How to Specify a PVC Colour: From a Name to a Number That Can Be Checked

A colour name carries a range, not a value. What the range contains is decided by whoever is interpreting it.

An order goes out for white PVC foam board. The container arrives and the buyer says it is not white. It is slightly cream, or slightly grey, or noticeably cooler than the last batch. The supplier says it is white, and it is, in the sense that the pigment is the same one used on every white board they make.

Both statements are accurate. White is a name for a broad region of colour space, and the region contains everything from a warm cream to a cool blue-white. When a buyer and a supplier both say white they may be describing two different points inside that region, and neither has done anything unreasonable.

Colour is the specification subject where the gap between how a buyer communicates and how a factory produces is widest. This article sets out the ways a colour can be specified, ranked by how much they leave open to interpretation, and what has to be written down before a colour can be checked at all. The production side of the same subject, why consecutive batches differ, is covered in our piece on PVC colour and batch consistency.

The names that appear on purchase orders cover territories, not points.

Names travel well between people and badly between documents. What a name describes is a region, and a factory has to make something specific.

There is a second problem underneath the first. Colour is a perception produced by light, a surface and an observer, not a fixed property of an object. The same board genuinely does look different against a different background, under a different light and to a different pair of eyes. That is not a defect in anyone's judgement, and it is the reason a colour specification has to describe the conditions as well as the colour.

There are four practical ways to communicate a colour, and they differ enormously in how much they leave to the reader.

The practical combination for a serious order is the top two together. An approved physical sample establishes what the buyer wants, and a measurement of that sample converts it into numbers that survive the sample. This turns a reference with a shelf life into one that can be reproduced years later, and it is the single most useful habit in this subject.

Where the colour is being chosen rather than matched, working from a system reference is fast and adequate. Where it is being matched against an existing surface, a tile, a window frame or another batch of the same product, the measured route is the only one that holds up.

The three systems that appear on purchase orders were each built for a different purpose, and none of them was built as a tolerance.

A code names a target. It says nothing about how close to the target is close enough.

A further point applies to all three. Reference books and physical colour charts age. A chart kept in a workshop for several years, exposed to light and handled daily, has drifted from the values it was printed at. A supplier matching a colour against a faded chart is matching against something that no longer represents the code.

For patterned products such as wood grains and stone-look finishes, the system codes do not apply at all. A marble or timber pattern is defined by an image rather than a colour, and the specification has to reference the pattern separately. The methods used to produce those surfaces, and how they differ in permanence, are covered in our guide to laminated and direct-printed wood grain board.

Colour can be measured, and the measurement is what converts a preference into a specification.

Instruments describe a colour as three values. One gives lightness on a scale from dark to light. The other two place the colour on three axes of hue and saturation, one running between red and green and the other between yellow and blue. Together the three locate a colour in a space, and any two colours can be compared by the distance between their positions.

The scale of that number is worth having a feel for, although the thresholds depend on the colour, the size of the sample, the lighting and the observer. A difference below about one is generally not visible under controlled conditions. Around two or three it becomes perceptible, especially where two surfaces are placed side by side. Beyond about five, most observers would describe the two colours as different rather than as a mismatch.

There is a second limitation that matters for demanding work. A single distance figure collapses a difference in three dimensions into one number, which means two samples can sit at the same distance from the reference and still look different from one another, one being slightly lighter and the other slightly redder. Where that matters, the requirement is written as separate limits on the lightness and on each of the two colour axes rather than as one combined figure.

Measured values survive the sample, the shipment and the years. A physical reference does not.

The measurements only become comparable between two parties if the instruments are calibrated and the conditions match. Two laboratories measuring the same board can produce different numbers if one instrument has drifted or if the measurement geometry differs. That is not a reason to avoid measurement, and it is a reason to state the conditions alongside the values.

This is the section that explains the most common colour dispute that reaches a supplier, and it is not a manufacturing problem at all.

A surface reflects light selectively. Different light sources contain different proportions of wavelengths, so the same board genuinely reflects a different colour under each of them. Daylight, an incandescent bulb, a fluorescent tube and a modern LED are four different lights, and a colour that matches under one will frequently not match under another.

The practical response is to name the light. A colour specification that does not state the illuminant under which the match is judged has left the most important variable in the subject unstated, and it is the variable most likely to be responsible for a disagreement.

The most widely used reference for this purpose is a standardised daylight, and other standard illuminants exist for particular industries. Instrumental measurement avoids the problem entirely, because the instrument supplies its own controlled light rather than relying on the room. This is another argument for measuring rather than comparing by eye: a measurement does not care what the weather is doing.

Where a visual check is still required, the useful practice is to carry it out under at least two different lights, one of them daylight-like. A pair of surfaces that matches under both is a genuine match to the eye. A pair that matches under one and not the other has a metameric difference, and the right question then becomes which of the two lights represents the condition the product will actually be seen in.

Pigments are supplied by weight, and where they end up determines what the surface looks like.

The first and largest effect comes from the material the pigment is dispersed in. A foamed board contains a large volume of small cells, and light entering the surface is scattered repeatedly within that structure before it returns to the eye. The result is that the same pigment loading produces a lighter and less saturated appearance in a foam than it would in a solid section. A colour that reads as a deep blue in a solid extrusion becomes noticeably softer in a foamed board of the same tone.

The structural reasons behind this are covered in our guide to skin and core in PVC foam board. For a colour specification the consequence is simple: a colour approved on one product line is not transferable to another without checking, because the substrate has changed.

The second effect comes from surface finish. A matte surface scatters incoming light broadly, which lightens the perceived colour and reduces its apparent saturation. A glossy surface reflects more of the light directly, which deepens the colour and raises the apparent saturation. The same pigment in the same formulation produces a visibly different surface depending on whether the tooling leaves it matte or polished, and the difference is large enough to be mistaken for a colour difference.

A cellular core scatters light internally, and a matte finish scatters it at the surface. Both lighten what the eye receives.

The third effect is time. Pigments fade and surfaces change with exposure to ultraviolet light, heat and cleaning chemicals, and different pigments behave differently. A colour that matches on delivery can diverge over years of service, particularly in exterior applications where one surface faces the sun and another is shaded. The mechanisms are the same ones covered in our piece on how UV, heat and moisture age a PVC surface, and they mean that a colour specification for exterior use should say something about the pigment system rather than only about the delivered colour.

A complete colour specification covers the target, the tolerance and the conditions, and most orders contain only the first of the three.

The commercial case for doing this is straightforward. Colour disputes are expensive out of proportion to their size, because they are usually discovered after the goods have shipped and because they are difficult to settle when the only evidence is two people looking at the same panel and disagreeing. A measured reference removes the argument rather than resolving it.

Where custom colours are being developed, the same discipline applies earlier. The process of turning a colour request into a producible specification, and what the request needs to contain, is covered in our guide to custom foam board production, which places the colour decision among the other choices that have to be made before a first run.