Learn the science behind:
How to Measure the Color of Food
Would you buy a raw brown steak? What about a blue apple? Green coconut flakes? Or a grey orange? Probably not. If the color of food isn’t ‘right’ consumers won’t buy it anymore, even if it is perfectly safe and delicious otherwise.
We all use color to make judgment calls on whether food products are of good quality. But we use color in so many more instances. Farmers use color to help decide whether his or her crop is ready to harvest. And you can use it to determine how your cooking or production process is progressing, e.g. shrimps that turn pink during cooking.
So color clearly is important, but at the same time, it’s very complex. You might have had disagreements with others about the color of something. When choosing a paint color in a store you might have walked outside to check on the color in sunlight, noticing that it looks very different inside than it does outside! If you’ve ever printed photos yourself, you might have also noticed that the photo on your screen can look very different than the one printed on paper!
Since color can be subjective, scientists have developed ways to make this more objective. Whereas this may sound obvious and simple, it isn’t! Measuring color is harder than you might think. For instance, you have to take into account how and which light falls on your color and whether the surface is glossy or matte. Scientists have had to develop methods to overcome these challenges, and nowadays we have several technologies available to help us measure color objectively. We’ll have a look at how they work and how you can use them for analyzing foods.
When you measure color, you need to step away from terms such as red, burgundy, purple, or lavender. Those are too descriptive. Instead, there are ways to quantitatively (in numbers) describe color. By using these harmonized and standardized systems we all speak the same ‘color language’. There are several of these systems. For instance, you might be familiar with hex codes that describe digital colors. The dark orange color on this website for instance is also known as #f26e3f, the light orange as #f2a04e.
In order to get to these numbers, this quantitative and objective way to define colors scientists had to dig deep into the science and physics of color.
Colors are wavelengths
So what really is color? The color that we see is electromagnetic radiation.
Electromagnetic radiation is all around us and consists of waves traveling through the atmosphere. An important characteristic of these waves is their wavelength. These waves make up and down movements, just like waves in a lake or sea. The wavelength measures the length of one such up and down movement.
Some electromagnetic radiation waves have a very long wavelength, they can literally be kilometers long. Radios use waves with this wavelength, they travel well over long distances. On the other side of the spectrum are the very short waves, these can be less than a nanometer long. An example of very short waves are X-rays. Other types of electromagnetic radiation that fall in between these extremes are microwaves and infrared radiation for instance.
As you may have noticed, electromagnetic radiation encompasses a wide range of waves and possible application. You may have also noticed that you can’t actually see most of these waves! The exception here are waves within the ‘visible spectrum’. These are waves with a wavelength between approximately 380 to 750nm. This visible spectrum encompasses all the colors that we can see, every color has its own wavelength. For instance, light with a wavelength of 610nm can be described as orange, around 500nm it’s cyan and 540nm is perceived as green.
Our eyes have cones
But a wavelength is not yet a color. In order for us to actually see the color, the light has to land on our eyes and be processed internally, by our eyes and brain. When light with a specific wavelength enters our eyes it will land on three different types of ‘cones’ inside our eyes. These cones capture the incoming light where each cone is optimized for a slightly different wavelength. Together, these cones send a signal to our brain, which then processes the signal into a color!
Describing color objectively
When we describe colors and try to do so with a number, we need to take into account how our eyes process these incoming wavelengths. How we process these colors impacts how we perceive them. In some color ranges we might be able to detect more subtle differences than in others.
We won’t go into too much detail here with regards to the theory. Instead, let’s have a look at what scientists have come up with over the years that combine the science of radiation with how we humans perceive colors.
The CIE & Tristimulus values
Over a hundred years ago an organization, the CIE, was established to define the first standardized way of describing color, using the scientific knowledge available at the time. This system, the CIE 1931, is still used in a variety of industries and various other models have been derived from it.
The core of the 1931 system is that it gives you so-called tristimulus values. These tristimulus values take into account the mechanism of those three cones in our eyes and use three numbers to describe every color. There exist different tristimulus value systems that each have a slightly different meaning to what each number stands for but the science behind them all is very similar.
One tristimulus system still actively used is the RGB system in which R stands for red, G for green and B for blue. You might have heard about this system since it’s commonly used for computer screens. It is not as commonly used in food research applications though.
Another common tristimulus system used to describe colors is the L*a*b* system. You will come across these numbers quite regularly within the food science world. The value for L* represents lightness. Something that is completely white will be 100 and something completely black is 0. a* is a measure for the green to red value (so how red and green something is) and b* represents blue to yellow.
From these values, using some formulas, you can then derive the hue (which is the name of the color, e.g. orange) and the intensity (or saturation) of the color.
Measuring color of food products
Luckily you can measure color without fully knowing all the math and physics behind every single measurement! Most measurement devices come with their own instructions on interpreting data to some extent and do a lot of the more complicated math for you behind the scenes.
If you want to measure the color of your food there are roughly two options available: colorimeters and spectrophotometers. Both have their pro’s and cons and are best suited for slightly different situations (and budgets!).
