If you’re a 3D artist or designer, it is important to educate yourself on the different types of 3D modeling, even if you won’t use them all. Why, you ask? Well, because knowing the theory of what you specialize in can always come in useful. Even if you use NURBS modeling 90% of the time, knowing about parametric modeling will let you know exactly when it’s the best option.

Let’s walk through 10 different 3D modeling techniques and talk about the popular use cases for each of them.

Polygonal Modeling

In polygonal modeling, every single object has a surface comprised of triangles and quadrilaterals. One of the most popular places where you’ll find polygonal modeling applied is in video games. 

Think of old Nintendo 64 games and how blocky the characters looked back then. It is easy to see those characters were made of polygons. These days, video game models are incredibly smooth, even spheres, but even what looks like a perfect sphere can still be comprised of thousands of straight-edged polygons.

Anything with 3D models moving in real-time will benefit more from using polygonal modeling since it is more efficient. It’s perfect for making characters, environments, weapons, etc.

NURBS Modeling 

Back in the days when I used to use Autodesk Maya, I learned about NURBS modeling. NURBS stands for Non-Uniform Rational B-Splines and it uses mathematical equations to determine the shape of curves and surfaces. 

With polygonal modeling, you can’t make a perfect sphere because at high magnifications, you’ll still see millions of flat-faced triangles or quadrilaterals. With NURBS modeling, you can, since equations are used to make the surfaces. This is perfect when you desire to create perfectly smooth objects.

NURBS modeling is best for 3D models that won’t be interacted with and are otherwise static. This could just be a high-detail 3D render of a car, or you could use it for architectural or engineering visualizations.

Subdivision Modeling

Subdivision modeling is also referred to as SubD. It’s interesting because it starts off with a low-polygon mesh that continues to be refined over and over until you get a very smooth and high-detail surface. 

If you’re trying to make a cube, you obviously don’t need subdivision modeling. But if you’re trying to design the complex curves of a hypercar’s chassis, then that’s where SubD could come in handy. 

In a way, it gives you a similar smoothness to NURBS modeling (though not directly comparable) while providing more speed than you can get when trying to build the same shape with purely polygons. This can come in useful for 3D animation, high-end video games, and industrial design.

Sculpting

If you’re trying to make a character or creature’s face, none of the 3D design methods really beats out sculpting. Similar to sculpting in real life, this is like using clay. You can choose to start off with a basic low-polygon sphere, and then pull, push, brush, and smooth over the sphere.

Using a combination of these, you can pull out a nose, push in wrinkles and eye sockets, add pores, and so on. If you’re looking to create organic forms, there’s no better way. The more detailed you want to get, the more powerful the computer you’ll need though (here are some great options for 3D artists in general).

Parametric Modeling

Parametric modeling is a type of 3D modeling that you’ll see being applied mostly in architectural design and civil engineering. The word “parametric” might sound fancy, but it is simply derived from the word “parameter”. 

If you’ve ever played a video game where you can design a character and adjust a “height”, weight”, and “musculature” bar and see those changes on your character in real time, then you already understand how parametric modeling works. 

In this case, think of a spiral staircase with railings. Imagine you have parameters for your baluster height, baluster width, rail thickness, staircase height, and spiral tightness. Adjusting any of these will modify the staircase model in real time, allowing you to develop a design that fits specific criteria you might have.

Procedural Modeling

If you know a little about video games, you might have heard of how a lot of games, particularly in the roguelike genre, have “procedurally generated” environments/maps. This basically means they are generated at random based on a specific algorithm, allowing for nearly endless options.

This is procedural modeling! It allows you to create numerous 3D models, usually for natural environments or cityscapes, without needing to manually design them each time. This is perfect when you need differing complexity and you need it fast.

Boolean Modeling

When it comes to 3D modeling styles, boolean modeling is certainly one of the lesser-known ones. However, it’s really not that complex when you actually hear what it involves. It’s about creating shapes by combining them or subtracting parts of one from the other. At its core, it uses true and false values, which is where the “boolean” comes from.

You can combine two objects into a single one, subtract one object from another one, or keep only the overlap between two intersecting objects. You need to get creative, but Boolean modeling can be very helpful under many circumstances, such as when trying to create hollow objects.

Volume Modeling

Volume modeling most popularly comes in the form of voxel modeling, which uses a 3D grid and voxels (three-dimensional pixels). Voxels can be added to this grid to create a shape; it’s like pixel art, but 3D. A real-world comparison would be building with plain Lego pieces.

Voxel modeling might be useful for some more stylized 3D video games, but in the real world, you’ll find it being applied mostly for medical and scientific visualizations as CT scans and MRIs are capable of returning volumetric data.

Photogrammetry

Photogrammetry is certainly one of the more interesting 3D art techniques that there is. As much as it involves a lot of work on a computer, it also requires a lot of work away from one. This is because photogrammetry involves feeding several photographs into software and allowing it to produce a 3D model from what it sees.

This basically allows real-world objects to be digitized and turned into 3D models and there are countless applications for this. It can be used in architecture to capture buildings, in films for photo-realistic CGI, and in engineering when attempting to reverse engineer something. Photogrammetry can make it incredibly easy to create for 3D model databases.

Hybrid Modeling

Finally, there’s hybrid modeling. There’s not much to say about this technique apart from the fact that it involves combining two or more modeling techniques to get the best that each has to offer. For instance, polygonal and NURBS models can frequently exist together in architectural projects, and a lot of the time when using SubD, you’ll find yourself combining it with regular polygons.

Conclusion

There are several different kinds of 3D modeling techniques and each one has its pros and cons. Even if your work will never take you beyond using polygons and the occasional NURBS, it doesn’t mean that you shouldn’t know about the others. There might be a day when photogrammetry could save you weeks of work (but that doesn’t help if you don’t know about it).

It's worth trying out these different techniques, even if just once, so you can figure out which one is the perfect option for each of your projects. Plus, if you want to make money by selling your models, this can be incredibly useful.