I want to use Zbrush for Hard surface (For Production)
Is this thing accepted from studios? I want to model/sculpt a hard surface in Zbrush and later use that for retoplogy and to be used in animation.
I really like how Zbrush handle hard surface, it is 100x better than Maya. So i was thinking, can i use Zbrush apart from character to sculpt also hard surface? Like mechanical stuff, vehicles etc.
I really like how Zbrush handle hard surface, it is 100x better than Maya. So i was thinking, can i use Zbrush apart from character to sculpt also hard surface? Like mechanical stuff, vehicles etc.
Replies
its probably not wrong to learn it, because there are certain workflows where zbrush hardsurface might be realy usefull.
but there will also be a lot of cases where its utterly useless and trying get the result you have to deliver will be certainly possible, but a lot slower then the alternative( wich is traditional subdivision modeling).
so i wouln'nt rely on it, and you should have an alternative for when it does;nt work.
The issue with using Zbrush as it is now for hardsurface work is that not only can you not go back and precisely edit edges, but it will sometimes produce a blobby result that cannot be easily fixed without re-doing the entire piece.
In an environment where changes are inevitable, using Zbrush probably isn't a good idea, as it would cost you more time at the end of the day if you wanted to change something.
A clean HP mesh will always be easier to work with.
A lot of people argue that doing it in ZBrush will give you a speed boost since you can just dynamesh awkward shapes, e.g. a piece of armor shaped like an eagle's head might be a pain to model, but could be blocked out fairly quickly in a sculpt (and then retop'd for better surface quality if required....) although, you then have to think about whether or not it would have been faster just modeling it in the first place.
Mike P also does some pretty cool character based stuff with it, some of his vids: https://www.youtube.com/channel/UCWiZI2dglzpaCYNnjcejS-Q
RexM - I kinda thought the same, but a lot of the Zmodeler tools were only demo'd on a small cube, I would be interested to see how they work on a more complex shape like a curved surface or more awkward slopes.
I think the general consensus is that hard surface in ZB sacrifices precision for speed. Again its debated a lot in the above thread, worth checking out!
Maybe because of that i asked this question
Zbrush4R7 should bring in some nice solutions for hardsurface modeling using Zbrush. But you still should know how to achieve what you want via SubD modeling.
Totally a viable route.
These are the dudes:
http://www.marcoplouffe.com
http://drawcrowd.com/fred2303/projects
http://cedricseaut.prosite.com
All done in zbrush.
There's your problem. You're trying to model in Maya, so anything would seem better by comparison.
There are better options than either ZBrush or Maya for hard surface work.
qft
i hear character artists wanting to shoehorn their character art workflow into making hard surface a lot, and it rarely ever works. traditional subd modeling isn't going away for a long while. It's fast, clean, and above all--predictable. it really is the best option, imo.
But Prav is being very clear here :
And that makes a whole lot of sense !
- Of course you most likely want your final deliverable meshes to be built with clean topology ready for subdivision ; and such models can also be complemented with straight-on polygon work, as demonstrated by Snefer in his recent videos.
- Of course sculpting can help in the process. Now I'll admit that I personally do have a bias against meshes delivered as a polygon soup by the Zbrush wizards mentioned above, simply because despite looking great they are very heavy to manipulate (I personally cannot stand to work in 3D under 60 or 30 fps, and Zbrush dips way below that when working with such models. It seems that some artists don't mind that, but I am just not one of them, and I wouldn't recommend anyone to work that way because of the inertia it causes). But when it comes to finding/solving shapes to make them work in space, sculpting programs with dynamesh-like capabilities are a huge timesaver. Now of course for blueprint-based work or very angular mechanical work it's all kind of pointless ; but as soon as surfaces bend in space in semi-organic ways, it would be a bit silly to not leverage the power of sculpting to get things resolved.
It's all about separating the artistic tasks (resolving the shape to be modeled) from execution tasks (producing the technical mesh). I feel like the only good argument against leveraging Zbrush for that is the fact that it requires OBJ exports/imports or Zapplink, which is always a drag and can get one out of the zone. But this whole debate would suddenly become irrelevant if programs like Max and Maya had solid sculpting tools integrated in their toolset ... like Blender does.
[ame]
Hybrid integrated workflows are the way to go !
i can see sculpting as a way to create the mayor shapes, totally but such fine tuning? how does one handle it? Or simple form changes requested by clients/leads/artdirectors? a simple basemesh is just so much easier to handle in a clean manner.
Pravely : no worries about the title man, you made your point very clear in your OP.
I guess for personal work it is great, you know what you want, you don't have to follow (or interpret) a concept by someone else, the way you do it is right and nobody will do the artdirection but the person in charge, you.
When it comes to making adjustments like edge bevels (or removing one), you really have to break out of the mindset of the traditional workflow. If I'm using dynameshed hard surface models in zbrush, then more often than not it will be either to:
- create something stylized with a hand-carved look to the edges (blizardy), in which case bevels are going to be carved in with a trim/flatten brush
- to combine several base shapes (additive or subtractive) into a single solid volume (in which case the bevels are already made and can be tweaked just like a normal mesh)
That is where the speed of zbrush hard-surface is when compared to all the "how do I make this shape flow into that" and "why is my cylinder getting pinched" threads. The trade off for the speed gained there is the time that will need to be spent redynameshing the model every time a change is made. Its a quick process all by itself, but it could easily stack if revisions get out of hand. The trick is to have all the base shapes stored either within the subtools, or within your project hierarchy in the most efficient way you can keep them. Then it doesn't take long to restore a mesh, adjust the subtractive bevels as needed, and recombine.Personally I'm an advocate of using zbrush in conjunction with other programs, so a bevel for me would be tweaked externally and then sent back to zbrush to reboolean. If I were doing it all in zbrush then I would store the sharpened additive state as a morph target and layer. That way any edge could quickly be restored to the original sharpness and rebeveled as needed.
It's not as fast as meshfusion, but it can still be faster than trying to make an acceptable sub-d mesh, if the project allows it.
In addition to being parametric, they are easy to modify after the fact. For certain objects like pipes and wiring, NURBS are definitely the way to go.
The nice thing about Maya is that it gives the best results with the NURBS to Polygon conversion process. You have quite a bit of control there especially with the recent reduction tool that was added since 2014 for a cleaner mesh. One final bonus is that you get the best possible UV map base to work with for cylindrical objects after the conversation.