Modelight / 3D Animation Studio / 3D Glossary
Modelight Studio — Knowledge Base

3D Glossary

From 3D graphics and rendering to ray tracing, PBR, and 3D cross-sections — 28 terms explained for B2B marketing and engineering teams. No unnecessary jargon.

Author: Jakub, Modelight Studio
Updated: May 2026
Reading time: ~18 min
28Terms
5Sections
9Fun facts

It was 1995. Pixar released Toy Story — the first feature-length 3D animated film in history. Rendering one minute of footage took 90 machine-hours. Today the same task takes seconds on a laptop. That's how far 3D graphics has come — and how thoroughly it has become a tool for everyone, not just Hollywood. The industry kept its jargon, though. This glossary is for anyone who wants to talk about 3D content production without feeling like they need a translator.

See also: Technical animations for B2B →
Section 01
Basics
What exactly is 3D graphics?
1.1 3D Graphics Beginner

3D graphics is the creation of objects and scenes in a virtual three-dimensional space with width, height, and depth.

Unlike flat 2D graphics (leaflets, posters, logos), 3D objects exist in space and can be viewed from any angle — like real physical objects.

For product marketing, this is a fundamental shift: instead of a photograph taken from one fixed angle, clients get an object that can be rotated, zoomed in, exploded into parts — showing precisely the technical advantage that makes the product worth buying. For a marketing team, this means: full control over how the product is presented — without a photoshoot, without a physical prototype, without the limitations of a camera.

Technical visualizations — Modelight
Fun fact

The first computer-generated 3D graphics appeared in a feature film in 1973 — a brief two-minute sequence in Westworld. It cost a fortune and took months to produce. Today, nobody would recognize it as a special effect.

1.2 3D Visualization Beginner

3D visualization is the use of three-dimensional graphics to show something that doesn't yet exist physically — or exists but is too complex or internal to photograph.

A product still in design. A machine mounted inside a housing that nobody can see. An HVAC system hidden inside a building wall. 3D visualization lets you show all of it before a prototype exists — or before anyone sets foot on a construction site.

For B2B companies, this isn't an aesthetic decision. It's a commercial one: shorter sales cycles and fewer late-stage engineering change requests.

Why this matters for B2B

3D visualization lets you show a product to a client at the proposal stage — before it's manufactured. Companies that have integrated product visualization into their sales process report 20–35% shorter decision cycles.

See at Modelight

Technical visualizations  •  Product visualizations

1.3 3D Animation Beginner

3D animation is a sequence of rendered three-dimensional frames played back at speed (typically 24–30 frames per second), creating the illusion of motion.

Think of it as a 3D visualization that moves. A standard industrial product animation runs 30–90 seconds. In that time, a client can watch how the product operates, how it's assembled, and what happens inside during real-world use — things no photograph can show.

Technical animations — examples
Fun fact

Animated films run at 24 frames per second — the same as classic cinema. Video games run at 60+ fps because interactivity demands more fluidity. Industrial 3D animation typically targets 25 fps — the PAL television standard, sufficient for the eye and practical for render farms.

1.4 Motion Graphics Beginner

Motion graphics is animation dominated by abstract elements — text, shapes, infographics, lines — rather than photorealistic objects.

It's a tool for communicating data and concepts, not simulating reality. When a company wants to show how its logistics network operates, how market share has grown, or how a process flows — motion graphics does it faster and more clearly than any static chart. Works well layered on top of 3D renders for explainer videos.

1.5 2D vs 3D Graphics — Key Differences Beginner

Both formats serve visual communication, but they differ fundamentally in capability and use case:

Dimension2D Graphics3D Graphics
SpaceFlat X/Y plane3D space X/Y/Z
Camera angleOne fixed viewAny angle, full 360°
Production costLowerHigher, but scalable
Production timeShorterLonger, multi-stage
Best forLeaflets, posters, simple animationProducts, architecture, film, VR/AR
InteractivityNoneConfigurators, VR, AR

Want to see how this works in practice? Browse our industrial 3D animation and visualization projects.

See projects →
01 Modeling → 02 Texturing → 03 Lighting → 04 Rendering → 05 Post-production
The 3D production pipeline — from model to finished asset
Section 02
Creation
How is a 3D model and scene built?
2.13D ModelingIntermediate

3D modeling is the process of creating three-dimensional objects in software by manipulating a mesh of vertices, edges, and polygons.

