Zspheres II

[Paint Guy]

In fact no one of the actual graphic cards supports native display of voxels. That's why voxels are actually a volumetric calculation method but all ends up with polygons these days, 3Dcoat too. Mostly used algorithm to transform voxels to displayable polys is called "marching cubes", but there are others too. Think of voxels as a bunch of spheres in a raster see wikipedia overlayed with a polymesh for displaying them.

You can't see those spheres in 3DC because they melted together in background mathematically and displayed live as a polymesh. As you may be imagine, there is lot of calculation running in background. Tthis is the reason for performance problems of higher voxel resolutions.

ZB is going a slightly different way...Zspheres are a polygonal representation of a volume. The combination of them is easy, flexible and usable as a rig too. The high resolution polymesh is not generated live as in 3DC but after you're satisfied with your Zsphere model. You can switch between Polymesh and Zspheres easily for editing.

I think this kind of workflow should be possible too for voxels. A bunch of polygonal spheres(or other geometrics) representing a volume with kind of bone like behaviour. Stick them together until you satisfied...and then transform them to voxels...

You said "All ends up with Polygons these days" That's a good understanding of it. So I guess both ZBrush and 3DCoat using polygons in essence but just with different mathematic equations? In ZBrush it seems you basically add details after you have created the general shape or form of the model, whereas in 3DCoat you are adding mass and details all at once.

But it does not even come close to voxels in terms of freedom. By freedom I mean cut,paste,merge,substract functions. You cannot lay out precise hardsurface pieces based on concept art out of Zspheres. It really is for organic basemesh generation,with a very stylish and comfortable system. But frankly I think voxel+quadrangulation is way ahead of that.

I agree I love the way voxels lets you sculpt without constraints but on the other hand the ability to pose and repose a model in ZBrush at any stage of the game is awfully tempting!

Someone mentioned on ZBrush Central that you will be able to import a low poly mesh and then add the ZSpheres to pose it so you aren't constrained to using the ZSpheres for sculpting a base mesh. This along with GoZ will really allow for rapid modeling.

I drool at the ability to be able to pose my models within 3DCoat like they will now be able to in ZBrush. Let's say for the sake of argument that in 3DCoat we wanted the ability to create an armature or bones either before or after we have created the voxel sculpt similar to what ZSpheres is doing.

I think this kind of workflow should be possible too for voxels. A bunch of polygonal spheres(or other geometrics) representing a volume with kind of bone like behaviour. Stick them together until you satisfied...and then transform them to voxels...

Wow, let's hope Andrew get's on it ASAP!

Lastly, do you see any "Disadvantages" to the new method ZBrush is using to apply the muscle mass? In you opinion from a technical standpoint are voxels any better than the new method of adding in mass over a ZSphere? I don't see why a person couldn't sculpt in a base mesh with ZSpheres, then import to 3DCoat, to UV and retop, then export back into ZBrush for details.

[GED]

Lastly, do you see any "Disadvantages" to the new method ZBrush is using to apply the muscle mass? In you opinion from a technical standpoint are voxels any better than the new method of adding in mass over a ZSphere? I don't see why a person couldn't sculpt in a base mesh with ZSpheres, then import to 3DCoat, to UV and retop, then export back into ZBrush for details.

its too early to really say what the disadvantages might be but if its like other features in zbrush it might suffer from a destructive workflow this is just speculation but once youve done a high detail sculpt added sub objects and paint etc will you still be able to go back to your zsphere2 armature and add some new legs and add all the zsphere2 muscle shapes and then go back to your high detail sculpt with all these new parts or will your subobjects mess up or your uvmap be messed or your detail lost or would it simply stop you being able to use zspheres after youve sculpted and applied texture? there are so many things to consider!? perhaps it will work but Im guessing there might be all sorts of workflows/work arounds in order to make the kind of stuff that 3dcoat does so very very naturally in voxel sculpt mode.

[Fabio Dona]

I'm not really impressed by this feature. I'm having so much fun with voxels right now...

I don't know. I'm not really interested in that kind of workflow.

I'm curious if Zbrush 4 will have voxels.

[jedwards]

I think the new zspheres will be a great addition, and will probably bump zb into a more useful tool for doing pure organics. They were already pretty awesome as it is. The ability to pose and define your volumes in the beginning is very important, just like with clay sculpting. That said, 3dc is still strong in this area even without zsphere functionality. I wouldn't complain if it got easier, like in zb, but it's still very easy to get any shape you want quickly in voxels without an armature.

I'm currently tackling a purely organic creature and a fully armored character in voxels and 3dcoat makes it very easy in the beginning for doing both. It's not perfect, but the new zsphere additions won't be either I suspect. At this point I'd much rather see 3dc get the ability to do higher detail in voxels because currently I still think ZB is a better all around tool for that. I don't see the new zsphere additions taking me out of 3dc at all. Maybe for some stuff where appropriate, but more than likely I'll continue to use both for what I feel they are best at.

