3D Topology

1. What is topology?

Topology in 3D is the structure and arrangement of a model's vertices, edges, and faces. It is the way to organise your 3D vertices in a detailed, efficient and neat way.

2. Three types of topology:

• Triangle

• N-Gon

• Quad

3. Four points that collated will make a polygonal 3D model/wireframe:

• Vertex

• Edge

• Face

• NURB's

---

Each type of topology and its pros and cons:

1. Triangle

   
   

Tri or Triangles are often used in low-poly modelling or making 'hard' surface structures like architecture. They have three vertices and edges.

Pros:

-triangles can be used to create any shape including quadrilaterals.

-computers will read triangles a lot faster as it's a flat three pointed shape

-fixing N-Gons

Cons:

-if you model with quads, putting in tris will create a dent into the model and may create a 'pinching' effect and blemishing on the surface.

-tris get extremely tedious unless modelling on very low poly

-sometimes the shading will be off despite looking fine in the surface viewpoint when rendering tris

2. Quads

Quads have four vertices and four edges, short for quadrilaterals or polygons with four points.

Most modelling is made by quads due to clean topology and versatility.

Pro:

-loop cuts work better on quads than they do on other topology forms like tris meaning you can save time

-quads form clean straight edges and are very suitable for models with use of perfectly straight lines and right angles as opposed to tris which can often have two long sides and a short one meaning it will not all be a 90* angle.

-quads will give you a predictable output polycount when you subdivide.

Cons:

-quads have a higher polycount compared to tris

-certain file formats may require models to be triangulated before exportation in some game engines or specific 3D formats

-in some specific detailed areas, tris may be more suited instead of quads.

3. N-Gons

An N-Gon is a face or a polygon in a mesh with 5 sides or more. They are used in hidden areas and sometimes to save polycount.

Pros:

-can be used perfectly for hard, flat surfaces that won't be animated or deformed.

-can be used in hidden areas that aren't seen by the viewer to lower polycount

-good for temporary quick mass modelling

-making caps of cylinders as its faster than using quads.

-saving polycount when bevelling planar edges.

-simplifying UV meshes

Cons:

-N-Gons might cause issues if you want to deform a mesh

-N-Gons can cause subdivision surface modifiers as they cause the geometry to overlap which animates weirdly and creates problems with shading.

-creates unclean topology which makes the final animation deformed and unclean

-they create render and smoothing issues occasionally

-if you wanted to export your model into another software, using N-Gons may change the appearance

-N-Gons used on real-time models will mess with the render output and render time.

---

Good vs Bad Topology:

The one on the left is incomplete with incorrect pinching and incorrect usage of quads. You can see the inconsistent sizing of quads which doesn't follow the geometry of the face at all. There's too much poly in places that don't need it. This topology will create deformation and bad frame rates.

The one of the right follows the proper geometry of the face lines and curves. There is the right condensation of quads where there needs to be, particularly around the eyes, ears, nose and lip areas and the pinching looks smooth.

---

Polygonal Modelling/ hard surface modelling:

Hard Surface modelling refers to creating man-made models with no curves or smooth edges. Usually used for making inorganic or non-living things like cars, buildings, furniture or other static non-moving machinery.

NURB's Modelling:

NURB's Modelling (Non-Uniform Rational B-Splines) are a type of Bezier curve. They follow the four point rule of: degree, control point, knots and an evaluation rule. NURB's modelling is using 2D geometrical shapes, lines and curves to create detailed and free 3D models using mathematical equations. They are highly accurate and flexible and use splines instead of polygons.

Subdivision Surface Modelling:

Subdivision Surface modelling is the process of manipulating lower resolution models to create a smoother surface and create high resolution models. It creates rounder curves by increasing the number of vertices on a model.

Non-Manifold Geometry:

A non-manifold geometry refers to any edge that is being shared by more than two faces. This usually occurs when a face or edge is extruded but stays in its original position which causes overlapping of two identical edges. It may cause difficulties for printers to read the file and interpret properly and cause other issues if you try to unwrap it.

