The selection of suitable glazing products for building design is no mean feat.

To meet performance specifications, we must tread ever so lightly through a maze of competing glass quality values to create inspired architecture and awesome façades.

Beyond the well-worn thermal performance discussion surrounding U-values, light transfer and reflection typically bring about a secondary debate that can put many a design team on hold.

Spectral sensitivity (the relative efficiency of light wavelength detection) is a tricky process, with visible and near infrared selection competing to control daylight distribution, glare risk (internal and external) and solar heat gain. Getting a bit more technical, ultraviolet light can lead to fabric fading, though this rarely becomes a point of discussion when it comes to design.

All of this is a highly manageable numbers game, but what of the glass colour, a major limiting factor in any glass catalogue and selection process? As the ultimate architectural showpiece of a design, assuming lots of glass that is, where is the numerical specification for colour?

Today, more often than not, we choose glass colour through a process of selecting an existing façade precedent (which can be a good method, but information is not always available) or by choosing a 300 by 300 millimetre glass sample of preference (which can be fraught with time penalties). Happy with a precedent or sample, we then go looking for alternatives in a glass catalogue, order them from the suppliers and potentially install them for comparison.

Glass Sample

Glass sample

The problem with this method is that it is qualitative, meaning we are throwing caution to the wind by leaning on 'colour' to procure the most fundamental element of many façade designs. It is also highly subjective as we all see colour differently, a real risk if many people are involved in the design decision.

Why can't we compare colour in a quantitative manner similar to how we treat performance values before we order samples from a supplier, saving time, money and design team headaches? Enter the CIE L*a*b* colour scale and the crazy world of colour science.

Specified by the International Commission on Illumination to represent the most complete representation of all colours visible to the human eye, the CIE L*a*b* colour scale provides a non-linear relationship between the values of L* a* and b*. This means it is useful for measuring the perceptual difference between two glass samples, because a uniform change in L* (green) a* (blue) and b* (yellow) will result in a uniform change in perceived colour.

CIE L*a*b* Colour Scale

CIE L*a*b* colour scale

Today we use this colour science within glass production runs to ensure minimum colour variation. For example, the ASTM standard (ASTM C1376-03) for aesthetic quality requirements for coatings applied to glazing sets a colour difference of ΔE* ≤ 4.5 for each run, therefore it is deemed to have an almost imperceivable visual colour difference.

So, we have the numbers to compare and can even convert them into CMYK or RGB for more accurate colour renderings but we are simply not using them. The next time you are heading down to a design meeting to discuss glass selection, when the subject of colour selection is raised, go quantitative and not qualitative, to save time, money and design team headaches.

Images courtesy of Darren O'Dea