Mon, Dec 23, 6:48 AM CST

Renderosity Forums / Poser - OFFICIAL



Welcome to the Poser - OFFICIAL Forum

Forum Coordinators: RedPhantom

Poser - OFFICIAL F.A.Q (Last Updated: 2024 Dec 22 10:18 pm)



Subject: Sculpting or Modeling... Which Gives Better Results?


EClark1894 ( ) posted Sat, 11 October 2014 at 1:48 PM · edited Sat, 30 November 2024 at 11:04 PM

Just wondering which one the modelers here prefer? I'm still learning and while I can see the benefits of learning to sculpt, the question is... "is it necessary to learn to sculpt, particularly if I can achieve the same results through modeling.




bopperthijs ( ) posted Sat, 11 October 2014 at 2:26 PM

It really depends on what your goal is: if you want to make buildings, cars and other man-made objects and you need precision, 3D-modellers is what you need. But if you want to make organic objects like rocks, people, animals and so on, sculpting is an easier way to go(nowadays). The difference is that with 3d-modellers you can keep control about the polygons you are creating and with sculpters like ZBrush you don't. As an engineer which I am, it was for me very easy to use a modeller like Rhino3D, because I got precision, but to use a sculpter like Zbrush it was something completely different (a different cookie, like we dutch say): you need to have some artistic skills and to let the urge for precision go.

That said, I like Zbrush very much and it has learned me a lot of things an skills. Today for me Rhino is for work and Zbrush is for fun/

best regards,

Bopper

-How can you improve things when you don't make mistakes?


shvrdavid ( ) posted Sat, 11 October 2014 at 2:27 PM

It depends what you want to do with the models you create and the way you prefer to create them.

If you want to create normals maps for the wireframes you create, scuplting is usually the way they are created. Normals maps don't require the mesh to have every detail modeled into it.

If you want to create wireframes from sculpts, that is a little different since you end up making a new wireframe by retopology after the scuplt is done.

 



Some things are easy to explain, other things are not........ <- Store ->   <-Freebies->


ockham ( ) posted Sat, 11 October 2014 at 6:05 PM

I was reminded of the difference lately when revising my anatomy courseware.  It uses some animations that I made back in 2000 when I was just 'discovering' CG.  At that time I was using the rather strange sculpting app Anim8tor to make things like lungs and larynxes.  Since then I've been using Amapi exclusively for everything. 

I realized that there's no way I could make those squishy parts NEARLY as well with Amapi, even after 14 years of supposedly improved skills.  Amapi is great for houses and cars and terrains, but not great for strictly organic stuff.

My python page
My ShareCG freebies


unbroken-fighter ( ) posted Sat, 11 October 2014 at 10:19 PM

normally i only work on hard bodies like  buildings and cars but even with those sculpting can be done for interior pieces like a couch or a pillow or seats

i learned in blender and still use it today because it allows for sculpting, modeling, texturing, rendering, and rigging all at a good price range  $000.00

im cheap but even if i had to i would buy blender because it is multi-use and as powerful than most paid softwares

but in the end which to use depends on the user


EClark1894 ( ) posted Sat, 11 October 2014 at 11:42 PM

Interesting... a nuimber of people say they use regular modeling on items like cars, but seems to me if there was at least one inanimate object I'd sculpt, it would be a car.




unbroken-fighter ( ) posted Sun, 12 October 2014 at 12:35 AM

most things can be made from modeling but can also been mae by sculpting

its one of those deals where its all about personal prefence

like MBR has been sculpting a sea horse alien thing that looks good and sculpting is perfect for it

 making a car body you need the low poly count and accuracy of modeling

there are people that combine modeling and sculpting

just do what is easiest for you


NanetteTredoux ( ) posted Sun, 12 October 2014 at 2:19 AM

I believe that you can't achieve the same results through modelling. You need both modelling and sculpting.

Poser 11 Pro, Windows 10

Auxiliary Apps: Blender 2.79, Vue Complete 2016, Genetica 4 Pro, Gliftex 11 Pro, CorelDraw Suite X6, Comic Life 2, Project Dogwaffle Howler 8, Stitch Witch


ockham ( ) posted Sun, 12 October 2014 at 8:02 AM

Correction: It wasn't Anim8tor, it was Amorphium.  That was a LONG time ago.

