A formula for evaluating colour differences for thread sewn into fabric samples

Coats plc.’s research efforts during the 1970’s led to a breakthrough in colour difference predictions. The result was a new colour difference formula ‘JPC79’, which was tested in a commercial environment providing fewer wrong decisions when compared with the best colour matcher from a group at Coats plc. The formula was designed from a large number of thread winding card samples from which colour difference pairs were formed. The formula was later
modified and became a standard in the textile industry known as the CMC(l:c) colour difference formula.

However, visual colour differences for two thread winding card samples, judged in a viewing cabinet side-by-side, may not reflect on a real world scenario, for instance, when a thread is to be matched while stitched upon or within another kind of material, such as a fabric or leather
sample. It was therefore of interest to assess such a scenario by using thread end products. Preparing those sample pairs can be time consuming and expensive. Therefore, it was also of interest to assess substitutes on a digital screen with the aim to find a suitable correlation
between two methods.

The CMC formula is the preferred equation for calculating colour differences for matching textile samples. The question here was whether the equation can also be used for predicting visual colour differences that were obtained from assessments of end products without compromising its prediction performances.

The primary findings for this project and experimental setups are; (1) that the variation in sample type and presentation can alter the perception of colour differences in human observers, significantly; (2) it is generally possible to use digital substitutes for psychophysical
experiments as such following the same trends as they occur from physical samples that are assessed in a viewing cabinet, (3) advanced colour difference formulae can predict those visual colour differences obtained from assessments for end-products once they are optimised
(parametric functions), and an overall size parametric factor is applied for various ‘thread sewn into fabric samples’.