What Fabric Manufacturers Want From 3D — And Why That Expectation Quietly Breaks

When we talk to fabric manufacturers about 3D, the request is remarkably consistent.

"We want to supply fabric files that can be uploaded into a specifier’s system — usually CAD or BIM tools — so those fabrics can be applied to many different furniture objects and produce results that are close to photorealistic."

It’s an entirely reasonable goal.

Specifiers want flexibility. Manufacturers want reach. Everyone wants a workflow that feels scalable and familiar. On the surface, this looks like a straightforward asset-distribution problem.

In practice, it isn’t.

The difficulty isn’t that this ambition is naïve. The difficulty is that fabric perception depends on variables a standalone file simply cannot carry with it.

Fabric Assets Don’t Travel Alone — Context Does

For a fabric to appear believable in 3D, a number of conditions must align at the same time:

• the geometry it’s applied to

• the lighting model of the scene

• the scale of the surface

• the way the pattern repeats

• the camera distance and angle

• the rendering system interpreting the material

• the device it’s viewed on

When a fabric file is dropped into a specifier’s CAD tool, none of those conditions are known in advance.

This is not an edge case — it is the default operating environment.

“Applies to Multiple Objects” Is the First Warning Sign

Fabric does not behave consistently across different forms.

A flat panel, a wrapped cushion, a curved shell, and a soft edge all change how a fabric reads. Scale alone can radically alter perception: a pattern that feels balanced on a chair may feel overwhelming on a sofa or disappear entirely on a large surface.

Expecting a single material definition to perform well across arbitrary geometry is not optimisation — it is compromise. At best, the fabric ends up tuned to nothing in particular. At worst, it looks wrong in different ways everywhere.

CAD Tools Aren’t Neutral Containers

Most CAD and BIM tools were not designed to author fabric perception. They prioritise geometry, documentation, coordination, and performance.

Material systems vary widely between platforms. Lighting models differ. Shader capabilities differ. Texture handling differs. Defaults differ. Optimisation rules differ.

So even before a user makes a change, the same fabric file behaves differently by design.

Photorealism Depends on Scene, Not Assets

Photorealism doesn’t live inside a fabric file.

It emerges from the relationship between lighting, geometry, scale, and camera placement. Those elements exist at the scene level — not the asset level.

The promise of “close to photorealistic” therefore relies on an assumption that someone else’s scene will accidentally align with the conditions under which the fabric was authored.

That’s not a strategy. It’s hope.

The Control–Responsibility Mismatch

There is a quiet asymmetry at play.

The moment a fabric file leaves the manufacturer’s system, control is gone. Users tweak materials. Software optimises them. Defaults override intent. Performance constraints flatten nuance.

And yet when something looks wrong, the manufacturer is still often blamed.

Responsibility lingers even after authorship has ended.

Why This Keeps Happening

The motivation behind this approach is understandable. It avoids building new systems. It fits existing workflows. It feels efficient.

But what it actually does is export the representation problem rather than solve it.

What manufacturers are really being asked to provide is portable realism without portable context — and that doesn’t exist.

The Unavoidable Conclusion

This doesn’t mean manufacturers shouldn’t support CAD workflows or supply assets. Those things still have value.

It does mean that if the goal is faithful representation, unrestricted reuse and photorealism are structurally incompatible.

Once that’s acknowledged, the need for a controlled environment stops looking like a limitation and starts looking like an honest boundary.

And that is why, as discussed in our previous post, retaining control over how fabrics are shown isn’t a preference — it’s a responsibility.