My python page
My ShareCG freebies


maxxxmodelz ( ) posted Sun, 12 October 2014 at 9:44 AM · edited Sun, 12 October 2014 at 9:55 AM

The workflow of most professionals outside the Poser community today begins with Zbrush, then ends in something like Maya or 3dsmax.  The topology is usually sculpted with very high detail first, then retopologized with standard polygons as quad topology, with proper edge flow, for rigging and animation.

Standard modelling will provide you with the best edge flow, and most manageable poly count.  Sculpting will provide you with much higher level of detail, at the expense of both of those other things.

You will want to learn, eventually, to do both very well.  You'll want to begin by modelling your base model, then taking it to something like Zbrush for high poly sculpting.  Then create normal maps or displacement, and then back out to another application like Topogun or 3dsmax for retopology again, where those details get baked into the lower poly mesh, and appear only at render time.

Some modelers use sculpting as a way to very quickly flesh out a base from which they could build a standard quad mesh with fast retopo tools, rather than just box modelling or surface modelling, which can be time consuming from scratch.  It's a faster workflow if you're just painting quads over a high res sculpt.

Either way, the end model should be reasonable in poly count.  If you're sculpting, you aren't going to offer a model at 1 million triangles for redistribution, when it could be modelled properly in quads, with the same detail, at say 100k polys with normal maps or displacement.  That's just silly.

Proper edge flow modelling will manage a detailed model much better than a triangulated sculpt.  So learning how to model traditionally is still the best way to create clean, usable models.


Tools :  3dsmax 2015, Daz Studio 4.6, PoserPro 2012, Blender v2.74

System: Pentium QuadCore i7, under Win 8, GeForce GTX 780 / 2GB GPU.


Cage ( ) posted Mon, 13 October 2014 at 7:43 AM

Quote - Correction: It wasn't Anim8tor, it was Amorphium.  That was a LONG time ago.

I was going to say.  :lol:  Amorphium was maddeningly imprecise.  Anim8tor was... maddeningly precise.  I suppose I may have learned something about how my brain works, when I found I could use the former just fine but I could do nothing with the latter.  Made some neat stuff with Amorphium.  Not great Poser meshes, but we had very few options, back then.  Pre-Blender, pre-Wings3D days!  Eep.

I'm most comfortable with Wings3D, now.  I've used 3DCoat for a few things, but I'm not very comfortable with it.  It's one of those programs I would probably need a hardcopy manual to be able to master fully.  I once spent... way too much on Modo, but have only managed to use it for texture painting.

 

And we're starting to babble now.  Back to the world of coffee....

===========================sigline======================================================

Cage can be an opinionated jerk who posts without thinking.  He apologizes for this.  He's honestly not trying to be a turkeyhead.

Cage had some freebies, compatible with Poser 11 and below.  His Python scripts were saved at archive.org, along with the rest of the Morphography site, where they were hosted.


moogal ( ) posted Mon, 13 October 2014 at 9:33 AM

Just noticed about an hour ago that Anim8or is being developed again...  Cool.

http://www.anim8or.com/download/index.html


bhoins ( ) posted Mon, 13 October 2014 at 9:36 AM

Quote - Just wondering which one the modelers here prefer? I'm still learning and while I can see the benefits of learning to sculpt, the question is... "is it necessary to learn to sculpt, particularly if I can achieve the same results through modeling.

In CG modeling and sculpting are the same thing, the difference is in perception on how the tool is designed to work and semantics.


AmbientShade ( ) posted Mon, 13 October 2014 at 10:47 AM

Zbrush is a modeler and a sculptor. 4r7 has added a ton of new poly modeling features that are unmatched by just about any other modeling software out there. And now it's finally 64-bit.

Many ZB artists have been using it to model cars, buildings, etc, for years, not just organics.

It all comes down to personal preference, and in the professional realm, what the client requires. And of course, having a thurough understanding of the software you're using.

 

 



EClark1894 ( ) posted Mon, 13 October 2014 at 12:10 PM

I probably would have learned how to model years ago if I knew how to read French. I could never find a copy of Amapi that wasn't in French.

I don't know if I consider modeling and sculpting to be the same thing. I mean sure, like baking and grilling are both cooking, maybe...




