NPL Technical Update - Calculating colour differences

There are several different ways to calculate a colour difference. They include Delta E absolute, Delta E CIE 1994, Delta E CIE 2000, Delta E CMC etc. As well, there are changeable parameters in the case of some, increasing the possibilities yet again.

So, which method should you and your customer use for the most suitable colour communication and control? There is probably no one definitive answer, because it will depend on the type of work and the method of reproduction. If you set colour matching standards too wide your customers won’t be happy, but if they are set unreasonably close your presses will be stopped too often.

A good start to the process of setting Delta E standards with your customer could be to look at your process capabilities. Measure the L*, a*, b* values within a job that is held under commercially achievable colour consistency by your most experienced printer. By taking measurements both across sheets and from sheet to sheet, a large database of values can be gathered.

The problem now is to calculate the Delta E values between each and every combination of samples, and for each and every Delta E type. This would be a prohibitively lengthy task if it were to be done by an instrument. For example, just twenty L*, a*, b* measurements give rise to 190 pairs for Delta E calculation!

The best way to do this is by using the formulae for the various Delta E calculations in a table that produces all the results of all the combinations. This matrix of Delta E values can then be analysed for ranges and maximums. You and your customer can study them against the visual samples concerned, and make good decisions on the suitability of the variation in colour.

The matrix of Delta E values can also be recalculated easily for any chosen Delta E type, and again allows you to compare the calculated Delta E data with the visual matches.

The National Printing Laboratory can set up spreadsheets to enable you to make these calculations. We can also measure the L*, a*, b* values for you, and help you with your analysis. If you’ve got a problem with colour, give us a call!

E-mail Lee Mellet for more information lee.mellett@eng.monash.edu.au