Pushing Blender Vertices
A mesh in Blender consists of a set of vertices that are connected by edges. Edges connect to each other to form faces. When you tab into Edit mode on a mesh, you can manipulate that mesh’s vertices (or edges or faces) with the same basic grab (G), rotate (R), and scale (S) tools that work on all objects, as well as the very handy extrude (E) function.
These actions form the basis for 3D modeling, so much so that some modelers refer to themselves as vert pushers because sometimes it seems that all they do is move little points around on a screen until things look right.
Of course, modeling has more to it. You actually have a choice between three primary methodologies when it comes to modeling:
Box modeling: As its name indicates, box modeling starts with a rough shape — typically a box or cube. By adding edges and moving them around, the artist forms that rough shape into the desired model. Bit by bit, you refine the model, adding more and more detail with each pass.
This technique tends to appeal to people with a background in traditional sculpture because the processes are similar. They’re both primarily subtractive in nature because you start with a rough shape and bring about more detail by cutting into it and reducing that shape’s volume.
If you need to add more volume to the mesh outside of the initial box shape, you select a set of edges or faces and extrude them out or pull them out. If you need to bring part of the mesh in from the initial box shape, you select those edges or faces and either extrude inward or just pull them in. Box modeling is a great way to get started in modeling, but you run a danger of ending up with really blocky models if you aren’t careful about how you move your edges around.
Point-for-point modeling: Point-for-point modeling consists of deliberately placing each and every vertex that comprises the model and creating the edges and faces that connect these vertices. The process is actually not as bad as it sounds. You can think about point-for-point modeling like drawing in three dimensions.
And as you may expect, this technique appeals to people who come from a drawing background (or control freaks like me!). The advantage of this method is that you can control the final look of your model, and you’re less inclined to end up with a boxy shape. However, some beginner modelers fall into the trap of getting too detailed too quickly with this technique, so you have to be careful.
3D Sculpting and retopology: Within the last handful of years, this approach to modeling has taken hold of the 3D computer graphics world to the point that it’s now the dominant method. The process works like this: using specialized sculpting tools in 3D software, you start by creating a model with no regard at all for topology, or how the vertices, edges, and faces are arranged in your mesh.
And then after arriving at a the form you want for your model, you retopologize (retopo for short), creating a second mesh with cleaner topology, based on the shape and form of your sculpt. The retopo step uses a combination of specialized retopo tools and the traditional modeling methods described in the preceding bullets. Initially, this technique may sound like you’re doing double the work, but it almost always produces better results and is a much more comfortable way to work for artists with a traditional art background.
The figure shows the difference between a rough human head started with box modeling techniques, a point-for-point method, and sculpting.