AmbientShade ( ) posted Mon, 13 October 2014 at 12:33 PM

Why wouldn't it be the same thing? In both instances you're pushing and pulling individual and groups of vertices into a desired shape. 

Traditional modeling requires you to always be conscious of your geometry and edgeflow. Sculpting allows you to focus on the shape instead, then worry about the geometry later. 

Pretty much everyone outside of the hobbyists have abandoned traditional modeling for sculpting because its much more fluid and allows you to focus on the art instead of the science behind it. 

 

 



bhoins ( ) posted Mon, 13 October 2014 at 1:21 PM

Quote -  

Pretty much everyone outside of the hobbyists have abandoned traditional modeling for sculpting because its much more fluid and allows you to focus on the art instead of the science behind it. 

 

 

I have to question where your numbers come from. The biggest two tools for creating models, in the Professional market, are still 3DS Max and Maya. Neither is a "Sculpting" program. ZBrush and Mudbox are far behind those and likely even behind Lightwave and Modo. Yes you can add Z-Brush and especially Mudbox to a Max or Maya workflow, but that does not mean people are abandoning traditional modeling, certainly not in large numbers.  


heddheld ( ) posted Mon, 13 October 2014 at 1:25 PM

@Ockham glad you made that clear a post or so down ;-) I LUV'ed anim8tor really cant get my head round amiphirom (sp)

@EClark not knowing French is no excuse :-( there was pictures ;-)

modeling and sculpting are different !!!! heck even a wood carver Vs a stone mason are dif ! ok so they would learn each others skills in mins not weeks but if you asked them to work in clay they would start with a huge square block and chop bits out rofl

practise practise the skills you have ;-) learn new ones when you can ! learn to "cheat" eg make that dress dup it add a multires sculpt in the wrinkles bake the displacement map

 

MOST of all have fun [else it turns into work an we all h8 that lol ]


Jaager ( ) posted Mon, 13 October 2014 at 4:22 PM

I suggest that Amapi (or Rhino) are not the best choices for producing items for use in Poser.  Amapi uses a method based on NURBS (in the main).  It makes for smooth and precise curves, mostly because it supplies as many points as needed to generate them.  This makes for an OBJ conversion product that has an untenable number of polys as far as being practical for use in Poser. It is my experience that in using a program to reduce polys automatically - the result is not a reduction where and how you would do it if modelled in a native Mesh Form modeller. 

Now that it produces quads, would not Marvelous Designer be the choice to produce clothing for Poser?

Because I come at this from a morphing point of view, I have been unclear about the difference between modelling and sculpting.  Looking at it for generating a new object, I can see the difference.  I think that if it is something that has or could have a blueprint a CAD type method would produce a more precise result - modelling(?).  For an organic subject without preconceived proportions -sculpting(?).


RorrKonn ( ) posted Mon, 13 October 2014 at 5:32 PM · edited Mon, 13 October 2014 at 5:32 PM

 http://www.zbrushcentral.com/showthread.php?174659-Andrzej-Marszalek-Art-Dump/page2

step 1 you sculpt the mesh

step 2 retopology / remesh for lower polycount

step 3 convert high poly count sculpt in to a normal map "games" or displacement map or vector map for the lower polycount mesh.

sculpting is how you make normal map "games" or displacement map or vector map.

============================================================ 

The Artist that will fight for decades to conquer their media.
Even if you never know their name ,your know their Art.
Dark Sphere Mage Vengeance


maxxxmodelz ( ) posted Mon, 13 October 2014 at 6:01 PM

Quote - Pretty much everyone outside of the hobbyists have abandoned traditional modeling for sculpting because its much more fluid and allows you to focus on the art instead of the science behind it. 

 

Absolutely NOT true.

Have you been to Polycount or CGArchitect.com lately?  Yes, Zbrush is used by most artists now, and is in almost every studio too.  However, your sweeping statement is incorrect.  Zbrush isn't even used at all in the arch-vis world, where some of the most technically proficient poly modellers are, and Maya and Max are still widely used in CG and Game modelling.

To say everyone outside of hobbyists have dropped traditional tools for sculpting is completely inacurate.  You'll still find tons of professional studios looking for Max or Maya modellers, and yes you need Zbrush experience, but not to replace the poly tools in those apps.