Two-Manifold Geometry:

Two-Manifold Geometry refers to polygons with non-overlapping edges in which they can be split along and unfolded to create a flat mesh of the model.

Polygonal:

Polygonal modelling is based on the usage of shapes with faces edges and vertices to create 3D models. It is preferred in VFX and gaming as you can use it to create meshes with accurate and deliberate topology.

Primitive:

Primitive modelling is the usage of standand 3D shapes like cylinders, cubes, spheres and pyramids and creating models with them. They are simple for the computer to understand and beginners to start with as they can be sculpted into anything and also come with all softwares.

Box modelling:

Box modelling refers to the usage of primitive shapes to create a basic draft of the model, of which is then slowly rendered with more and more detailing. Box modelling is used because it can quickly create an outline of the model and also be controllably rendered as simply or as detailed.

Extrusion:

Extrusion is the act stretching out a flat 2D asset into a 3D shape vertically. Only points, lines and polygons can be extruded. Extrusion benefits the creation of walls, boundaries and building blueprints.

Edge and Contour modelling:

Edge and Contour modelling refers to the process which a model is made polygon by polygon and the gaps in between are filled. Instead of using a primitive shape, the polys are looped around the outline to achieve a specific model mesh that box modelling would make it difficult to do.

Procedural:

Procedural modelling is when the computer creates 3D models that the artist can modify the parameters based on the software and needs. This is mainly used for natural environment like trees/forests/beaches or any other package the software can offer that are difficult to do by hand.

Digital Sculpting:

Digital sculpting is when the model is created using brushes that can pinch, push, pull, smooth, grab akin to real-life clay sculpting. The artists can create models with higher pace, accuracy and efficiency making higher resolution meshes with millions of polygons.

Image based modelling:

Image based modelling is when artists make 3D art from the likes of a 2D static image. It is used when time is limited or to visualise real-life objects in VR/AR.

3D Scanning:

3D Scanning is when technology is used to capture real-life objects or environments by recording the 3 dimensional size, colour, texture and shape which is then used to make 3D models. This is used for a more accurate, to-scale realistic model of inanimate objects/environments.

---

3D softwares:

1. Autodesk Maya

Autodesk Maya is a massively comprehensive 3D modelling software used in the industry, equipped with animation, modelling, rigging, texturing, UV mapping to a mature extent. It covers all of the CG pipeline moduele and is flexible, catering to most scripting languages and using CAD which means it can be simultaneously used with other CAD softwares like Inventor and SolidWorks. The price is $225 per month, or $1785 annually. Maya is a favourite of 3D modellers, animators, lightning and FX artists in the film and games industry. However, it often crashes a lot in the beginning and takes a long time to understand how to use. Modelling especially with details is often done in other softwares as its sculpting program is still lacking in comparison to Zbrush and Blender.

2. Zbrush

Zbrush is a software made primarily for the pursuit of art. It allows users to create models like how you would sculpt clay and comes with a tremendous amount of tools and also 'Zclassroom' where they teach every skill to you for free. You can manipulate millions of polygons to create fine detail and organic matter in Zbrush more efficiently than other softwares so its often used in tandem with Maya. Zbrush is around $795 one time purchase on Mac; there are also other student financing options avaliable. However, Zbrush works differently from other softwares as there is no camera or quad views so essentially you can only turn the actual model around and only see it from one angle. Its specifically geared towards organic modelling so hard surface modelling and rendering can be preferred on other softwares like Blender.

3. Blender

Blender is a free to use open source software that has a strong set of all-in-one rounded modelling tools, digital sculpting capabilities, powerful renderer. It is widely used by beginners and small indie game comapanies or simply for a quick modelling exercise. It has a small download package and is quick to set up as theres no need for a license for easy installation and since its free, it has one of the biggest supporter base for tutorials online more than any other 3D software. Learning blender first for sculpting can lead to easy transition to a paid software like Zbrush. However, it has quite a steep learning curve and crashes/bugs more often than Maya or Zbrush. Although it's free, it is much better with paid add-ons from the community. Most companies will also not accept Blender as professional industry standard and often times the rendering process can be slow and remove all fine detail.