Meshes
3D Modeling Fundamentals: Understanding Meshes
What are Meshes?
A mesh is fundamentally any collection of three basic components: Vertices, Edges, and Faces.
The Building Blocks of Meshes
Vertices: The Foundation
The vertex is the first fundamental unit of any mesh. Each vertex contains:
A Position: X, Y, Z coordinates in 3D space
A Direction: A normal vector indicating orientation
Other data: Additional information such as color, UV coordinates, or weights. We will go into this later.
For now, think of a Vertex as a floating point in space, that has no surface, no volume, and can be linked to other vertexes by means of an edge, and that is capable of holding other forms of data.
Edges: The Connections
If you have two vertexes, you're capable of defining an edge - which is a "line" connecting the two.
Edges serve as the connective tissue of meshes:
Edges define the wireframe of the mesh, and therefore:
The fundamental topology that determines how this mesh deforms and bends. Proper edge flow and triangulation of edges will be a common and reoccurring theme in this field.
Edges Continued: Edge Loops and Edge Rings
Two critical concepts emerge from edges and their relationships to other proximal edges - the Ring, and the Loop.
Edge Rings: Successive edges that are parallel to each other. These represent a conceptual "ring" around the mesh.
Edge Loops: An unbroken chain of edges connected end-to-end. These create continuous paths through the mesh.
Key insight: You may have noticed that rings and loops are inter-related, and naturally form a grid structure.
Rings and loops run perpendicular to each other, and in fact all rings can be freely bisected with loops to create additional rings. These concepts are fundamental for effective modeling workflow.
Edge Metadata
While vertexes typically store the "data" associated with the mesh, manipulating that data is commonly done via edges. For example:
Seams: Mark where UV unwrapping should split the mesh
Creases: Define where subdivision surfaces should maintain sharp edges
Bevels: Information for edge beveling operations
Sharp/Smooth: Control how lighting calculations treat the edge
Note: I want to highlight however, that this meta data isn't "universal data", its DCC specific, and is only valid while we're in Blender. Once we send the model to another program, it is anyone's guess how that program interprets this data.
Triangles: The Foundation
From Vertexes, and Edges, we can now define Surfaces, and the simplest and most fundamental surface type would be the Triangle.
Named because it is naturally made up of three edges, and therefore three angles - tri +angle.
Created by connecting three vertices via edges and enclosing that structure with a surface
This is not a triangle - this is just three vertecies, and three edges, in the shape of a triangle. 
This is an actual triangle. The edges and verts have been enclosed and designated as a surface.
Known by various names: tris, polys, polygons, faces, geo etc
All geometric detail eventually becomes triangles in the rendering pipeline.
Triangles are "geometrically unambiguous" in their surface definition. There is only one way to make a triangle from three vertexes.
Triangles possess a "Face Normal", which is direction that determines the way the triangle is "facing". Triangles are therefore said to have a Front Side (front face) and a Back side (back face).
The face normal of this triangle, pointing up and away. Blender has a toggle to display the front and back faces of triangles, and when enabled (as of blender 4.4.3)front faces will render solid (as above), while back faces will render red.
The back face.
So, they are the simplest, most fundamental polygon, and can be represented in a way that provides absolutely no ambiguity in surface or direction, and this is why they are fundamental - they are the only polygon type that has these attributes.
Note: You may have noticed that this word "Normal" keeps showing up. Think of "Normals" as a direction, and in the case of triangles they point perpendicularly away from the triangle.
Quads: The Modeling Standard
When we join two triangles together, and remove the inner edge delineating them, the structure that remains is called a Quad.
Quads (4 vertices) are the second most basic type:
Named because they are comprised of four edges, four verts, four angles
Internally are actually represented as two triangles, but this representation is often hidden from the user.
Also known as Polys, Polygons, Geo, etc.
Note - Triangles are often ALSO referred to as a Face, Poly, Geo, etc and you may be aware of the term "polycount".
Whenever you refer to an objects "polycount" - always be referring to the number of TRIANGLES this object is comprised of. Tricount and Polycount are therefore synonyms.
