Character model base mesh?

Just wanted to look for different opinions on the use of a base mesh for character modelling. Do you sculpt straight from a cube or other basic mesh, do you have a specific base that you use as a template, or do you prefer to work with polygons directly? What are the pros/cons of each (if you can think of any)?

I’ve tried to sculpt straight from a cube and ended up with a badly disfigured blob, so if anyone here has any opinions on how they go about this sculpting characters I’d be interested in them.

Moved from “General Forums > Blender and CG Discussions” to “Support > Modeling”

I prefer the more calculated method rather than the organic one of sculpting. Like you, my attempts at sculpting ended up in the same disfigured blob.

IMO, I think it speeds up learning to have a base mesh. A prefigured mesh provides is a template for comparison initially. For instance, if you are going to model a hand poly by poly, then I would use a base mesh hand for comparison. Same for an arm or leg.

But if you are going right in to create a character mesh, then I would absolutely start with a base head, and body if needed. It absolutely speeds up work.

The base head should not be too detailed nor have too dense of polys. Except around the eyes and mouth. Those loops could be very fairly numerous. That allows for flexibility in creating more density and loops.

The body can be more defined as desired.

Personally i prefer to model my base mesh first, not only because i can, but because it makes sculpting much easier when you have already all your proportions in place and the form is already there.

If you’re not very used to modelling, there’s a possible alternative i have been playing with sometime : using the Skin modifier.

By example, with the skin modifier i build that
http://i.imgur.com/6Rpx634.jpg

I apply then the modifier, but as the skin modifier has some unresolved problems (face sometime overlapping, holes etc…) , in Edit Mode you fix them, and while you’re there you get your base model more in shape (using proportional editing by example), and if you want, you can add little details to make it easier for sculpting (like the hands and fingers i modelled in that case)
http://i.imgur.com/iyS8PB1.jpg

Then from there, you can go sculpting, for this kind of base model without much modelled details, the best thing to do in my opinion is use the Dynamic Topology function instead of Multiresolution
http://i.imgur.com/lsod87v.jpg

The big advantage of Dynamic Topology is that it does not care at all if your base model topology is not good, while Multiresolution really require a very good topology to start with, so you don’t face some detailling problems at higher multires levels

I recommend you read this post from this artist is very good and gives good tips on modeling in general, This is in Spanish so use google to translate:




Thanks for the responses. Didn’t even know what skin did until now, but it looks particularly useful for getting proportions of the body right, etc. Currently I’ve been practicing with polygons + dyntopo to gradually add detail. Possibly once I get some anatomical studies done (e.g. /uploads/default/original/4X/8/d/b/8db621a440f7d645f9db884e279e35b5ffd85050.jpgd=1389334853), I can snap a few planes to them to create low-poly references. Also, bookmarked the tutorial link for future reference.