zbrush workflow
I find myself in a vexing, familiar situation: Uncertainty.
I bought Zbrush a few weeks ago, and have been messing around with the software, eager to do some sculpting, and I realize now, I have no idea what I'm doing. I'm missing a critical step between understanding the software and executing the things I want to do.
I try to model something, and then I ask myself "is this the right way? I'm sure there's other ways, but is this one even working? How can I tell? How can I tell I'm doing something that'll hurt me later?"
Those are my questions, that's my situation. It's not a matter of putting the hours in, that's a given; it will have to be done no matter what anyone tells me. It's a matter of direction, of me having none.
to not waste your time I do have things I can show, if poor and feeble. If this post is in the wrong forum, by all means move it somewhere more appropriate.
I bought Zbrush a few weeks ago, and have been messing around with the software, eager to do some sculpting, and I realize now, I have no idea what I'm doing. I'm missing a critical step between understanding the software and executing the things I want to do.
I try to model something, and then I ask myself "is this the right way? I'm sure there's other ways, but is this one even working? How can I tell? How can I tell I'm doing something that'll hurt me later?"
Those are my questions, that's my situation. It's not a matter of putting the hours in, that's a given; it will have to be done no matter what anyone tells me. It's a matter of direction, of me having none.
to not waste your time I do have things I can show, if poor and feeble. If this post is in the wrong forum, by all means move it somewhere more appropriate.
Replies
Looks like you're doing a mix of stuff, which is what zbrush is great at, so it might be a good idea to head over to zbrushcental and watch some of their process videos.
From what I'VE seen, everyone develops their own workflow in ZBrush. Whenever I get to watch high end guys work they always seem to do things their own way. ZBrush has SO many tools and SO many options in it, it seems like it can handle any task in 2 or 3 different ways.
For example, you can watch people work while you work and get some tips ... like Hazardous: http://www.twitch.tv/hazardousarts/profile
I know there are 10 different ways to do something in zBrush. I want to know what the 10 ways are, and try each of them?
the vexing/uncertainty part is that I don't get the right kind of feedback from the software yet; I can only tell something is wrong because it's not what I want yet, and I get nothing on"How to solve the problem."
As for what I want to do, I want to do hard surface models/machines, as well as organic characters/sculptures. Again a lot like HazardousArts or Mike Jensen.
Here's a specific question: I've been modeling a face, and it looks lumpy at high subdivision, and no matter what I do I can't seem to make it smooth. One path that gave me reasonable success was to go down a subdivision level and smooth those polygons.
For my experience, i spend a lot of time just doing random stuff. Experimenting with the various features and see what i get. Most of the times it's just random shit, but 1 out of tens or hundreds might be something useful and reproductible.
When experimenting, look for reference and try to mimic them as an exercise. The first step is always to REALLY understand how things are done in real life, and only after figuring out how to replicate them.
Often, you'll be stuck and wonder how to do certain things, just google them and see how others have done it. A lot, in game art (and not only) is just simple problem solving. You just get used to it. Still, seeing how others have solved the same or similar problem before us, helps so much.
For the "when should I when i should work low poly, or high poly, or when to subdivide, when to remesh, or how" part, it'll eventually come to you with practice. For example you'll know at one point when you actually need to subdvide a mesh, just because you can't push any more details in it even if you begged for it.
You'll know when to retopo/remesh, just because you've been pushing and pulling geo for a whole day on your mesh and it's starting to get uneven and you've got problems getting the result you want.
It's only just a matter of experimenting all you can, then push more and see when and where problems arise and solve them.
I've been where you are, and I'm certain everyone does, so I hope this helps
http://eat3d.com/zbrush_hardsurface2
Heh, I actually bought Hard Surface #1 already, the skullface robot guy is the direct result of me trying to follow along.
And Fnitrox that was very helpful, yes! I think it's a sort of quandary of asking for art-help; I want help with the media, a direction so I can practice. Your comment re: topology is the stuff I wanted to see, because I don't recognize yet what those points are, but now I can look for them!
Anyway, points of order:
I've tried to use zremesher with curve guides, to little helpful effect, and I have no idea if remeshing parts/polygroups (as seen here) is a good idea or a bad idea. Also when zremeshing i lose my multiple subdivision levels, so this here has been a one-way experiment into 4.4 million polys (the whole model, not her face).
And do you want this to be a useable game asset?
And No, I am not aiming for game asset, I'm aiming for 'Good Zbrush Sculpture'.
I just want to better learn the tools to get from 'blank canvas' to 'finished model'.
SubD- Standard geometry to work with for most people. You have the upside of subdivision levels which allow you to modify big forms in a forgiving way, but the downside of being restricted by your original basemesh.
Dynamesh- Dynamesh isn't overly smart, all it wants to do is make nice uniform square at x resolution for you. This makes it very powerful for rapid freeform creation, but lacks subdivision levels. I mostly have sculpted things like heads in dynamesh, upping res as needed. Its predictable and simple, which I like.
zRemesher- I haven't really integrated zremesh into my sculpting too much yet (mostly used it for generating starting points for gameres), but its kind of like a smarter dynamesh with a more complex ruleset. You can control global resolution, specific area density and edgeflow. I guess it could help bridge the gap between subd and dynamesh. For instance, you could sculpt a dynamesh head and use zremesher to generate a nice basemesh for you to subd sculpt on in the future. zRemesher could also be used to lower/raise poly density on certain areas of a dynamesh, to squeeze out a sudivision or two without going overboard on polys.
In your images above, your density is way too high to make working with it easy. I'd recommend zremeshing a head base to something not so dense, or using a low-res dynamesh base if you think you'd prefer that route.
Alright, so Zremesher is arguably not good at re-topologizing? Then what is a method of redoing topology? I haven't been able to find another one in the software yet.
And by re-topology i mean like creating edge loops and arcs for eye sockets and mouths and muscles. I'm at the point of understanding where I don't even know if i can do that in zbrush, or if I should. And If I cannot/should not, what do I do instead?
See video and it will explain wth to do with ZRemesher and what it will not do:
https://www.youtube.com/watch?v=CVoFi2mnyhA
If you're confused about what to do, take a piece of concept art and tackle recreating it in a sculpt. Right now it sounds like you're not necessarily thriving on these "experiments" that you're aiming to look pretty to begin with. Get something you like from CGHUB or something and try to recreate it.
And what are you worrying about final topology for if all you want is a pretty final sculpture?
Zremesher can help with the direction of edges, but on the eyes for instance, its more about density than edgeflow for sculpting. So you could use the polypaint density controller if you'd like to try that.
If you want to make a specific low-poly mesh, look up zshpere topology. It works in the same vein as 3dcoat, topogun etc.
Thanks for the video: watching as I type.
RE: Final sculpture: Well, so far all my experiments have looked terrible due to geometry doing things i don't want, like not having the right amount of density around eye sockets and the like; it creates ugly jagged sculpt lines.
My experience with non-zbrush models usually tells me that's a sign of bad polyflow, hence me poking at it.
and.. Okay so, Take concept art, try to recreate it... and I have almost no idea on how to do anything. Okay...
Edit: I can't even get polypaint to work the way I expect it too. Every time I try to apply a material, it sticks to every subtool.
Cuz there's Google Hangout, and I'm available to do some quick 1 on 1's
I'm floored by the offer. I might want to take you up on that, but probably later. Even so, thank you kindly for offering.