That would be helpful. 
But yeah, Zbrush looks like a lot of fun
I think there are some good tutorials over at Digital Tutors for Mudbox. You can have a look see over there to start.
But I can more or less break it down technically.
First of all Zbrush forces you to paint on a mesh that is high enough poly count to get you the detail you are looking for because it is poly painting. Similar to Ptex but tied into the resolution of the mesh.
Mubox allows you to paint based on the resolution of the map on any kind of poly model. It also has Ptex which allows you to paint on the model without the use of UVs. This can then be baked down to UVs, or even used in a limited way - no bumpmaps for example - in Maya without UVs, In my Opinion Ptex is not there yet as far as pipeline workflow is concerned. They need to do some more work in Maya/Mental Ray I think to make that happen. But it is interesting and can be useful for some things.
Zbrush Polypaint is also not tied into UVs. But you bake it down to UVs on the lowest subdivision of the model to export. Just as you do with PTex in Mubdbox.
Just overall I like the layers interface and workflow in Mubdbox over Zbrush. There are not as many brushes and I really miss a lot of the painting functions in Zbrush when painting in Mudbox. So I am still a bit mixed in my opinion of it.
I think in the end you’ll just have to play with both and decide.
Thank you. That is still a bit confusing but I think I kinda know what you mean but I guess I will understand it even more once I get to that point.
I just have to make sure that I have the model ok for animation and that the sculpting works. (I remember zbrush 7 years ago when you had to export displacement and bump map but apparently these days you can choose the level of detail you want to have in mesh and the rest can be bump or displacement which I think is better - unless i got the wrong information)
Overall, Zbrush looks great and love the potential and I’m saying that after only seeing the “getting started” tutorials. Once I get to more and more I’m sure i will be impressed even more
OK, if you want a definitive answer. It is simple.
Zbrush is the best solution for sculpting currently on the market.
Don’t worry about the animation side of it. These details have been worked out years ago and are still being improved on. You will find tutorials on this kind of thing and you can get the answers you need here to specific workflow points.
It is very simple.
All sculpting detail is baked back to a low poly mesh for animation. You will not be able to make that mesh in Zbrush or Mudbox alone. You will need to retopo the mesh in another app. then bring that mesh into Zbrush and up the resolution of that mesh to match the detail of the sculpt. Then you transfer the detail to this mesh by what is called “projection”. Now your sculpted detail is on the mesh you will use for animation that has levels of detail. You then put the resolution of that retopoed mesh back down to the base level of detail and make normal and displacement maps for it. Zbrush does this by “seeing” the highest level of detail on that mesh and making a map of it to the current level you have selected.
The maps are baked to a UV map you have made for that mesh at its base low poly level.
EDIT: At no time is a high poly mesh ever used in the animation. Only the retopoed mesh. Levels of detail are not exported from Zbrush at the mesh level. Only as image maps.
A similar thing is done in Mudbox.
These maps are then exported to your animation app and applied to the mesh during render time using texture mapping with the same UVs on that mesh in Maya.
You always animate the low poly base mesh. And when you render it subdivides the mesh and applies the details with the displacement maps and finer details with normal maps on a bump channel.
Thank you Richard, that is very helpful and slowly starts to make sense. I’m sure that as soon as I get to that stage I might come here again with some questions but until then I will keep learning and maybe I will even know the answers when I reach that stage.
I guess the tricky part might start when the mess is animated as some of the deformation might need to be redone (my guess) but that is for later.
Right now I will do baby steps. Thank you for your input, it is very much appreciated
Regarding your point about deformation you are on the right track.
Basically when you retopo you want to capture as much mesh detail as you can of the mesh.
Capture the major cavities and so on. Nothing wrong with even going for some of the large wrinkles in skin depending on the character. The more you can get away with at this stage the better in my opinion.
This will give you the most realistic representation.
Now you could actually get away with no displacement at all. In this case Zbrush just becomes a tool to sketch out and detail your mesh in a free form manor. Then you retopo that to a more efficient mesh. For example the level of detail of a mesh to sculpt a clean fold in skin or the details in an ear is quite high. But to go back and retopo that mesh and put control edges where needed is far more efficient.
Then for all of the finer skin pores and all of that you are actually using only a normal map to the bump channel.
In the case of scales on skin where there is a definite profile you’d want to displace. But you’d want to be sure and retopo the larger profiles and details. So it depends on the character.