3dc doesn't just provide easy, seamless merging... it's a very powerful boolean approach to building stuff as well, and should be even further refined to handle more precision modeling as it matures. This is the kind of thing ZB isn't so good at.

[Nemoid_]

I think the new zspheres will be a great addition, and will probably bump zb into a more useful tool for doing pure organics. They were already pretty awesome as it is. The ability to pose and define your volumes in the beginning is very important, just like with clay sculpting. That said, 3dc is still strong in this area even without zsphere functionality. I wouldn't complain if it got easier, like in zb, but it's still very easy to get any shape you want quickly in voxels without an armature.

I'm currently tackling a purely organic creature and a fully armored character in voxels and 3dcoat makes it very easy in the beginning for doing both. It's not perfect, but the new zsphere additions won't be either I suspect. At this point I'd much rather see 3dc get the ability to do higher detail in voxels because currently I still think ZB is a better all around tool for that. I don't see the new zsphere additions taking me out of 3dc at all. Maybe for some stuff where appropriate, but more than likely I'll continue to use both for what I feel they are best at.

3dc doesn't just provide easy, seamless merging... it's a very powerful boolean approach to building stuff as well, and should be even further refined to handle more precision modeling as it matures. This is the kind of thing ZB isn't so good at.

ok my 2 cents here.

Zbrush and 3dcoat are two totally different approaches and technologies. Z brush is polygons, and 3d coat is voxels.

while polygons are voxel representation though marching cubes or other algorythm, technology behind is different from polygons and it allows cut, paste, boolean operation, and its resolution independent, while polygons are not like this, you have to subdivide the mesh or parts of it to get more detail. you can have booleans, but the final shape has to be calculated and closed into a polygonal mesh and this can bring errors and create an unclean mesh.

z spheres II is very interesting, mainly because its a different workflow to build up characters.

It will be more natural for perople coming from a sculpting background in real life foe sure.

It will offer better rigging/deformation of the final mesh for sure, and, regarding zspheres will offer better results spheres placed into complex ways that are sometimes difficult to manage right now.

The volume mass you obtain, is something like a unified skin mass, very clean and subdivided equally, thus allowing good sculpting after youre happy with the general shape.

Notice that a mesh can be rigged even now in Zbrush You can use zpheres as a sort of armature yet to deform a character, but surely pixologic has enhanced this making this the focus of this zsphere II workflow. So characters will be posed with that system, now, rather than with the transpose tool.

Very probably users will be able to store in ztl file zspheres tructure plus the muscle mass they obtain, for further modification. this will allow to use models as a base for several ones as it happens yet now with z spheres /adaptive skin, but into an empowered way.

About flexibility, i dunno if you'll be able to further add modification to a model when you are in painting or detailing stage. Maybe editing it when you're in muscle stage is better and more possible.

Voxels have great advantages too, btw. They allow for inorganic modelling too, very well,and allow to grow detail resolution independent which is way cool. I dunno about rigging /posing. however at the end of the process, u can always generate a low poly mesh and rig that one, and project details with disp and normal maps.

right now, after all, you'll be forced to do this, when working into traditional 3D apps, for further animation, because they use polys and subdivisoon surfaces, and rig poly meshes with deformers (bones)and lattices.

So, even with Zbrush you'll be forced to retopo the final mesh or have a system through which you obtain a low poly mesh good for animation, with good topology for correct deformation, to work with other traditional 3D apps. Lets say that as time passes, this is more and more a limitation of traditional 3D apps rather than a disadvantage of apps like ZBrush.

But still, traditional 3D apps offer a real 3D environment in which you can build your scenes with several objects, put lights, cameras, setup dynamics and more, and rig into a controllable way.

It willl come a time where the different approaches that apps like 3dcoat /ZBrush, and traditional 3D apps offer, will somewhat collide, since these sculpting apps offer a more artistic approach to DCC for sure!

Sorry for the long post. Hope it makes some sense

[cuffins]

ok my 2 cents here.

Zbrush and 3dcoat are two totally different approaches and technologies. Z brush is polygons, and 3d coat is voxels.

That's the point. These two are totally different approaches and technologies. I think it's upon the artist which tool is used for which kind of task. I'm using both ZB and 3DC. Each one for it's benefits to get out the best result in the shortest time. The feeling of real clay is great in 3DC and very intuitive. I can let go and concentrate on sculpting only... If you satisfied with a static object it's all you need. But if you seriously have in mind doing an animatable Mesh you have to retopo and texture yur work and bring it over to an other app for posing and animation. I would actually never do hard surface work in 3DC seriously because NURBS is state of the art for this because of freedom in filletting, booleans, blending and so on. Yes, you can do it in 3DC too but not effective at the moment.