A 3D artist sculpts a digital object the way an architect builds a structure. There are several distinct approaches:

  • Polygon modeling — building the object from small flat faces, like assembling a mosaic.
  • Sculpting — the artist works as with clay, freely shaping organic details.
  • CAD/NURBS modeling — used in engineering, based on precise mathematical curves rather than polygons.
Modelight 3D Animation Studio — overview
Fun fact

A single photorealistic human head model can contain up to 50 million polygons. A car model typically runs 300,000–2 million. A full factory scene with machinery can reach 10 million. Every polygon is calculated during rendering.

POLYGON COUNT BY OBJECT TYPE5KCNC Bolt50KCNC Cutter300KMed. implant2MCar model10MFactory scene50MFilm character
Typical polygon counts across 3D applications
2.2TexturingIntermediate

Texturing is the process of applying image maps to the surface of a 3D model to give it the appearance of a specific material — steel, plastic, rubber, wood, glass.

The model itself is like a mannequin's skeleton. Textures are the clothing. In practice it's never just one map, but several working together: a color map (albedo), a normal map, and a roughness map. Combined, they create the visual impression of a real surface.

2.3UV MappingIntermediate

UV mapping is the process of unfolding a 3D model's surface into a flat 2D layout so textures can be applied without distortion.

Imagine trying to wrap a globe with a flat world map. That's exactly the problem — and exactly what UV mapping solves. Poor UV mapping means stretched or misaligned textures. It's one of those things clients rarely notice when done right — and immediately notice when done wrong.

2.4Rigging and SkinningIntermediate

Rigging is giving a 3D model a skeleton — a set of bones and joints that can be animated. Skinning links the model's surface to that skeleton.

Without rigging, every part of a model would need to be animated independently. With it, an animator moves one bone and the rest responds like a real mechanism. At Modelight, we rig primarily industrial mechanisms — valves, actuators, robot arms.

Fun fact

Thanos in Avengers: Infinity War had over 1,000 facial muscle control points. The principle is identical to our industrial rigging: the better the rig, the more believable the motion.

2.53D SceneBeginner

A 3D scene is the virtual environment in which models, light sources, and the camera are placed — everything needed to generate an image.

Building a scene isn't just about positioning objects. It's a set of decisions about lighting, composition, and framing — what's in shot and what isn't. A good 3D artist thinks like a photographer and a film director simultaneously.

2.6Key 3D Software

Software choice depends on the industry, goal, and budget:

Modelight workflow overview
SoftwarePrimary useKey strength
BlenderModeling, animation, renderingOpen source, highly versatile
MayaRigging, animation, VFXIndustry standard for film
3ds MaxArchitectural visualizationArchitecture standard
Cinema 4DMotion designIntuitive workflow
SolidWorksEngineering CADPrecision, CAM integration
HoudiniVFX simulationsDestruction, fluids, particles

We work with Blender, SolidWorks, and Cinema 4D. Want to know which is the right fit for your project?

Ask us →
Section 03
Rendering
How does a computer turn a model into an image?
3.1RenderingIntermediate

Rendering is the computational process in which a computer calculates the geometry, materials, lights, shadows, and reflections of a 3D scene and produces a final 2D image or a sequence of animation frames.

This is the moment when the virtual world becomes an image. Rendering a single frame can take anywhere from a fraction of a second to several hours — depending on scene complexity and available computing power.

Fun fact

Toy Story (1995) was rendered on a farm of 117 Sun Microsystems workstations. Today, an equivalent GPU setup cuts rendering from overnight to under an hour.

3.2Render FarmIntermediate

A render farm is a network of computers or cloud resources working in parallel to render a 3D scene.

Instead of waiting for a single machine, many render nodes each process a few frames simultaneously. Delivery time becomes a question of scalable infrastructure, not a single computer's capacity. For clients, this means predictable timelines regardless of animation length.

3.3Render EnginesIntermediate

A render engine is the software that processes a 3D scene into a final image. The choice of engine affects visual quality, rendering time, and interactive capabilities.

EngineCharacteristicsBest for
V-RayPhotorealism, precise PBRArchitecture, products
CoronaNatural light, ease of useInteriors, products
ArnoldHollywood standardFilm, VFX
Unreal EngineReal-time, 60+ FPSConfigurators, AR/VR
CyclesOpen source, path tracingIndependent, education
KeyShotSpeed, simplicityProduct visualization
3.4Ray TracingAdvanced

Ray tracing is a rendering technique that simulates the physical behavior of light — tracing the path of each light ray from its source, through reflections and refractions, all the way to the camera lens.