Now on the subject of displacement on animation you are right on the mark as far as feedback is concerned.
This is actually currently the cutting edge of technology. And in fact you have some of this right now in Maya with the new real time DX11 shaders that will do tessellation (subdividing) of mesh for displacement maps in real time. Useful for feedback while animating.
http://area.autodesk.com/blogs/stevenr/maya-2013-extension-dx11-in-viewport-20
http://area.autodesk.com/blogs/marcel/maya-directx-11-technology-preview
This is currently being further developed with the new Pixar Open Subdiv project. Where the goal is to get much more efficient placement of detail where needed when tessellating the mesh.
You may find this entire video very enlightening.
Thank you.
Yeah, I’ve noticed the new stuff in VP2 so that should make things a bit easier.
One thing I was wondering about as the tutorials didn’t touch it (yet) is that when you are in 2.5d drawing and you draw instances of things is there a way to then be able to sculpt all of it or just the last one? The reason why I am asking is this scenario:
Let’s just say that I want to make a table and 4 chairs and then I will put a bowl on that table etc etc etc. After that I would import it to maya. Is that possible or do I have to do it individually?
The reason I wonder about this is because I’m thinking of creating a whole cottage house in zbrush as with zspheres and sketching it looks like it will be very fast compared to maya.
Btw. so far I love the program. It’s amazing and every tutorial I watch and then test I go like this: “OMG thats amazing! LOVE IT!!”
It makes rainy days very enjoyable 
Ps.: I’ve seen the Pixar stuff. It looks really good and hope it will be in the next maya revision. Shame they didn’t make it for 2014
Yep Zbrush is amazing.
I wish the documentation would treat 2.5 D better.
It did not click with me until I realized that 2.5 D is a powerful tool to create depth drawings that can be converted into alpha maps and used as brushes of different types. I mean that is just one thing. There are other uses.
But for all practical purposes think of 2.5 D as a legacy feature they should kill from the standard interface and just bury as a feature you can pull up is you need it. Granted, maybe that would not be possible. But from an artist perspective it is best to file it as yet one of the many irritating Zbrush interface quirks.
Drag out your mesh, hit the edit button first thing and then forget about it.
Moving on…
The way to set up multiple pieces of a scene if that is what you want to do is to use subtools.
This is a powerful feature that goes beyond just setting up separate objects within one tool container. You use the subtools to also perform many other functions. But never the less you can use it to simply place various parts of a character, the body, eyes, clothes, effects and even props. Same for environments and set pieces.
Zbrush is a powerful modeling tool. But I will stop at saying it is a replacement for certain things I still think are better to set up and do in something like Maya. And for me that would be something like a set with furniture. Not something I’d want to do and set up in Zbrush. That’s me. But Subtools is one way if that is what you want to do. You’ll get there soon enough.
Thank you Richard,
that makes sense now. I thought it could be the case but the idea of creating alpha for brushes is an interesting one. That can actually be something really great to use this for.
Yeah, I’ll look into the subtools and yes I wondered how to separate eyes for example from the character so you have cleared that out.
I think soon I’ll post my first test so lets hope its not a total garbage
Thank you for help and input, very helpful and nice of you.
Thank you so much
ok, i’ve finished the getting started videos and did some of the sculpting one.
I’m a little bit curious about what others think its best for character creating.
I love the zspheres approach to create the basic “skeleton” of the character, then use sketch brush to sketch the mass to a more detailed model and then convert it to polymesh and continue normal scultpting process. That seems to make sense and looks fairly simple.
Now, what is the dynamesh or is it any use these days?
What is the most uptodate workflow (I know everyone is different but I want to avoid using “old” approaches) that most people use, please?
Thanks
Yeah you can use zspheres like that. The cadge you get from zspheres is rarely symmetrical so the center line is often off and you can fix those and center diamond polys in a standard 3d app. If you use goz it is even quicker and you end up with a cadge that you can resym at will.
Dynamesh is another really useful approach as it lets you grow meshes out of your start shape and by using a remesh stroke the resolution of your sculpt gets distributed over the whole piece when its needed. You can also use the project feature to stop the form from going mushy when you do that. You can insert meshes and dynamesh has a super boolean function as well. You cant go up and down in resolution like you can with a standard zbrush mesh but the possibilities for experimentation seem endless. At the end when you have a more worked out form you can always retopo or qremesh your sculpt.