We should think about what's the goal of 3DC? In my oppinion it's not good for the app to mimic other packages but bringing new features (like voxels) to stay in the market. Filling the gaps...

Retopo tools in 3DC are another great exemple of this. For me they're the best in the market for now. So I suggest to improve these tools consequently instead of implementing features other applications offer.

my 2 pence

[Nemoid_]

That's the point. These two are totally different approaches and technologies. I think it's upon the artist which tool is used for which kind of task. I'm using both ZB and 3DC. Each one for it's benefits to get out the best result in the shortest time. The feeling of real clay is great in 3DC and very intuitive. I can let go and concentrate on sculpting only... If you satisfied with a static object it's all you need. But if you seriously have in mind doing an animatable Mesh you have to retopo and texture yur work and bring it over to an other app for posing and animation. I would actually never do hard surface work in 3DC seriously because NURBS is state of the art for this because of freedom in filletting, booleans, blending and so on. Yes, you can do it in 3DC too but not effective at the moment.

We should think about what's the goal of 3DC? In my oppinion it's not good for the app to mimic other packages but bringing new features (like voxels) to stay in the market. Filling the gaps...

Retopo tools in 3DC are another great exemple of this. For me they're the best in the market for now. So I suggest to improve these tools consequently instead of implementing features other applications offer.

my 2 pence

well unless of huge technology changements you'll always have to retopo the mesh, if you build it entirely with sculpting, both in ZBrush or 3D Coat. Traditional 3D app work with polygons and deform low/mid poly meshes, not meshes with millions of polygons because they start to literally crawl down their knees updating animation and solve IK and deformations in such situation.

Whats actually cool about apps like ZBrush and 3D Coat is they allow user to think at shape before, sculpting it quite from scratch, and only afterwards to how the mesh is interms of optimization and polyflow.

This is great especially for organic modelling, where you want a natural and believable shape. With poligonal modelling you can do great models however, but if forces you to think to polygons first and then actual shape so it ends to be a bit more technical.

This also brought sometimes a division creating pipelines in which you can have a sculpting artist creating the shape for a character or a creature, and then, a modellig artist creating the poligonal cage for it for animation purposes.

Another thing thse apps brought are displacement and normal maps to give to models a more natural and detailed aspect at render stage.

About retopo i would add that 3D coat has great tools for it so no real problem having to retopo meshes into it. Zbrush i hope will have better tools than current ones for retopo purposes.

I agree with the fact is better to refine and enhance existing 3D Coat toolset and performace before starting to add new stuff. Voxel is an excellent technology to keep pushing and an advancement even over current Zbrush one. When 3D coat will be as smooth as Zbrush for sculpting, competition ill be really huge!

[renderdemon]

ZSphere 2 seem to have a great advantage over voxel sculpting,a not destructive approach.

Even if I really like voxel sculpting,posing a mesh is hard,and you can't do mistakes,you can't jump back and forward between poses.

[Paint Guy]

ZSphere 2 seem to have a great advantage over voxel sculpting,a not destructive approach.

Even if I really like voxel sculpting,posing a mesh is hard,and you can't do mistakes,you can't jump back and forward between poses.

I agree. Using Voxels in 3DCoat is fine if you only want a static model but the new ability in ZBrush to "pose" and "repose" a model at any time during the sculpting process is a definite show stopper! I think the current 3DCoat posing tools should be scrapped for a "bone" type posing system that would allow you too easily pose a voxel model and then making the bones exportable to other 3D apps for animating.

[Fabio Dona]

I think the current 3DCoat posing tools should be scrapped for a "bone" type posing system that would allow you too easily pose a voxel model and then making the bones exportable to other 3D apps for animating.

The current pose tool in 3d coat is fine. I know it's not perfect, but it's more than a "pose" tool. It's like a generic transform tool. You can do more than just pose characters. The name is misleading.

Could be useful to have skeletons nonetheless, Just keep a separate set of tools for that.

Yet, I'd like to have simple things that are still missing in 3dcoat, like masking, than any of that...

[ggaliens]

I see no reason why a SKELETON could not MORPH Voxel surfaces

almost as well as they seem to morph dense meta-ball like structures (ZBrush).

Seems like it would ne easy, at least for me.

[calilifestyle]

i have to say after watching that ZII video i really want to buy it. i mean being about to build you model around a structure you want as the base. Then moving it around to better sculpt.

Man i would kill for some kind of skeleton system in 3dc and for those Lw3d users to also have weight maps come along with the structure.

