3D Architectural Rendering: How 3D Rendering Works

Before a building exists, most people first meet it as an image. Competition boards, sales brochures, planning consultations and client meetings all rely on pictures of spaces that have not yet been built. Almost all of those pictures are produced by 3d rendering: the process of calculating a two-dimensional image from a three-dimensional model, its materials, its lights and a virtual camera.

This article explains how architectural 3D rendering works, the difference between offline and real-time approaches, the steps that take a design model to a finished image, where rendering helps the design process, and the limits that anyone commissioning or reading renders should keep in mind.

What 3D rendering is

A 3D model is a mathematical description of geometry. Rendering is the computation that turns it into a picture, by working out, for every pixel, what the camera would see: which surface is visible, what material it has, and how much light of what colour reaches the camera from it.

In architecture, rendering serves several purposes. It helps designers judge proportion, light and materials. It communicates intent to clients, communities and authorities. And it markets projects before they are built. The same underlying technology supports quick design studies and highly polished marketing images, which differ mainly in time, skill and the detail of the scene.

The main rendering techniques

Rasterisation

Rasterisation projects each triangle of the model onto the screen and colours the pixels it covers. It is extremely fast and is the basis of most real-time graphics, including video games and the viewports of design software. Lighting effects such as shadows and reflections are approximated with additional techniques.

Ray tracing

Ray tracing follows rays of light between the camera, surfaces and light sources. It naturally produces accurate reflections, refractions through glass and sharp or soft shadows, at the cost of far more computation.

Global illumination and path tracing

In a real room, much of the light has bounced off other surfaces. Global illumination methods calculate that indirect light; path tracing does so by following many random light paths per pixel and averaging the results. This is what makes renders of interiors look naturally lit. Production renderers such as V-Ray are built around these methods.

Physically based materials

Modern renderers describe materials with physical properties: base colour, roughness, metalness, reflectivity, transparency and surface detail through texture maps. Using physically plausible values is one of the main reasons current renders look convincing.

Offline versus real-time rendering

Architectural visualisation now uses two broad families of tools.

Offline (production) renderers calculate each image in detail, taking from seconds to hours per frame. They give the highest quality and the most control, and are favoured for final marketing images and complex lighting. They are typically used inside modelling programs such as 3ds Max, SketchUp, Rhino or Revit.
Real-time renderers, such as Enscape and Lumion, use the graphics card to draw the scene many times per second, so designers can walk through the model and see changes immediately. Hardware-accelerated ray tracing has narrowed the quality gap considerably.

Many practices use real-time tools throughout design and reserve offline rendering for a small number of final images, often prepared by visualisation specialists.

From design model to finished image

A typical architectural rendering passes through several stages, whatever the software.

Model preparation. The design model, from a BIM or CAD tool, is imported or linked. Geometry is checked for gaps, overlaps and missing detail that will show in close views.
Context and entourage. Surrounding buildings, landscape, furniture, people and vehicles are added to give scale and life. This is often the largest part of the work.
Materials. Surfaces are assigned physically based materials, ideally reflecting the actual specification.
Lighting. Sun and sky are set for the project’s location, date and time; artificial lights are placed and given realistic output.
Cameras. Views are composed at realistic eye heights and focal lengths.
Rendering and post-production. Images are rendered, sometimes in separate passes, and finished with colour correction and minor edits.

For homeowners, a set of good renders can make a design far easier to understand than plans. Our guide on how to read house plans is a useful companion: the plan says what will be built, and the render suggests how it will feel.

Where 3D rendering helps the design process

Rendering is most valuable when it is used to make decisions rather than only to sell them. Comparing façade materials under real sun angles, checking whether a room will be bright enough, testing how a stair reads in a double-height space or how a building sits beside its neighbours are all questions that a quick render can answer better than a drawing.

Renders also support engagement. Clients, residents and approving bodies who cannot read technical drawings can comment meaningfully on an image. In that role, accuracy and honesty matter: camera positions, lens choices and planting should represent the design fairly rather than flatter it.

The limits of 3D rendering

A rendering shows what has been modelled, lit and composed. It does not show whether a design works. The most important limits are these:

Structure is often assumed. A slender column, a deep cantilever or a frameless glass corner can be rendered long before it has been engineered. Real member sizes may differ.
Services are usually absent. Ducts, sprinklers, access panels and drainage rarely appear in renders but will appear in the building.
Materials are idealised. Real finishes have joints, tolerances and weathering.
Lighting can be generous. Exposure settings can make spaces look brighter than they will feel.

None of this makes rendering misleading by nature; it simply means renders should be read as intent. When an image shows an ambitious structural idea, it is sensible to have that idea checked before it becomes a commitment to a client. Our structural engineering consultancy is often asked to do exactly that.

Frequently asked questions

What is 3D rendering in architecture?

It is the process of generating images or animations from a 3D model of a building, using software to simulate materials, lighting and a camera. It is used to study design options, communicate with clients and authorities, and market projects before they are built.

What is the difference between real-time and offline rendering?

Real-time rendering draws the scene many times per second on the graphics card, so users can move through a model and see changes immediately. Offline rendering calculates each image in greater detail over seconds to hours, giving higher quality and more control for final images.

What software is used for architectural rendering?

Common tools include real-time renderers such as Enscape and Lumion, production renderers such as V-Ray and Corona, and 3D packages such as 3ds Max that host them. Many design programs, including Revit, ArchiCAD, SketchUp and Rhino, can also render directly or through plug-ins.

Are architectural renders accurate?

They are accurate to the model and settings used, but they represent design intent rather than verified construction. Structure, services, tolerances and real lighting conditions may differ, so renders should be read alongside drawings and specifications.

Can a render show that a structure is safe?

No. A render shows appearance only. Whether a structure is safe depends on engineering analysis and design to the relevant codes, carried out and endorsed by a qualified engineer.

Want the structural ideas in your renders checked before they become promises? Talk to our consultancy or get in touch.