The result: reflections in glass surfaces show what's actually surrounding the object. Shadows have soft, natural edges. Glass materials refract light exactly as real glass would. This is the foundation of photorealism.

The trade-off: it's computationally expensive. Nvidia RTX cards (introduced in 2018) brought real-time ray tracing to interactive applications like product configurators.

RAY TRACING vs RASTERIZATIONRay TracingSimulates physical light raysTrue reflections and shadowsSlower, but photorealisticFilm, product visualizationRasterizationProjects triangles onto pixelsApproximated light and shadowsFast, interactive (60+ FPS)Games, configurators, AR/VR
Two rendering philosophies — quality vs. speed
Fun fact

The concept of ray tracing was described by Arthur Appel in 1968 for optical simulation purposes. The technology reached consumer video games 50 years later — with Nvidia RTX cards in 2018.

3.5HDRI — Image-based LightingIntermediate

HDRI (High Dynamic Range Image) is a spherical panoramic photograph with a wide brightness range, used as a light source in 3D scenes.

Instead of manually placing dozens of lights, the artist wraps the entire scene in a single HDRI image — and instantly gets realistic lighting that matches a specific environment, whether a factory floor or a conference room.

Why this matters for B2B

HDRI lets you show your product in exactly the environment where it will be used. The client sees their product in context — not on an abstract neutral background.

3.6Post-production and CompositingIntermediate

Post-production is the stage after rendering — color grading, visual effects, editing, sound design — everything that transforms raw rendered frames into a finished deliverable.

Even a perfectly rendered scene goes through post-production. Common tools: Adobe After Effects, DaVinci Resolve, Nuke. This stage often accounts for 20–30% of total production time on a complex animation.

Want to see the results of our rendering and post-production work? Explore the portfolio.

See projects →
Section 04
Advanced
For those who want to go deeper.
4.1PBR — Physically Based RenderingAdvanced

PBR is an approach to material and lighting creation based on the laws of optics — not on artistic approximation.

The core principle: energy conservation. A surface cannot reflect more light than it receives. PBR uses physically measurable parameters: metalness, roughness, albedo — values that can be taken from a lab measurement of the actual material.

Why this matters for B2B

A PBR visualization shows exactly how a product will look under real-world lighting conditions — on a factory floor, at a trade show, or in a client's facility — not under idealized studio conditions.

4.2Normal MapAdvanced

A normal map is a special texture that simulates surface detail without increasing the polygon count of the model.

Welds, chamfers, scratches, embossed text on housings — all achieved through a normal map, without extending rendering time. The visual result is nearly indistinguishable from actual geometry, at a fraction of the computational cost.

4.3LOD — Level of DetailAdvanced

LOD is a technique in which the same 3D object exists in several versions of different complexity, switching automatically based on distance from the camera.

An object close to the camera renders in full detail. In the background, a simplified version weighing 10× less is used instead. Standard practice in games and 3D product configurators.

Fun fact

In Unreal Engine 5, traditional LOD has been replaced by the Nanite technology — billions of polygons in real time, no manual LOD required. It powers the car configurators of BMW, Porsche, and Audi.

4.4PhotogrammetryAdvanced

Photogrammetry is the process of scanning real-world objects using many photographs taken from different angles, from which software reconstructs a 3D model.

Instead of modeling by hand — you photograph the object from 100+ angles and the algorithm builds a model with photorealistic textures already baked in. The results are difficult to match with traditional modeling for organic or highly detailed subjects.

4.5Real-time RenderingAdvanced

Real-time rendering is the generation of 3D images fast enough (minimum 30 FPS) for the user to interact with the model in real time.

It's the foundation of product configurators: a client clicks "change color to red" and sees the result instantly — no waiting for a render job. The same technology powers VR training simulations and AR product overlays.

Fun fact

Unreal Engine — best known from video games — is now the industry standard for automotive configurators at BMW, Porsche, and Audi. Showroom visitors see their customized car in real time, in any color or trim.

4.6NeRF — Neural Radiance FieldsAdvanced

NeRF is a machine learning technique that trains a neural network from a set of photographs to generate new views of a scene from any angle — without a traditional 3D model.

Instead of building a polygon mesh, NeRF learns "what the space looks like." Published in 2020 by UC Berkeley, it has found applications in e-commerce product photography and building documentation. Not yet production-standard, but closing in fast.