Tools :  3dsmax 2015, Daz Studio 4.6, PoserPro 2012, Blender v2.74

System: Pentium QuadCore i7, under Win 8, GeForce GTX 780 / 2GB GPU.


moogal ( ) posted Mon, 13 October 2014 at 8:13 PM

The next version of zBrush is going to have box modeling tools.  That alone could be taken as evidence that sculpting hasn't replaced modeling.

Sculpting is great for organic shapes, complex contours, and for the way that you do not need to concern yourself as much with poly counts and topology until later.  But there are also many things that one would likely not sculpt.

Take a look at various programs galleries.  These are usually the best works produced by that software and they can show you not just what the program is strong at, but what it is weaker at.

In the zBrush and Sculptris galleries you'll see many portraits, body sculpts, and creatures.  But in the zBrush galleries are also numerous mechanical concepts, some of them very stylized and many quite complex.   If you were to look at a collection of images from SketchUp you'd see many more buildings, interiors and exteriors, mechanical components and designs but very characters or creatures.  What you see the most of usually represents the programs strongest features.  One of zBrush's strengths is the speed with which an experienced user can illustrate a concept.  Similarly SketchUp's is valued for the ease with which lets users model geometric forms quickly, cleanly and accurately.  And similarity in two models produced by those programs would belie the huge differences in those programs methods of working and tools provided.     


lupus ( ) posted Tue, 14 October 2014 at 4:46 AM

Quote - It really depends on what your goal is: if you want to make buildings, cars and other man-made objects and you need precision, 3D-modellers is what you need. But if you want to make organic objects like rocks, people, animals and so on, sculpting is an easier way to go(nowadays). The difference is that with 3d-modellers you can keep control about the polygons you are creating and with sculpters like ZBrush you don't. As an engineer which I am, it was for me very easy to use a modeller like Rhino3D, because I got precision, but to use a sculpter like Zbrush it was something completely different (a different cookie, like we dutch say): you need to have some artistic skills and to let the urge for precision go.

That said, I like Zbrush very much and it has learned me a lot of things an skills. Today for me Rhino is for work and Zbrush is for fun/

best regards,

Bopper

I'd say this answere is the best.

If you're going for  making morphs or conforming clothes to existing models in Poser, then you'll need a 3d-modeler.
As people say, use Maya, 3D-max or whatever you want.
I use Lightwave and is happy with that.

Sculpting like in ZBrush I use for fun or when doing a sketch to later work in modeling.


AmbientShade ( ) posted Tue, 14 October 2014 at 12:55 PM · edited Tue, 14 October 2014 at 1:00 PM

I probably should have been a bit more clear: in terms of design where precise measurements are required such as architecture and design firms go, where you're building prototypes for real-world tools and such, then yes, traditional modeling would be a better approach. 

When you're building something like a human figure, or clothing, or scenery, etc, there is no logical reason to still be trying to box model something like that if you have access to zbrush or mudbox as you're just creating more work for yourself. 

The notion that zbrush is not capable of producing architecture and other hard-surface models is completely inaccurate and it doesn't require millions of polys or even 100k polys to do so. ZBrush artists have been doing these types of models for a number of years now and it doesn't require building a base model in another app like maya first. ZBrush has all the primitive shapes you need to start with, with a number of settings and options for each one in its initialize pallet. It can also add in edge loops, extrude, change topology, delete faces, etc, at any point in the modeling process. Those features have existed in zbrush for a while now. The new poly modeling features coming with 4r7 just expands on that. 

The notion that you need a million polys or even half a million polys to sculpt with pretty much always comes from novice zbrush users who haven't learned how to sculpt properly or taken the time to actually learn zbrush itself. I never take my models anywhere close to 1 million polys until i'm ready to start painting or adding in fine details. My models generally stay under 10 - 20k, especially in the early stages, and I never have any problems sculpting them. I only increase resolution when I need it in certain areas. That's another thing with zbrush - you don't have to divide the entire mesh, just mask out the area you need more resolution in, reverse the mask and divide that. You can do all your retopology in zbrush - you don't have to go out to another program. For the moment, I find it's easier to do retopology in topogun, but that will soon become a non-issue as well. I'm using topogun and other modeling apps less and less these days, because I can do more and more in zbrush faster. 