Ztools are separate models and you can see them in the tool dialog where they appear in a list like ps layers. The body of a character could be one tool and clothing pieces other tools or ‘subtools’ for example. In turn subtools can be made up of many elements or just one. You can hide subtools and turn their textures on and off and you can also use them as booleans. Ztools are an incredibly convenient way to separate parts so you can handle them individually. Those ‘tools’ or parts can be combined and split whenever you please depending on their resolution the tools will also maintain their sub level history as well.
Hope that helps.
Opinion forthcoming… take it with that in mind.
I don’t like Zspheres at all and I find them incredibly odd to work with. I agree that they look simple and logical.
But…
They are actually an older method. A newer method is Zsketching and the most dynamic and intuitive to me is Dynamesh. It is also the more “current” technology.
Of course one can argue - especially if one is a long time Zbrush user - that all of these tools have their place. But I would argue that the only reason some of these tools have their place is because they became comfortable from use mainly because there was no other way. And but now there is a better way - in my opinion - so why mess about learning an older way?
That is just me and the way I decided to approach it as I worked through the vast documentation. I very early on decided that I hated dealing with Zsperes and was just as happy to check that off the list an eliminate hours or days of training time.
However it is a good idea to be familiar enough with them as they can be used as a base for making topology with the topology tool.
That said, beyond just a basic character Zpheres/Zsketch is a good way to sketch out more odd shapes. Not for me, but for others… up to them.
But regarding the problems with symmetry you can always convert to a quad mesh and also use the mirror function. There is also the slice brush and other tools to use to correct the problem within Zbrush. The more familiar you are with the tools the more solutions you’ll find.
With Dinamesh I simply drag out a sphere and start sculpting. Insert brushes come in handy. But I like to just push and pull and shape something into what I want.
I am also not ashamed to use one of the included human meshes as a starting point to sculpt my own character in Dynamesh as a starting point. Or you can import premade meshes from other sources and start that way as well.
In general there is this workflow:
You sketch out your basic shape - by any method.
Once you get it close as a base shape then you can then convert to quads and start using levels. There are a few ways to go about this but projection is a key component. You could retopo first and then come back to add details.
But I have been settling on the workflow of using dynamesh and then convert to quads, reduce the resolution as far as I can using Qremesher and then add levels and then project back to the highest level to get the details from the Dyamesh back. Then from there add all of the finer details on the sculpt and then finally retopo in an external app and bring it back for projecting details needed for displacement and/or normals occlusion UVmapping and so on.
Just my opinions based on how I have found I like to work. FWIW.
thank you guys, I will reply properly a bit later when i test certain things I wanna try first before I ask but here is something i wonder.
You mentioned retopology in early stage on the model but I thought that I can just use zspehere, then zsketch and then convert it and start sculpting and only after that do the retopology, was that incorrect?
Because my impression is that it doesn’t matter what method you use as in the end you will have the same mesh as a result, correct? (ie. zspheres will not be inside the mesh anymore same as zsketch etc. the mesh will be just surface like you would do in maya and it will be the same regardless of the procedure, correct?)
I hope you don’t mind me asking but I though I would get this right the first time so i’m on the right track.
Yes. In the case of Zphereres you would basically convert it to a mesh and then start sculpting.
And yes, just in general the workflow is to first sketch out a basic shape before adding details.
In Dynamesh you can take this quite a ways. But it is just generally an intuitive way to sculpt and keep mesh integrity. For example you stretch and pull then you can remesh it on the fly and it will fill in with new geometry so that the mesh is not stretched out.
Any kind of topology can be done at any stage of the process, including sending objects back and forth to a 3D app or retopoing parts as you go to give you a better basic shape.
But me personally I don’t like to get too tied down to topology until I am prepping for animation and maps.
There are no rules. As you go through you will find things that work best for you.
So actually until I’m ready with the model for animation etc. then there is no “wrong” way to do things as I can always fix that even when the model is done, right?
Well in that case it makes it easier I guess to start and headache will come later when animation will start 
Thank you Richard
so now that 4R6 is out would you recommend using that for retopology or are we still looking for topogun as a better retopology solution?
You experts will probably know straight away if the new update replaces the need for me to learn topogun :)))
Thank you
p.s.: Mari 2.0 still best for texturing?
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.