Blender 3D: Noob to Pro/Performance vs. Quality

From Wikibooks, open books for an open world
Jump to navigation Jump to search

Mesh[edit | edit source]

Polygon Count - when you work with meshes in 3D, be aware that the memory usage in Blender with reference to meshes comes from the number of faces (or vertices) the mesh contains. That means that if you have many useless faces (such as 55 faces on the side of a mesh that are as flat as one face), the number of faces will directly affect the Render time of Images and Animations. Also remember that for smooth rounded meshes, you do not need an infinite number of faces to make it look really nice and smooth. You actually don't need that many faces to keep a smooth slope looking smooth when viewed from the side. The viewer's eye will not actually identify every flat face, but will trace it as a smooth curve.
Face Structure - 4-side (quad) vs.3-side (tri) - Poly Count is very important when working in 3D graphics, and depending on the type of mesh you are making you could be wasting a vast number of faces because they are not the right structure. For example, if you want a smooth object, you can make it look mostly smooth when viewed from the side if you make it with well-structured 3-vertex faces. You will also yield the same results with fewer faces than if you use 4-vertex faces. Another example is that if you need to apply UVTexture, or have something to animate, a mesh being evenly deformed will yield a lower face-count if it is a 4-vertex versus a 3-vertex face-structure.


It is worth bearing in mind, however, that most graphics libraries deal exclusively with triangles. This means if you are planning to export your mesh to a game or other external 3D application, whether you use quads or triangles will make no difference as all quads will likely be converted into two triangles anyway.

Lighting[edit | edit source]

Using RayShadows or Buffer Shadows - There are a few options you have for using shadows. Ray Shadows use an advanced algorithm to trace the edge of any interfering objects and create a perfect shadow onto the receiving object(s); however, the Ray Shadow calculation is memory-intensive and can seriously slow down your Render-Time. Buffer Shadows, on the other hand, use a different algorithm for similar results. The difference is that buffer shadows use a bit-rate of shadow "pixels" that fill in the shadowing area. You can adjust the bit-rate to make the shadow higher or lower quality. This calculation is much more memory-friendly, and your Render-Time will not jump up as much as with Ray Shadows. An interesting topic because this will change with better hardware performance and eventually won't matter. Until we get into 3D.