Dynamesh changed a lot of the ways people go about sculpting, and is a great tool for quickly conceptualizing a model as it redistributes your geometry on the fly, where ever your mesh is being stretched too thin. But you don't need dynamesh for everything. I don't use it for architecture, as it is primarily intended for organic shapes. 

Just browse the zbrush central galleries to see numerous examples of architecture and vehicles all created entirely in zbrush. Considering it is an industry standard for just about all the CG films being made today, and have been being made for the last several years, I have to say there is a lot of disinformation being thrown out there in this forum. No, it does not do animation or complex material setups. It's a modeling app first and foremost, and always has been. For animation and complex shaders, you do need another package, at least for now.  

Mudbox began its life as a plug-in for maya, developed by an independent group of students and was bought by autodesk later. I've never used Mudbox as it's price is too high in conjunction with its constant upgrade fees, and I've yet to see it do anything zbrush can't do, but I don't pay attention to what it's latest features are. Zbrush upgrades have always been free.  

 



bhoins ( ) posted Tue, 14 October 2014 at 4:30 PM

Quote - I probably would have learned how to model years ago if I knew how to read French. I could never find a copy of Amapi that wasn't in French.

I don't know if I consider modeling and sculpting to be the same thing. I mean sure, like baking and grilling are both cooking, maybe...

Effectively, sculpting is just using a boolean tool in a traditional modeller. Yes, you have all sorts of presets to choose from to either add or remove mesh, but it is still boolean operations the difference is still just semantics and perception, but the process and principal is the same. 


RorrKonn ( ) posted Tue, 14 October 2014 at 6:02 PM · edited Tue, 14 October 2014 at 6:03 PM

mudbox remesher you can place loops where you want.
hope zbrush 4r7 zremesher can too.

zbrush finally making it to 64 Bit :)

============================================================ 

The Artist that will fight for decades to conquer their media.
Even if you never know their name ,your know their Art.
Dark Sphere Mage Vengeance


moogal ( ) posted Tue, 14 October 2014 at 8:06 PM

Quote - > Quote - I probably would have learned how to model years ago if I knew how to read French. I could never find a copy of Amapi that wasn't in French.

I don't know if I consider modeling and sculpting to be the same thing. I mean sure, like baking and grilling are both cooking, maybe...

Effectively, sculpting is just using a boolean tool in a traditional modeller. Yes, you have all sorts of presets to choose from to either add or remove mesh, but it is still boolean operations the difference is still just semantics and perception, but the process and principal is the same. 

No, it's not.  You could indeed use booloean operations to remove parts of a larger shape, which might be a bit like sculpting marble.  "Sculpting" apps work more like sculpting with clay.  You start using soft selection to form an approximate shape, and then refine that form with various brush types.  Using sub-objects (as brushes) could be considered similar to the boolean "add" operation, but most of sculpting has nothing to do with the traditional boolean operations a "modeling" program might have.  Boolean tools create a specific result based on the boolean operation (AND, OR, NOT or AND NOT) performed on one or more sub-objects. 


EClark1894 ( ) posted Tue, 14 October 2014 at 11:09 PM

You can't use sculpting to make a morph, though... right?




AmbientShade ( ) posted Tue, 14 October 2014 at 11:15 PM

Quote - You can't use sculpting to make a morph, though... right?

yes. I do it all the time. 

 



RorrKonn ( ) posted Wed, 15 October 2014 at 12:29 AM

This is how box modeling in C4D & Sculpting in zBrush works.

box modeling you select a polygons ,lines or a vertices and move them.

sculpting you do not select a polygons ,lines or a vertices
you just slide the brush over the mesh.
zbrush has 100's of brushes that do diffrent things.


box modeling you can add or subtract polygons ,lines or a vertices.

the basic sculpting brushes you do not add or subtract polygons ,lines or a vertices.
your just moving them around.
ya polycount could be a billion so it's a bit mute to worry about adding topology

but we can add or subtract with dynamesh.add horns to a monster.

and we can zremesh there's a lot of tools to do any thing we want.
list of tools
http://pixologic.com/zbrush/features/ZBrush4R6/

videos to see them in action :)
http://pixologic.com/zclassroom/homeroom/


============================================================ 

The Artist that will fight for decades to conquer their media.
Even if you never know their name ,your know their Art.
Dark Sphere Mage Vengeance


AmbientShade ( ) posted Wed, 15 October 2014 at 1:11 AM · edited Wed, 15 October 2014 at 1:15 AM

file_507834.jpg

Here's a quick example of morphing via sculpt.

