animation technique 3/4: understand everything about 3D animation in 10 minutes

image credit:
Toy Story 4, by Josh Cooley at Pixar
10
minutes of reading
The third article in this series focuses on 3D design and its production constraints, from modeling to final rendering.
published on
13
October
2021

You now know almost everything about animation, whether traditional or computer-based. All these concepts lead us naturally to the third dimension.

The depth

I spoke earlier about vector animation, based on points and X and Y coordinates. Well, in 3D, it's almost the same thing, except that the points also have a depth coordinate, Z.

In 2D, two points are all you need to draw a line and see a shape appear.

In 3D, the smallest visible element is a polygon, defined by at least three points.

3D vector and polygon representation in 3D space


Objects can therefore only be assemblies of polygons, which create surfaces and therefore volumes.

To create an object, you need to imagine it entirely made up of polygons, then calculate each of the points of intersection of these polygons.

And that's exactly what the fathers of modern 3D, Ed Catmull and Fred Parke, did in 1972, transforming the former's left hand into a combination of polygons, and calculating (by hand) each of the points of intersection. Thus was born the very first 3D animation.

Note thatEd Catmull, who probably doesn't ring a bell, is no less than one of the founders of Pixar, and the former president of Disney Animation Studios !

However, we had to wait 12 years and the arrival of John Lasseter before admiring the first 3D animated short, The Adventures of André and Wally B..

Modeling

The first stage in the production of a 3D film is therefore the creation of objects from polygons, and this is called modeling.

Today's algorithms make it possible to create objects with four-sided polygons. A cube is therefore simple to model: just assemble 6 polygons.

But modeling a sphere becomes much more complex, because in 3D, the only way to round corners is to add polygons.

So the more detail you want in your scene, the more polygons you'll need to add, and the more polygons you have, the more calculations you'll need, because it's not enough to add a single polygon, as you can see in the animation opposite (from 6 to 600 polygons).

Computing power is therefore one of the first hurdles to overcome when creating in 3D.

Low-Poly animals by Pavel Novák

And that's why the Low-Poly trend emerged a few years ago: objects built with a very small number of polygons, which become visible, for ever faster calculations. A constraint that has finally given rise to a style.

On the other hand, some softwares allow you to go very far in the details of modeling by allowing you to sculpt directly in the software (like ZBrush, for example, or even Blender, a very powerful open-source 3D software).

Now that we've modelled our object, let's apply a texture!

The material

Textures from the AmbientCG library

Glass has reflective properties that sand does not.

Wood is rough, marble can be polished, water diffracts.

All these parameters are adjustable potentiometers that allow us to imagine an infinite number of original materials, which can make this work very time-consuming, especially if you're looking for photorealism.

But the material only shows its full potential thanks to an essential element of a 3D scene: light.

The light

And the light was there.

After the creation of heaven and earth, God created light (for yes, God is the first 3D artist).

And so are we.

Because it shapes, it reveals, it atmospheres.

Technically, you need to place the right lights in the right places, at the right intensity, and in the right color, just like a photographer in real life.

It sounds simple, but it's not. Good light management makes all the difference. Light isn't just essential, it' s life. You've been warned!

Our object is now modeled, materialized and illuminated. How do we get an image of it?

The point of view

The big advantage of 3D is the use of cameras. Once everything is in place, all you have to do is find the best angle to get the best image.

3D cameras have the same properties as real ones: depth of field, angle of view, focal length.

And unlike 2D, you can look ahead, behind and above without having to recreate anything. Which means you can offer original camera movements.

Framing, and therefore point of view, is where storytelling begins 🙂 🔥 !

Photographic printing

Objects, lights and cameras are in place, and now the little bird has to get out.

In 3D, this is called rendering: the machine will calculate the path that each ray of light will take between the camera and the various lights.

As you can imagine, these calculations are managed by highly complex algorithms that take up a lot of resources on the computer's processor (the CPU - Central Processing Unit). That's why most of them have been deported to graphics cards (or GPUs - Graphics Processing Units) to reduce calculation times.

The Mandalorian in a 3D set (the screen behind the actor)

eal-time display is the holy grail of any 3D artist, and certain environments such as Unreal Engine now enable the creation of entire realistic backgrounds that can be used directly on a filmed scene. The Mandalorian series was the first to experiment with this.

In all cases, you'll need a large machine with as many graphics cards as possible, or use render farms (sites full of servers equipped with graphics cards) which can be very expensive if your scene contains millions of polygons.

The movement

As you now know, video animation is simply a succession of images.

3D animation is no exception to this rule. However, a single image can take several seconds or even minutes to calculate, for the reasons outlined above.

Imagine that an image takes 30 seconds to calculate, so a one-minute film at 25 frames per second would take 12 hours to calculate!

This means not only optimizing the scene so that each image can be calculated as quickly as possible, while maintaining a certain quality, but also providing the equipment needed to carry out these calculations, which can require a substantial budget.

Finishing touches

As a general rule, 3D animation often comes out very rough, and integrating certain effects directly into the rendering can greatly extend calculation times.

That's why it's best to go through a final finishing stage, called compositing. This consists of assembling several image sources (photos, 3D images, illustrations) to create a final image that is homogeneous and pleasing to the eye.

3D rendering by Nidia Dias

At this stage, we add a host of effects such as depth-of-field blur, motion blur, colorimetry, grain (to simulate film), smoke, and many others at the artist's discretion.

There are professional software packages dedicated to this task for film (such as Nuke), but for motion design, After Effects is widely used.

And that's it! Your film is ready to be exported, edited and mixed with vocals, music and sound!

A few tools

To create and animate 3D objects, you have a choice of tools, which are becoming increasingly sophisticated.

In motion design, most studios use Cinema 4D, an easy-to-use program that's relatively quick to learn (consider that 3D software is extremely complex to learn). Numerous plug-ins complement this program, such as the Redshiftand Octane rendering engines, or the X-Particles particle engine. Zbrush, a popular sculpting program, is now integrated into the Maxon subscription.

Other studios and freelancers also use 3DSMax, another powerful 3D software package.

You may prefer to use the free Blender software, which already lets you create complex scenes with a high-performance rendering engine.

To create professional character animations, studios generally turn to Maya software.

Another comprehensive program, Houdini, is also often used by production studios, but is much more difficult to access because the way it works on a scene is very different from all the other programs. This one is based onnodes .

Last but not least, Splineis an online 3D software package that lets you create computer-generated images directly in the browser, as well as interact with 3D objects. Their software is expanding very rapidly, and their community is very active.

You'll also find extensive online libraries of 3D models, textures and HDRI images, free or not.

Conclusion

Illustration by duo Cabeza Patata for Spotify

3D has many attractions, but it is complex and time-consuming to set up. It can also be (very) expensive.

The constraints associated with this technique also give rise to very different and original styles that will enable you to stand out from your competitors!

So, are you ready for a 3D movie?

Photo de profil d'Alexandre Sobrier, motion designer freelance
Alexandre Soubrier

Alexandre Soubrier is a freelance motion designer with a passion for illustration and animation. He created the podcast Exquises Exquisses in which he interviews author-illustrators, and produces this blog.
Contact him here or on Linked In.

Discover other articles:

To stay informed, subscribe to our newsletter:

Thanks! Your registration has been taken into account!
Oops! There was a problem, try again in a few moments.