As I begin to work on the Circo Volador’s Project, I propose making different characters on 3d that feel and look as 2d, to adjust the graphic style they already have. So first thing I did, was taking all of the currently drawn characters and select one for each workshop they have: photography, circus arts, capoeira..and many others.
This is the draft I choose for this project:
Then what I needed for each one, was a projection which I can use as a blueprint to create the models on 3d. Everyone must have at least a frontal view and a side view, which in combination with some imagination can gives an easy way to create each 3d model.
Right after that, it was necessary to chose a 3d program, at first it was supposed to be Maya, but thinking that this project is meant for learning, I thought Blender, an awesome free 3d program, was the best fit. I had made some little things on it before so I knew its potential.
So first things first, what is needed is to adjust yourself to the way Blender moves and behaves, for that there are many many videos and tutorials, and what I found best, is to take a look at the tutorials, then try whatever you are learning and if there is a doubt on any subject or option or value, take a look on the blender wiki, which in most cases has exactly what you need.
One of the best places to look for neat videos is: http://cgcookie.com/blender/
Next step is know how to model. Blender has a really neat way of model things, it has easy commands and hot keys. The ones that I use the most: extrude (e), make edge rings (ctrl r), fill faces (f) and there is one plug in that make the life easier when extruding faces but trying to keep the shape of the model: Mesh:Insert Extrude
Then there are this things called Modifiers, they let you make changes to the models but with the great advantage that you can change its values at any time, so for example, you can see your model on different amount of polygons at any time.
There are two mesh modifiers that I used on most cases, when modeling any character, and they are Mirror and Subsurf.
Mirror lets you have an exact copy of the mesh you are making on the other side of any axis, The option on this modifier that you have to keep an eye on, is clipping , it merges the vertex that go near the mirror axis, and once they are merged they will not separate until you uncheck that option.
The other one is subsurf and lets you see the model multiplying its poly counts, make them smoother. This is great for previews and little tweeks on how the mesh will look in the end.
The last thing I will cover on this post is what I suggest when choosing between an orthogonal or a perspective view. This is crucial as I discovered that unlike other 3d programs the cameras or perspective view on the viewport, follow really accurate the function of normal camera lenses, which gives you plenty deformation depending on how far or close you are to the objects, which can lead to inaccurate modeling. And as my very personal advice, it’s better to make all the modeling on a orthogonal view, which lets you see the model as it really is.