Digitise your product once, reuse it in every sales channel.

We work from STEP, IGES, Parasolid, SolidWorks, Inventor, CATIA, Rhino, Revit and IFC files. Where no file exists, we model from measurements and photographs of the physical product. Incoming CAD data is never used directly for marketing: simplification, surface repair, material assignment and web optimisation are a separate production step. Which file is sufficient for which output is decided together during scoping.

If sensor data, live operation, predictive maintenance or production simulation is required, the project becomes a technology integration. The MORFAXIS standard service is a marketing- and sales-oriented 3D product twin; the boundary is separated explicitly before any proposal.

We set this up as a reusable product asset rather than a single render job, and the difference shows in the cost line. The first production is more expensive than one image because the master model, material system and variant matrix are built. In exchange, the second, third and thirtieth output drops to derivation cost. A new colour, a new module or a new market no longer restarts the shoot–redesign–retranslate loop.

Every new product, every new variant and every trade fair season restarts that loop in a company without the system. The cost looks one-off but repeats on every cycle.

Scenario 1 — A machine that cannot be photographed

The product is assembled on site, cannot be moved to a studio, and only exists fully built inside a customer's facility. Catalogue images were shot in different years under different light; placed side by side they do not read as one product family. Here the 3D master asset establishes a consistent visual language in one pass: same lighting, same angle set, same material definition. Adding a new model later does not break the set.

Scenario 2 — Variants that photography cannot scale to

Four body sizes, six colours and three module options is seventy-two combinations. Physical photography cannot cover that, so in practice the company shoots a handful and leaves the rest to the buyer's imagination. With a variant matrix, combinations are derived from the same master model and e-commerce, catalogue and the dealer portal all feed from one source.

Scenario 3 — A product nobody understands until it is explained

In a closed-body system the real value is inside: the flow path, the filtration stage, the sealing solution. A buyer looking from outside cannot see the difference between two competing products, and the decision collapses to price. A cutaway and a working-principle animation make that difference visible and move the conversation from price to technical merit.

The most common failure of interactive 3D experiences is performance, not design. The rules we treat as binding on the web side are:

• No 3D scene autostarts. A poster image and an explicit “start” button are used; neither the model nor the WebGL context downloads before the click.
• The compressed model budget is under 3 MB per scene, with geometry compressed via Draco or Meshopt and textures in KTX2/Basis.
• Polygon budget for a web scene stays under 250k triangles. A CAD model is never used directly; simplification is its own production step.
• Loading shows percentage progress, an estimated size and the ability to cancel.
• On low-memory devices and under prefers-reduced-motion, an image sequence is served instead of 3D; without WebGL support a 360-degree sequence or an explained poster is shown — never an empty box.
• Every 3D and image slot reserves space with width/height or aspect-ratio; the cumulative layout shift target is zero.

These are commercial requirements rather than aesthetic preferences: a configurator that will not open on a phone costs more than one that was never built.

Four inputs make the first call productive: the format and version of the source file, the variant logic across the product family, the target channels, and any calendar pressure (a fair, a launch, a season). If no file exists that is not a blocker; the measurement- and photograph-based modelling route is assessed separately.

By the end of the call you get, in writing, which output is possible with which input, which step requires additional production, and at what point the scope crosses into operational digital twin territory.