Topology vs retopolgy?

[ATTACH=CONFIG]330395[/ATTACH]I was wondering, basically topology is a bunch of meshes basically and retopology is making less faces for better animation? why is that a lot of people do box modeling then go over the model again with retopology? is it because they need the model to map the quad mesh easier and for better edge flow on the more dense model?

I mean, I take it retopology is better for animation correct? Less mesh and better quads, and easier for the animator to rigg?

But I thought, for movies and animation the more detail, the more realistic and up close viewing it gets? so this is why i am confused on topology and retopology in which is better for animation and rigging.

I would like to get this detail ( picture below) but not sure if its retopology that will give this much detail? if so, wouldn’t it be better to study topology edge flow to be able to get the model right the first time and not have to do a second mesh over the first one for better deformation?

Topology is the surface and how it is shaped. You try to arrange things so that deformation works correctly, and use as little as necessary to get your form. Retopology is the practice of creating a new mesh to get clean topology on top of a mesh that is correct in form but messy for deformation.

modeling is the merger of 2 characteristics – art and geometry. Art has to do with the shape, size, details, basically how the model looks. Geometry has to do wih how the model is made, quad faces, even face area, consistent edge flow. In the process of modelling it is often easiest to separate the two – i.e. start sculpting or box modeling, and ignore all rules for good geometry, just get the model to have the shape you want it to have. then, go back over it and retopologize it, this time the shape is taken care of, so all you have to worry about is good geometry.

If you’re goal is to create a single still image, retopo may be unnecessary, but if your goal is animation, clean topology is critically important.

Actually, from my understanding, topology simply refers to the way the quads are arranged in the mesh to form the shapes. Good topology makes it easier to animate.

Retopology, on the other hand, is the process of creating a clean, usable mesh for animating and texturing. While it is used sometimes on meshes that are box or edge modeled to help reduce the polygon count, the process of retopology is more commonly used to take a sculpted mesh (particularly one done with dynamic or multires sculpting) and make it a clean mesh with good topology. Retopology uses the same techniques and processes used in box or edge modeling, but instead of doing it from a reference image, you are using the sculpted version as a guide of where to put the parts of the clean mesh.

Quads have a reputation of being able to subsurface and deform better and in more expected ways than other polygons, which is why they are sought after by those wanting to do animation. You’re right in that the more detail it has, the better the result, but when you don’t have clean topology, the results that you get with deformations or in texturing might not be quite what you expected.

Topology is a knowledge of surfaces. In CGI, how to apply effective geometry for specific purpose.
Retopology is applying that knowledge to a predefined / sculpted / scanned / non effective mesh.
I go by:
Only polys allowed.
Less is more and faster.
Tris for stills, quads for animaton.
Keep the flow, continuity.
Beauty is in order.

I thought, modeling the topology is more for high resolution, then you can create and bake the normal maps then, use the retopology for a less topology mesh and use the maps for the high detail?

that makes sense, then somehow bake a normal map and use that on the retopology mesh?

You can do retopology from any mesh, whether from box or edge modeling or from sculpting (either dynamic or multires). Retopology is used for reducing poly counts, but it is also used to create a clean mesh to use for animation and deformations in posing.

It is completely possible to do modeling and not have to do retopology. In fact, a lot of meshes, particularly “hard surface” meshes, won’t even need retopology. But “organic” forms, particularly characters, get retopology done to them frequently. And yes, the high frequency detail, in those cases, are mapped to be applied to the “low poly” mesh.

I’ll admit, I haven’t done a lot of character modeling, so I haven’t done a lot of baking of normal maps, but what I’ve said is my understanding, based on what I’ve learned from the tutorials I’ve watched, the books I’ve read, and going through the Mastering Modeling in Blender course at CG Cookie.

MeshLab is also nice and open.