Short answer
In a translucent material light does not bounce off the surface: it enters, scatters many times inside and leaves some distance from where it went in. That spreads light sideways, softens shadows and edges, and lets light pick up colour from deeper in the material — or from whatever is behind it. How strong the glow is depends on how far light travels before being scattered or absorbed, the thickness of the piece and what backs it.
An opaque matt surface returns light from essentially the point where it arrived. A translucent one does not: photons wander under the surface and emerge millimetres or centimetres away. Jensen and colleagues' 2001 model of subsurface transport, now standard in computer graphics, treats this as diffusion and was built to render materials such as marble, milk and skin that looked wrong when drawn as surface reflectors. The visible consequences are consistent across materials: shadow edges are soft and slightly coloured, fine surface detail is blurred, thin edges and corners glow, and a point of light on the surface spreads into a halo. That is the 'depth' people describe in marble, alabaster, jade, candle wax and bone china.
Because light that travels further inside passes through more material before it comes back, it is filtered more. In a lightly coloured translucent material the light that has gone deep is warmer or more saturated than light scattered near the surface, so edges and backlit areas show a stronger colour than the face. Marble veins seen under the surface look softer than veins at it; the pink of a thin porcelain cup held to a lamp is not visible face-on. Backlighting shows the effect most: onyx and alabaster panels, lampshades of paper or parchment, and thin marble cladding lit from behind show transmitted colour that is often much warmer than the reflected colour, because long paths filter out more blue.
Any light that reaches the back of a translucent layer is partly reflected by whatever is there. A thin marble tile on white adhesive looks paler than the same tile on grey; a translucent glaze on a dark clay body is darker than on porcelain; paper looks greyer over a dark desk; and a sheer curtain changes with what is behind it. This is the same physics as hiding power in paint, covered on the substrate page, and it is why translucent materials must be approved on their actual backing. Adhesive colour for thin stone, mosaic and glass is a real specification item, not a detail.
Spectrophotometers assume light returns close to where it enters. With a translucent sample, some light spreads beyond the measurement aperture and is lost, so smaller apertures read darker than larger ones — an error known as edge loss or lateral diffusion. The backing under the sample changes the reading too. For comparable numbers, measure translucent materials at a consistent thickness, over a stated backing (often white, sometimes black as well, so the difference can be reported), with the same aperture size each time, and record transmittance separately if light passes right through. Visual comparison should be at the viewing distance and lighting of the installation, because the glow is often what people are approving.
| Material | Translucency ordinal | What to watch when measuring |
|---|---|---|
| Glass | 4 | Measure glass in transmittance as well as reflectance, state thickness, and view tints edge-on and in the installed thickness. |
| Ceramics, glazes, porcelain and vitreous enamel | 2 | Every firing varies; approve against fired samples from the production kiln, and measure glossy glazes with specular excluded as well as included. |
| Marble | 2 | Translucency lets the backing and the instrument aperture affect readings; measure on a consistent white backing and report aperture size. |
| Paper and card | 2 | Paper is translucent: measure over a stated backing (white or a stack of the same sheet) and state whether the instrument's UV is included. |
| Plastics | 2 | Texture, gloss and wall thickness change measured colour; compare parts of the same texture and thickness, with specular excluded as well as included. |
| Cotton and linen (cellulose fibres) | 1 | Pile, weave and yarn twist change readings; fold samples to opacity, measure several orientations, and compare under more than one illuminant. |
| Natural stone (granite, limestone, sandstone, slate) | 1 | Stone is inherently variable; specify a range from representative slabs, and measure honed and polished samples separately because gloss changes the reading. |
| Plaster (lime and gypsum) | 1 | Surface texture and residual moisture both change readings; measure only fully dry plaster, averaging several spots, and compare repairs after weathering. |
| Polyester and other synthetic fibres | 1 | Fluorescent brighteners and lustre make readings depend on instrument UV content and geometry; state both when comparing. |
| Wood (solid timber and veneer) | 1 | Grain direction, figure and cut change readings; measure with the grain aligned consistently, average several positions, and remeasure after light exposure because fresh wood moves fastest. |
| Wool and silk (protein fibres) | 1 | Silk's directional sheen makes readings depend on warp orientation; record it, and measure lustrous fabrics with specular excluded as well as included. |
| Aluminium (anodised, powder-coated, mill) | 0 | Anodised colour depends on alloy and batch; approve against physical range samples and measure at more than one angle. |
| Concrete | 0 | Pours vary with mix water, curing and formwork; measure several areas of dry, cured concrete and treat a single reading as unrepresentative. |
| Copper, brass and bronze | 0 | Colour changes within hours of polishing and differs with finish direction; measure at a stated time after finishing, at more than one angle. |
| Fired clay brick | 0 | Colour varies within a pack and within a single brick; blend packs on site and assess panels of many bricks, not a single sample. |
| Laminates and engineered stone | 0 | Printed patterns vary across a sheet; measure a defined background area and compare gloss separately. |
| Leather | 0 | Natural grain, nap direction and uneven dye uptake make single readings unreliable; average several areas and specify the finish type. |
| Steel (stainless, weathering, galvanised) | 0 | Brushed metal is directional: readings change with sample rotation. Record orientation, use multi-angle or both SCI and SCE, and judge highlights visually. |
Why: Uneven or coloured adhesive shows through the translucent stone.
Fix: Use a white adhesive with full, even coverage; test a tile on the chosen adhesive first.
Why: Edge loss from a smaller aperture or a different backing.
Fix: Measure both with the same aperture, thickness and backing.
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
FactStrong evidence
Subsurface scattering — light entering a material, scattering internally and leaving at a different point — explains the soft appearance of materials such as marble, skin and milk, which surface-only reflection models render as hard and flat.
Source: A practical model for subsurface light transport (SIGGRAPH 2001)
FactModerate evidence
Calcite's relatively low refractive index lets light penetrate some distance into marble before it scatters back out, which gives marble its characteristic waxy look.
Caveat: Penetration depth varies with grain size, purity and finish; published figures are indicative.
Source: Marble
FactModerate evidence
Porcelain becomes translucent when it is fired close enough to melting to become glass-like, and iron and titanium impurities reduce that translucency.
Reviewed 29 September 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.