Colour assessment, our eyes versus spectrophotometry.

Some Colour experts passionately swear by the visual colour assessment; a few even declare this sacred.

Colour is a visual property, there is no doubt about that. And if you don't see a colour difference visually, then it should be fine, right?

My best experience has to be that of a black colour development for a dashboard part of a well-known car brand; that colour expert was of the opinion that this should be done visually, colour spectrophotometers out of the question, things like that.
In my opinion, the colour was an exact match, but the customer saw too much difference and wanted it slightly (!?) redder... Long story short: after 15 corrections, during which I had long since lost sight of the differences, we secretly sent the very first match again.
And what do you think: Success!

Now, black is difficult to match anyway, the colour spectrophotometer also has great difficulty with this. With the D/8° geometry, which, due to the too high amount of collected light reflection does not or hardly correlate to the visual perception, it is very difficult to match the black colour exactly spectrophotometrically; although the 45°/0° geometry supports well on the a* and b*, it is the L* value that is difficult here, partly because the surface can have quite a bit of variation when injection moulding an Engineering Plastics plate.

You perceive colour with your eyes, but you see with your brain. And that brain is not always reliable, it is sometimes tired, it sometimes lets you look at objects a little too long and, usually due to too little experience, it does not interpret the differences properly. Our eyes are not equally differentiating for all colours; we can distinguish differences in green better than the other colours due to the wavelength sensitivity of the cones in our eyes. And both the brain and the eyes can differ from person to person.

And then I'm not even talking about the way of visually assessing colour: must the inside of the light cabinet (non-negotiable condition!) be white, black or (dark) gray - have you set the angle at which you look correctly (at 45° ?) - should the light in the analysis room be turned off - whether or not to measure the gloss - set how long you look at the sample plaque - how do you deal with differences in polymer between reference and sample - how do you define (together with your customer) which differences are still acceptable and which are not.

So much more often the visual assessment of colour is called subjective, it is very difficult to quantify and even more difficult to communicate.

Factually the colour spectrophotometer is and remains the most objective and most reproducible and repeatable analysis technique, which does not require extensive experience, which makes colour (difference) quantifiable and where the colour, via the L*a*b*, is easy to communicate.

But this too has its limitations.
The most common geometry, D/8°, works very well for the colours in the visible range, but, as mentioned, for black colours there is hardly any correlation with the visual perception. And the 45/0 geometry can only partially solve this, especially if it is very deep black, although BYK Instruments has released a very interesting and promising 45/0 spectrophotometer for this purpose.
Because a colour spectrophotometer works purely mathematically, it is able to express very small colour differences numerically, which also has a disadvantage. Customers tend to tolerate only very small colour differences, just because you can measure it, but this is not always visible and is colour depending. For the gray and white tones, where the smallest differences are already visually visible, and which are often used for applications with more complex plastic compositions (just look at your  electrical circuit breaker panel), the desired extremely low tolerances are even unworkable!

So, what do you do in practice, do you only use the spectrophotometer or do you assess it visually?

In my opinion it is a matter of perspective!
Because if you are in general developing a colour, the spectrophotometer and visual assessment (by a well-trained and experienced colourist!) go hand in hand. The spectrophotometer provides accurate numerical information as to whether you are close enough to the target and the visual assessment is necessary to establish the correlation with practice, in absolute sense with the corresponding relevant maximum tolerances.
Because if the colour to be developed is red-orange-yellow-blue-purple and even green, the spectrophotometer D/8° is an indispensable and powerful instrument, together with good development software, to quickly achieve beautiful colour formulations and results.
Because if the colour to be developed is deep black, then your eyes are of great value in assessing colour differences, sometimes in harmony, and the spectrophotometer is supportive in determining tolerances.
Because if the colour to be developed is white or grey, the spectrophotometer is of enormous value in assessing the value of the visible differences and determining relevant and workable tolerances.
And once the colour has been approved by your customer, you set the spectrophotometer colour values ​​as a specification and also use them for quality control in batch productions.

It is important that you, in collaboration with your customer, clearly define how you are going to analyse, what the relevant maximum tolerances/approval criteria are, what the strength and limitation of the equipment (and corresponding geometry) are and use all resources available to you, both spectrophotometrically and visually, to develop and produce the right colour (according to the customer).