Quads suffer from surface ambiguity, in that it is often difficult to know ahead of time which way it is triangulating
Lets take the vertex on the right, and lift it up and over, folding the quad in half like a sheet of paper. 
Wait what. We couldn't fold it the way we wanted to? This quad wasn't triangulated along the Red edge, but infact along the blue edge.
This ambiguity can result in confusion, and shading errors, so while modeling you should keep your quads "as planar as possible" (as flat as they need to be).
In cases where this is not possible, then we must triangulate them.
This ambiguity can ALSO resolve differently in the viewport, vs in the game engine. A quad exported from blender may triangulate one way, but a quad exported from Maya may triangulate another way.
We must therefore model in such a way that this ambiguity will not dramatically change the final shape of the mesh.
Why Quads Are Preferred for Modeling:
Natural compatibility with ring and loop edges
A sheet of quads is by definition also a sheet of loops and rings.
They create regular grid geometry
High compatibility with fundamental techniques and modeling tools.
They subdivide uniformly into themselves.
Can be easily bisected with loops.
Important note: While quads have historic preference in modeling, use common sense rather than dogmatic thinking. Quads are preferable in many cases, but are not universally required as a rule.
N-Gons: Avoid as a beginner, and always use with Caution
N-Gons are defined as "Any polygon having greater than four edges", and therefore define all other polygons.
Like quads and all surfaces, they are also comprised of triangles internally.
NGons are almost universally avoided for a number of reasons:
They have significantly more surface ambiguity than quads
They cannot be triangulated unambiguously
Thus, they also cannot deform and animate predictably
Cannot be bisected with loops like, like triangles
When triangulation finally does occur, the different programs may have wildly different implementations on how to triangulate an NGON
Some programs functionally are incapable of importing them (ZBrush)
Some programs will crash when importing them (much less of a problem in 2025 than it used to be)
Unwrapping commands arn't always accurate with particularly shaped NGons, and will often yield a highly distorted UVIsland
So, there are many cons to NGons, and admittedly several pros that we havn't covered here. Therefore, as a beginner, I want you to universally avoid them. Keep your meshes triangulated to either triangles or quads.
Learning approach: Avoid N-Gons entirely, but as proficiency increases, you'll discover edge cases where they're quite beneficial.
Additional Mesh Attributes
There's a lot more data stored in a mesh then what we described above, but here is a rough list of some of them:
UV Coordinates
UV coordinates create a mapping between the 3d world, and the 2d world. :
Transform 3D mesh surfaces into 2D texture coordinates
Typically operate in a 0-1 range, where (0,0) is bottom-left and (1,1) is top-right of the texture
Can extend beyond 0-1 range for tiling effects or special mapping techniques
Allow textures to be applied correctly to complex geometry
Essential for texture mapping, materials, and surface detail
Material Information
Material assignments: Per-face assignment determining which material/shader affects which faces
Commonly implemented as material indices (Material Index 0, 1, 2, etc.)
Vertex Colors
RGB Color data stored per vertex for basic coloring without textures
Part of vertex information, not material information
Normal Vectors
Control surface orientation for lighting calculations:
Vertex normals and Face Normals
Normally calculated automatically, but can be manually adjusted.
Animation and Rigging Data
Weights - Used with bones and animations
Morphtargets / blendshapes - Used for fine control over specific expressions and surface displacements.
Well done!
I understand that is a fair bit of theory to lead with, but this is pretty fundamental stuff, and its ok if you don't fully understand it yet. We just want these concepts in your mind as we move forward.
Key Takeaways for Students
Meshes are fundamentally: Vertices → Edges → Faces
Vertices hold data- they are not just positions.
Edge loops and rings are crucial: You will learn to think in terms of loops and rings.
Always prefer triangles for final output: They're unambiguous and render-ready
Use as many quads as you reasonably can: They provide the best workflow flexibility
Avoid N-Gons initially: Master basic polygon types first.
Pay attention to normals: The back face of Triangles often render invisible or see through. If your mesh isn't showing where you think it should, think back to your normals!
Last updated