Question About Modeling Methods

[jtemple]

Hello,

I’m currently working my way through this course as well as another Udemy course on character creation and they’ve each used a different approach for character modeling.

Method 1 (this course), blocking out the character and then sculpting it.
Method 2 (the other course), starting from a single plane just above the mouth and extruding/loop cutting/adjusting to create an entire body.

I’ve attached my current progress on the models below.

My main question is if anyone can briefly discuss the general pros and cons of each of these methods?

For context, I’m looking to use these models as game assets in Unity, so an amount of modularity (especially for enemies) and low performance cost is ideal. Most characters would be wearing armor or clothing (which is a whole other skill set to tackle, but that’s for another time ), so while getting a good body shape is important, it doesn’t exactly have to be super high detail.

Also a secondary question, the way we’re currently sculpting the hands, is this a standard workflow? Will they be able to open and close normally when we’re done with the course or will these particular hands be balled up as fists forever?

Thanks for any responses I receive.

[FedPete]

It all depends on your end goals!

• Sculpt adds realistic details to the model. But then for games dev you need to retopology the model.
• With block modelling (extruding vertices, edges) you have control of a goof topology. But it lacks in details if you need them. The next would be sculpting and less need for retopology.

It also depends on how your mind works. So like mesh model/hard surface, others are comfortable with a block or concrete. How does a sculptor see a figure (possibilities) in a block of wood?

Sculpting takes a lot of experience, tools. But much faster in designing organic shapes with a lot of details. Making this mode fit for fast animation (games) is difficult also and time-consuming. But much better control on how your mesh is designed. Especially where animation is involved.

Block modelling gives you clean mesh, but it is difficult to create an existing model. Like your toes. correct mesh maybe, but too much for such a tiny detail. Or just good when you want to use close-ups of the feet … This means, know upfront where and how you are going to use the model.

Don’t worry too much about this. Many students say I’m going to reuse this later when I have more knowledge. But with more knowledge, you see shortcomings in your model. better to start fresh again.

So Learn all different work processes Blender. So you can choose the right technique for a project.

[FedPete]

@jtemple

Also a secondary question, the way we’re currently sculpting the hands, is this a standard workflow? Will they be able to open and close normally when we’re done with the course or will these particular hands be balled up as fists forever?

Hands are difficult, so the tutor opted in for a quick solution as in a fist. But if you want real hands go for it. You know how it needs to be done (block or sculpt). The Orc is not about sculpting an Orc, but the blender tools to create something.

[jtemple]

Great, thank you for your response! I think my brain is more comfortable with block modeling, but I’m enjoying the organic feeling of the sculpting. I haven’t gotten to the retopology portion of this course yet so I do not know how painful that is yet.

I suppose there’s no real rules against using block modeling to set up a basic body shape (which is going to be obscured by clothing anyway) and then coming in with sculpting for areas like the face and hands which can use the extra detail, ya?

[FedPete]

@jtemple

and then coming in with sculpting for areas like the face and hands which can use the extra detail, ya?

Yes is a work process. It is like being a carpenter. A wide range on tools from chisels to sandpaper, from saw to axe. Pick your tool, save regularly to a new version number. Especially when destructive work is done.

[bOBaN]

Basically FedPete adressed all of your question in great way, but let me add my 2c, from Blender beginner, but a bit more experienced (still beginner though :P) gamedev perspective.

If you aim for low poly look, box modeling should be faster. The more complex the character you make the harder is to make them with box modelling techniques. I actually think that at this time sculpting is used for almost every high-poly character in games (and in movies?).

Even if as FedPete pointed out - box modelling is giving you ‘cleaner’ topology… it’s not cleaner by default. I certainly make a mess of my topology quite frequently. It comes with experience, maybe even many years of experience.

And another point to consider is that sculpting is extremely useful tool, even if you do hard surface modelling. Putting dents, distorting the shape a little, etc. (i.e., things that make objects more believable, as nothing is perfect in real world) are in my experience much easier to do with sculpting tools.

As for clothing. If you do it in Blender, then again in my opinion combining hard surface (basic shape) and sculpting (adding more organic/believable look to it like folds in cloth or dents in armor) is good approach. Though if you look at what industry is using for clothing - marvelous designer seems to be dominating that space (it’s not cheap program though).

It’s great that you are learning both sculpting and hard surface. After a while you will learn what tool is best for YOU for a given job. And knowing both will give you a greater flexibility.

PS. It would be great experiment if you would re-do both the characters in opposite techniques afterwards. That is, doing the Orc in a hard-surface way, and the boy (?) with sculpting. Don’t know if this exercise is worthy time investment though.

PS2. The “Orc course” covers some basics of armor and clothing creation

PS3. Both character are coming along nicely

PS4. As for retopologizng - if you find the process unsatisfactory (I certainly did), there are tools that making it easier like Instant Meshes (there are also paid plugins for Blender like quad remesher that are producing excellent results). I don’t know how long it will take, but I have a hunch (call it an educated guess as I work in AI, but not graphics related), that in not too distant* future tools for automatic retopologizing will get so good that this step will not be necessary. (* I think it’s better to be more precise this days about it - I’m talking in perspective of couple of years for Blender, not couple of months - specifying Blender here as some tools already are close to it - like 3ds max)