What is a 3D Rendering?

You may have heard the word render a thousand times, but you might still not know what it means exactly. Here, you'll find the answer.

In general terms, rendering for a 3D and CG job, also known as 3D rendering, is the process of reproducing an image based on three-dimensional data stored in a computer. This takes raw information from a 3D scene (polygons, materials, textures and lighting) and calculates a final result, which is known as "output" and is usually a single image or a series of compiled images.

A rendering can simulate realistic lighting, shadows, atmosphere, color, texture and optical effects such as light refraction or blur seen on moving objects. It may also not be realistic at all, made to transform the output into something that looks like a painting or an abstract image.

What is a 3D rendering used for?

3D renderings give audiovisual creators new possibilities. It allows videos to be made without having to record real locations or objects. Industrial designers use them to simulate specific situations and use cases. For architects, they are a powerful way to visualize buildings, complexes, houses or even to recreate ancient structures. Advertisers and graphic designers use them to get more striking results. They are also used in video games, digital art, medicine or even in gastronomy, to offer more immersive experiences.

The 3D rendering process

3D rendering is a creative process similar to photography or cinema: starts with an idea, takes shape through composition and design, then its materialized and illuminated–even staged–to produce the final image. The big difference between the former and the latter is that, in 3D rendering, the scenes that are portrayed do not exist and everything we observe must be created on the computer before it can be rendered.

And while it is true that this can mean a lot of work, it also allows enormous creative control over everything that appears on the scene and how it is represented. Perhaps the most complex aspect of this process is the time it takes. Sometimes it's a lot, even on computers designed to excel at 3D processing. The reason is that the software has to create each pixel for the image, a process that involves a large number of calculations, including the tracing and handling of light, its behavior and how it bounces on the different objects of a scene.

Something important to keep in mind is that, although rendering is usually the last stage of a creative process, an exception can be made when a rendering is taken into Photoshop for further processing.

Main 3D rendering methods

There are two common types of rendering: CPU (central processing unit) rendering and GPU (graphic processing unit) rendering. The CPU is the computer's processor, its main function is to execute sequences of instructions and do the general processing of data. The GPU, on the other hand, is the graphics card installed in the computer, which focuses on performing specific calculations and reducing the workload of the processor (CPU) in terms of graphics, vectors, matrixes, vertices and pixels.

· CPU based rendering

CPU rendering is mainly used in movie studios and is also the favorite process for architectural visualization. This is due to its accuracy in creating photorealistic images and because rendering times are not a consideration for these industries. A scene with flat lighting and materials with simple shapes can be rendered in a matter of seconds, while a scene with lighting and HDRI models can take hours.

It is also possible to render more advanced lightning effects using the CPU, with techniques such as:

- Ray tracing: each pixel in the final image is calculated as a particle of light that simulates interacting with objects in the scene. It is excellent for creating realistic scenes with advanced shadows and reflections, but it requires a lot of processing power.

- Route plot: calculates the final image based on how the light will illuminate a certain point on a surface of the scene and how much it will be reflected in the camera. The process is repeated for each pixel of the final image. It is ideal when striving for photorealism.

- Photon mapping: the computer shoots "photons" (light rays) from both the camera and any light source used in the final scene. This technique is often used to simulate caustics–the envelope of light rays reflected by a curved surface or object–as light refracts through transparent surfaces.

- Radiosity: similar to route plot, although it takes into account the light sources that have already been reflected on other surfaces of the scene. It is perfect for simulating the softer and more realistic shadows of a complete scene.

· GPU based rendering

GPU rendering is used for real-time processing. Its used in video games and interactive applications and very common when you need to render 30 to 120 frames per second and get a smooth experience.
Real-time rendering does not allow some of the previous CPU rendering techniques to be used. To overcome these limitations, other effects that deceive the eye and make the elements look smoother may be implemented, like motion blur on moving objects.

However, GPU rendering does not always have to be used in real time, it can also help in complex CPU renderings and it is a good method to show the first results of a final piece–a kind of preview for the –without having to wait hours. This makes it a very useful tool in terms of 3D workflow, especially when applying light and textures.

What rendering engine to use

A rendering engine is the software or plugin that allows you to generate a realistic final view from a 3D model. Each render engine works on the basis of the GL (global lighting), that is, it calculates the light produced by the different light sources at a point and its average on a given surface.

There are dozens of rendering engines; In fact, virtually any 3D software has its own built-in rendering engine, but they may have limitations compared to specialized rendering engines.

Three of the most important and recommended rendering engines are:

- V-Ray: perhaps the most common one. It is capable of using both CPU and GPU processing, so it is very flexible. It is available for Maya, Blender and virtually any 3D software.

- Corona: the favorite of many designers and architects. It is very powerful, although only available for 3DS Max and Cinema 4D.

- RenderMan: developed and used by Pixar studios for their productions. It can be used as a complement with Maya, or as a standalone product on Windows, Mac and Linux computers.

You may also be interested in:

- Cinema 4D Tutorial: cloning methods in MoGraph
- The 20 essential shortcuts to boost your Cinema 4D productivity