M4 at default resolution. No subdividing, used 4 brushes (move, inflate, clay buildup, snake-hook) and a quick mask to pull out the ears. Done in less than 5 minutes. Of course it's a quick and crappy example, but it's just a demonstration to show you don't need to divide a mesh to sculpt a morph. Dividing the mesh can create problems when loading the morph back into Poser, so you never want to divide. 

Now how long do you think that might take if you had to select each face, edge or vert in a traditional modeling app? This is how morphs used to be done, before sculpting came along. 

 



moogal ( ) posted Wed, 15 October 2014 at 2:13 AM

The OP question as asked can't be answered without knowing more...

Do you want to push and pull on a form until you like the results?  Use a sculpting app.

Do you want to make a specific known object cleanly and neatly?  Use a modeling app.

Despite the overlap, they are very different.  But you shouldn't say one is better than the other.  Always use the right tool for the job at hand.


moogal ( ) posted Wed, 15 October 2014 at 2:20 AM

Quote - You can't use sculpting to make a morph, though... right?

The only practical difference between a sculpting program and Poser's own morph brushes is that the latter doesn't add/subtract polygons.

Of course a real sculpting app is more fine tuned to this purpose than Poser's brushes, but basic familiarity with the morph tools would be a good start to using any sculpting program.


EClark1894 ( ) posted Wed, 15 October 2014 at 11:24 AM

Okay, then, something I saw recently in a video got lost in translation about sculpting. I actually thought it made things smaller and subdivided vertices and polygons.




AmbientShade ( ) posted Wed, 15 October 2014 at 11:46 AM

Quote - Okay, then, something I saw recently in a video got lost in translation about sculpting. I actually thought it made things smaller and subdivided vertices and polygons.

The model only "shrinks" if you have a smoothing modifier on it. The shrinkage is caused by the smoothing. If the smooth modifier is not on then it will only increase the poly and vertex count of the model.

A vertex can't be subdivided. It's only a point in space. The edges that connect between two vertices are divided in equal lengths by a new vertex. So if you have a single quad face, that is 4 vertices connected together by 4 edges. One subdivision level will add a new vertex to each edge and connect a 9th vertex in the center point, giving you 4 faces from the previous 1 face. 

 



AmbientShade ( ) posted Wed, 15 October 2014 at 12:31 PM

file_507842.jpg

Here's an example of a tri, quad and pentagram (known as an n-gon, meaning more than 4 sides), being subdivided.

In poly modeling, all polygons want to be quads. So when you divide one level, anything that is less than or more than 4 sides will always become quads. 

If you spend enough time memorizing how the polys divide, then you will get to a point where you can look at a model's mesh and see the underlying base geometry. It also helps greatly when building models, because you will know where to use different types of faces (quads, tris, n-gons) to acheive the subdivided edgeflow that you're trying to get. Very useful when building a model intended for animation (or retopologizing one).

 



moogal ( ) posted Wed, 15 October 2014 at 6:14 PM

file_507849.jpg

> Quote - Okay, then, something I saw recently in a video got lost in translation about sculpting. I actually thought it made things smaller and subdivided vertices and polygons.

I hope I don't sound pedantic, these forums are here to benefit people of many skill levels, and I think it's important to understand and use proper terminology whenever possible.  I wouldn't want others being misinformed by my own generalisations or over simplifications.

Whether something is made larger or smaller or subdivided doesn't matter, Boolean operations are a very specific thing.  Achieving the same result of a booloean operation does not necessarily require a boolean operator.  I can use Wings to manually merge two spheres for example, but it would not be a "boolean merge" if it was done manually without a boolean tool.  

In this image you can see the only logical results of a boolean operation performed on objects "A" and "B"; a new object comprised of both "A" and "B", an object defined by the intersection of "A" and "B", object "A" minus object "B", and object "B" minus object "A".

 


Privacy Notice

This site uses cookies to deliver the best experience. Our own cookies make user accounts and other features possible. Third-party cookies are used to display relevant ads and to analyze how Renderosity is used. By using our site, you acknowledge that you have read and understood our Terms of Service, including our Cookie Policy and our Privacy Policy.