4.7Gaussian SplattingAdvanced

Gaussian Splatting (2023) is a 3D visualization technique that represents objects as a collection of millions of Gaussian point clouds instead of polygons.

Exceptional detail with generation times shorter than NeRF. Enables photorealistic scene creation from a smartphone video. In 2–3 years it will likely become an industry standard for certain product categories — watch this space.

Section 05
B2B Applications
How do industrial companies use 3D content?
5.13D Cross-sectionBeginner

A 3D cross-section is a visualization technique that virtually cuts through a 3D model to reveal its internal structure — mechanisms, material layers, hidden components.

This is exactly what Modelight has been doing since 2008: showing what the camera can't see. The interior of a CNC machining center. The layers of a medical implant. Airflow inside an HVAC system. A photograph can't show any of it. A cross-section animation can.

3D cross-sections are consistently among the highest-traffic terms bringing visitors to Modelight — because very few studios do them well at the technical B2B level.

Technical animations with cross-sections — Modelight
Why this matters for B2B

A cross-section eliminates the "I can't see what I'm paying for" problem. The client understands the product's internal mechanics, grasps its technical value, and makes a decision faster.

5.23D Product ConfiguratorIntermediate

A 3D product configurator is an interactive browser-based tool that lets users change product parameters — color, material, dimensions — and instantly see the result in a photorealistic 3D view.

Powered by real-time rendering. Instead of a catalogue with 200 variants, a single configurator shows all of them on demand. Companies using 3D configurators report 20–40% shorter sales cycles and higher average order values — because the customer can see exactly what they're ordering.

5.3Digital TwinAdvanced

A digital twin is a virtual replica of a physical object, process, or system, updated in real time with data from sensors on the physical counterpart.

Imagine a copy of your production line on a screen — reacting to the same data as the real line. You can test changes, predict failures, and train operators without stopping production. Digital twins combine 3D modeling, IoT, analytics, and visualization. This is the direction of Industry 4.0.

Fun fact

The term "digital twin" was popularized by Michael Grieves in 2002, but the concept is older — NASA used digital replicas of spacecraft during Apollo missions in the 1960s.

5.43D Render vs Photography — When to Use WhichBeginner

3D rendering and professional photography are two different tools — each with its own applications, costs, and limitations.

The typical client question: "Wouldn't it be cheaper to just take a photo?" The answer depends on the situation:

Scenario3D RenderPhotography
Product doesn't exist yetOnly optionImpossible
Show the interior3D cross-sectionRequires cutting a prototype
Multiple variantsOne scene, many rendersEach variant separately
Small existing productPossible, but slowerFaster and cheaper
Context (factory, lab)Any environment on demandRequires location shoot
ScalabilityModel once, render 1,000×Each session separately

Not sure which format is right for your case? We'll help you decide.

Let's talk →

Got a product to show?

A conversation and brief are always free — no commitment, just straight answers.

Write to us
FAQ
Common questions
Answers optimized for Google and AI Overview

Photorealistic 3D visualization is a computer-generated image indistinguishable from a photograph. It is achieved through PBR materials, ray tracing, realistic lighting, and post-production. For product marketing, it means a complete visual presentation before the physical product exists.

It depends on scene complexity and computing power. A straightforward 60-second product animation on a render farm takes 12–48 hours. Complex scenes with detailed interiors or large environments can triple that.

3D visualization produces static images or still renders. 3D animation is a moving video. Visualization is faster and cheaper; animation gives you more — it shows movement, assembly, and the operation of internal mechanisms.

Yes, but they need preparation — materials, lighting, and polygon optimization. Preparing a CAD model for photorealistic rendering takes anywhere from a few hours to several days, depending on complexity and the number of assemblies.

A 3D product configurator is a browser-based tool that lets users change product parameters and instantly see the result in a photorealistic 3D view. It runs on real-time rendering. Companies using configurators report 20–40% shorter sales cycles.

A 3D cross-section virtually cuts through a model to reveal its internal structure — mechanisms, material layers, hidden components. It's one of Modelight Studio's core specializations: showing what a camera never could.

In art, "to render" means to depict or represent. In computer graphics, the term was popularized by Loren Carpenter at Lucasfilm in the early 1980s to describe the computational generation of images from 3D scene data.

→
Sources and links
For further reading and verification

Related 3D guides