[polyxo]

Zspheres 2 show a fascinating technology. I can imagine getting access that this feature is a Must for anybody

who only creates creatures and also animates these.

But is it also useful for anything but humans/animals and maybe trees - for geometry without "extremities"?

[kay_Eva]

i have to say after watching that ZII video i really want to buy it. i mean being about to build you model around a structure you want as the base. Then moving it around to better sculpt.

Man i would kill for some kind of skeleton system in 3dc and for those Lw3d users to also have weight maps come along with the structure.

I don't want to buy Zbrush =) but yeah that feature is pretty sweet. Andrew is very smart. I think he can also see how great it is.

[sgiff]

Why does it have to be like ZSpheres, a long time ago, Impulse Inc. made a software called Organica which could take any primitive shape and twist and bend it like clay but it was a metaballs like system which worked well but it was old technology. Andrew could do this better I think.

[geo_n]

Amazing thing is all these updates are free up to zbrush 2.0 I heard. They made lots of money being number one for a long time with no competition.

[philnolan3d]

Just thought I'd throw this out there. Meats Meier posted the news about ZS2 on Twitter the other day. I said I was frankly happy with 3DC curves for this purpose (although only for creating one pose) and he replied that he looked around and found my video online (I assume he means

) and that it "looks like a nice tool.". I think the 3DC marketing needs some work if the big guys seem like they've never heard of it. Obviously Andrew doesn't have a huge marketing department, but we can all do our part.

BTW I agree that a skeleton would be really nice, I was asking for that in the early days of voxels.

[stevecullum]

Wow, just saw the demo - that's some pretty awesome base mesh creation method!

I always felt trying to create a base mesh from curves in 3DC a bit clunky and limited, but with ZS2 I can create the base mesh in there and then import to 3DC for detailing and texture painting

[ozukaru]

I think the big problem with z-brush is still its interface. And you need plug-ins to do things that in 3d-coat are built in.

So if I was pixologic I were concerned about that mainly, for its next release. Anyway this z-spheres II is still very interesting, if they follow this way they'll end up with a complete skeleton -> muscle -> skin -> animation system!

[Javis]

They only thing I thought was impressive is the rigging structure to manipulate the zspheres. This could easily be done with voxels and polygons of course. I have details on a planned feature but, I don't know if I should really mention it? Andrew would be better suited to comment on that one of course.

[GED]

I dont really see how a skeleton would transform voxels, surely if you moved the hand over to scratch the head and then pulled away bits of the hand/ fingers would be stuck to the head :p

[Paint Guy]

They only thing I thought was impressive is the rigging structure to manipulate the zspheres. This could easily be done with voxels and polygons of course. I have details on a planned feature but, I don't know if I should really mention it? Andrew would be better suited to comment on that one of course.

OK stop teasing us like that. that's not fair! lol. Does it have something to do with "Posing" voxels because if it does you might have just saved me from spending $600 on ZBrush.

So what is it?

[ggaliens]

Well ... if you RIG the Voxel shape with "bones",

before you re-pose it into a collision based mess,

then you still have a point of reference from which to

recover and PULL a sword out of a rock or a hand back

out of a head.

Simple. I can think of technical hurdles ... but they

don't seem very confusing to me.

[stevecullum]

This could easily be done with voxels and polygons of course. I have details on a planned feature but, I don't know if I should really mention it?

Yeah come on, spill the beanz!

Don't want to spend money on ZB unnecessarily if 3DC is going to get something along these lines...

[JamesE]

They only thing I thought was impressive is the rigging structure to manipulate the zspheres. This could easily be done with voxels and polygons of course. I have details on a planned feature but, I don't know if I should really mention it? Andrew would be better suited to comment on that one of course.

First, conirm with Andrew that it's ok. Second, SPILL IT!!!

I'll wait right here till you get back with an answer =]

[Taros]

They only thing I thought was impressive is the rigging structure to manipulate the zspheres. This could easily be done with voxels and polygons of course. I have details on a planned feature but, I don't know if I should really mention it? Andrew would be better suited to comment on that one of course.

Oh oh... this post was a big mistake.

[philnolan3d]

This is all news to me as well.

[cuffins]

They only thing I thought was impressive is the rigging structure to manipulate the zspheres. This could easily be done with voxels and polygons of course. I have details on a planned feature but, I don't know if I should really mention it? Andrew would be better suited to comment on that one of course.

come on... give it to us...

[David Walters]

I always felt trying to create a base mesh from curves in 3DC a bit clunky and limited

I find them a bit unweildy too, I probalby just need a bit more practice -- perhaps a manipulator-less 'tweak' feature for positioning would be a good request? I find that very useful for rapid mesh layout.

[Javis]

Lol, I don't know much guys! I just know what Andrew has told me.