Interactive websites

WebGL, rendered frames or video for a website?

Choose the production format from the control a visitor needs. This comparison covers viewpoint, editing, accessibility, fallback and the work behind each option.

Read in Spanish

Choose the format for the task

A short working slice can settle the choice better than a list of effects. Compare the visitor's control, fallback and editing needs with a real project before quoting production.

01

Start with control

Video gives the maker control over every frame and gives the visitor play, pause and seek. A rendered image sequence can tie a fixed viewpoint to scroll, but changing the camera later means rendering frames again. Real-time WebGL allows the visitor to rotate, inspect or navigate a scene; that freedom introduces more input, camera and device states to design.

None of these formats makes the surrounding page useful by itself. The offer, labels, navigation and action still need readable content. If the interaction is only a decorative background, first ask whether a still or short video communicates enough. Use real-time graphics when a person's action changes what they can learn or decide.

Start with control
FormatVisitor controlProduction question
VideoPlay, pause and seekWho films or renders and captions it?
Rendered framesUsually a fixed path tied to scroll or timeHow many frames and responsive crops are needed?
Real-time WebGLViewpoint, object or world can respondWhich controls and graphics fallbacks are required?
02

Budget for changes and fallbacks

A fixed film concentrates effort in production and editing. A frame sequence adds delivery weight and responsive timing. A live scene adds asset optimization, graphics engineering, input handling and testing across target devices. The cheapest option depends on what will change after launch: replacing a product color or adding a chapter can have very different costs in each format.

Ask for a working non-animated path, captions or equivalent text where needed, keyboard access to controls and a reduced-motion behavior. Name what happens if media fails to load or the graphics context disappears. Then compare quotes with the same deliverables and maintenance assumptions instead of comparing a 3D label with a video label.

03

Use one representative scene as the decision test

Provide the same content and asset to each proposed format. Ask what a visitor must do, how quickly the essential explanation appears, whether the action remains reachable, and how an editor changes the content. Test the result on an agreed phone before committing to the full production.

Janus Observatory shows why formats can coexist: an atmospheric opening establishes a world, while the atlas and observatory provide orientation and exploration. That division is a design decision, not a prescription that every project needs a real-time scene. The appropriate build is the one that preserves the intended understanding across the supported devices and fallback path.

Sources & editorial notes

Sources

How we built this guide

EvidenceThis guide combines 2 linked sources, Dardo’s analysis and clearly labeled project examples.

LimitsVisible design and content decisions are not presented as measured ranking, traffic or conversion gains.

Read our editorial policy