XO3D Studio · Thinking
Material Obsession.
If the material doesn't behave the way the real object would, the viewer knows — even if they can't say why. That's the thing we're always solving.
There is a particular look that commodity product CGI has. You've seen it in thousands of product listings, in most corporate product videos, in the overwhelming majority of CGI produced for e-commerce. The surface is metallic but not quite right. The specular highlight is in roughly the correct position but moves wrongly as the camera passes. The plastic reads as plastic but not as that specific grade of polycarbonate with that specific surface treatment. Everything is approximately correct and nothing is actually right.
The viewer doesn't consciously identify any of this. They feel it. The product seems slightly less real than it should. The material seems slightly less trustworthy. The overall effect is a reduction in the viewer's willingness to believe the product's quality promise — which is the exact opposite of what the CGI was commissioned to produce.
I've spent my career trying to understand why this happens and how to not let it happen in my work.
The library lie
Why preset materials produce this problem every time.
Most CGI software comes with material libraries. "Brushed aluminium." "Matte plastic." "Gloss paint." These presets describe a category of material, not a specific material. They are averages — statistical midpoints between the range of real materials they nominally represent. They are useful for rapid prototyping and terrible for final production.
The brushed aluminium preset in any software's material library was created by someone studying some brushed aluminium at some point. It is a reasonable approximation of a reasonable example of that material. But your product does not use a reasonable example of brushed aluminium — it uses a specific aluminium alloy with a specific brushing process at a specific depth producing a specific anisotropic reflection pattern. That specific combination of properties is not in any library.
When a CGI studio applies a library material to your product, the viewer's eye — informed by a lifetime of looking at real surfaces — registers the mismatch. Not consciously. At the level of instinct. The surface doesn't quite behave the way their mental model says it should. The product seems less real. Less trustworthy. Less worth wanting.
What reconstruction actually involves
The process behind a material that reads as real.
When I start a new project, before I open any software, I study the physical object or the specifications that describe it. I look at how light behaves on the surface at different angles. Where does the reflection become tight and specular? Where does it spread into a broader, diffuse glow? Is there anisotropy in the brushing — does the reflection change character as the light source moves perpendicular to the brush marks versus parallel to them? What is the base colour of the material in the absence of reflections? What happens at grazing angles?
These are not questions a library material can answer correctly. They are questions I answer by looking at the specific material and encoding my observations into physically accurate values in the render engine. The roughness value is not a guess from a preset — it is derived from observation of how sharp or diffuse the specular reflection is on this specific surface. The metallic value is not a binary toggle — it is calibrated to the specific balance of specular and diffuse reflectance this particular finish exhibits.
This process takes longer than applying a library preset. It takes significantly longer. And it is the entire reason the result looks like a photograph of the real product rather than a CGI approximation of a similar product.
The most difficult materials
Where the problem is hardest to solve.
Liquid
Beer, spirits, water, oil
Liquid requires volumetric simulation — the light doesn't just reflect off the surface, it enters the liquid, scatters within it, and exits at a different point. The colour of a spirit in a glass depends on liquid depth, the specific absorption properties of the liquid, and the refraction index of the glass container. Every variable interacts. There is no shortcut.
Translucent plastics
Diffusers, covers, casings
Semi-transparent polymers — the frosted diffuser on a smart home device, the translucent casing of a medical device, the smoked lens on a consumer electronic — require subsurface scattering alongside correct transparency and refraction. The interaction between internal light scatter and surface reflection is where most CGI fails with these materials.
Natural materials
Wood, cane, leather, stone
Organic materials resist perfect reconstruction because they are never perfectly consistent. Wood grain has pattern but not repetition. Leather has texture but not uniformity. The challenge is capturing the character of the material — the quality that makes it read as natural — without producing either a sterile digital pattern or an overcrowded surface.
Coated metals
Anodised, PVD, powder coat
Metal finishing processes — anodising, physical vapour deposition, powder coating — dramatically change the optical properties of the base metal. The same aluminium substrate reads completely differently in natural anodise versus black anodise versus colour-anodised versions. Each requires separate material reconstruction; a palette swap does not produce correct results.
Why it matters commercially
Material accuracy is a brand-trust mechanism.
The viewer's response to material quality in product imagery is visceral and pre-conscious. They do not decide whether the material looks right — they feel it before they can think. That feeling becomes a judgement about the product's quality, authenticity, and value. Material accuracy is not a technical nicety. It is the mechanism by which product CGI either builds or erodes brand trust.
This is why I don't use library materials. Not because I can't. Because they produce work that I know will fail at the level that matters most — the viewer's instinctive assessment of whether this product is worth wanting. That failure is too important to risk for the time a library preset saves.
Questions
Common questions about material reconstruction.
How long does material reconstruction take on a typical project?
Do you ever use library materials as a starting point?
Start the conversation
Want to see what correct material reconstruction looks like on your product?
Send us your product specs and a reference image. We'll show you what the difference looks like.