A colorimeter is a slightly simpler device than a spectrophotometer, so we’ll get started with this one. A colorimeter works by projecting light onto your sample. It filters this light through three filters, to mimic your eyes and then analyzes how the light is reflected from your sample. As the outcome, you get a tristimulus value, this can for instance be an L*a*b* value.
Colorimeters are relatively simple devices and don’t need a lot of advanced software (or hardware) to operate. They are generally the cheaper option of the two, but they also don’t provide you with as much data. Because of the way a colorimeter works, by always using one and the same light source, it might not be able to pick up on all differences between colors.
A good choice for quality control
Sometimes though you don’t need to know that many details about your sample. All you might want to know is whether the color stays the same during a production run. For these types of applications, a colorimeter works very well. Colorimeters work fast and are pretty sturdy.
However, if you want to know all there is to know about your color, or your samples are somewhat more complex, a colorimeter might not provide you with all the data you need.
‘Old’ vs. ‘new’ colorimeters
You will find a lot of articles on colorimeters that refer to a colorimeter that does not give an L*a*b* value, but instead just one number. In these systems, you choose one wave length of color and then compare various samples at this wavelength. This is a common way to calculate the concentration of colored solutions (using Lambert-Beer). Nowadays though, colorimeters have become a lot more advanced and can do more than just provide this single value. A reason for still finding a lot of articles about these ‘old’ colorimeters seems to be that they are part of the high school curriculum in many countries.
If you need a bit more information about your color you might need a spectrophotometer. A spectrophotometer works according to the same principle as a colorimeter: it shines a light on a sample and then analyzes its reflections. However, a spectrophotometer can do a more advanced analysis of the sample due to a few subtle changes.
First of all, it can work with different light sources, instead of just one. This can be important for some samples which might look similar under one light source, but are different under another.
Also, it doesn’t use those three light filters a colorimeter does. Instead, it can completely analyze the spectrum of the light, showing exactly which wavelengths are reflected from your sample.
Lastly, another convenient property of some types is that it can beam light onto a product under several different angles. This is especially relevant when you’re trying to analyze the color of very uneven samples.
Generally speaking, when you’re doing research on your food and really want to understand color differences, you’d want to use a spectrophotometer. These devices used to be large and bulky, but nowadays they’ve shrunk in size (and costs!) making them more accessible.
Deciding on a technique
So how do you decide which one to use? And, how do you decide how and whether to measure the color of your samples? Let’s have a look at some of the considerations.
Goal of your measurement
As with any analysis you might do, you should always have a clear goal in mind. It is of no use to just collect a lot of data and then try to figure out what you can do with it. A few questions you could consider are:
- Is sample A the same as sample B? Or, similarly, are all my product made in the factory of the same color?
- A reason for looking into this question might be to ensure that your product always looks the same. Or it might be to check whether certain changes you’ve made to your product have impacted the color.
- Do my samples have a color of a specific pre-determined spectrum?
- If you’re making products that need to be of a very specific color, for instance, because of the identity of a brand, you want to make sure you hit that exact color.
- How does my color change over time?
- If the color of your product isn’t stable over shelf life you might want to investigate just exactly what is changing. This can then help you to determine what the cause of these changes might be.
As with any analytical technique, you might want the fanciest device, but if you can’t fit it in your workflow, there’s no use for it. Consider where you want to do your measurements and how fast a measurement needs to be completed. Do you have well-trained people available to do the analysis, or should just about anyone be able to run it. This is also a good time to consider your budget, both for the equipment itself as for possible costs required to run it.
Last but not least, you need to have a look at your sample. No matter how fancy the device, food can still be pretty complicated to measure properly.
Color homogeneous vs. heterogeneous
Is the color of your sample very homogeneous or more heterogeneous? If your sample is homogeneous measuring the color tends to be pretty straightforward. However, if your product has a lot of different colors (which, in all honesty, is the case for a lot of food products!) measuring color becomes a lot more complicated.
Think of an apple that’s ripening, it might have, yellow and some green areas. Which area do you decide to measure? The yellow one, the green one, all of them? What does the information that you get out of such an analysis even teach you?
Reflectance vs. Transmittance
Another aspect to think about is whether the food lets light through or not. If a food lets light through, this could be colored vitamin water for instance, you need to use a technique that uses this. You can use transmittance tests. During such a measurement light is beamed through a sample. The light that comes through the sample is then analyzed. Not all colorimeters can do these types of measurements, so would be something to look out for.
In other cases, when the light can’t pass through (e.g. an orange or a piece of bread) you will need to use a reflectance method. In this case, light is bounced onto a sample, and the device measures what is reflected by the food or drink.
Once you’ve had a good look at your sample, and aligned your objectives, you are in a far better place to look at choosing the most appropriate color measurement technique!
CIE, International Commission on Illumination, link
Tim Mouw, Colorimeter vs. Spectrophotometer, Oct-7, 2019, link
Ken Phillips, Spectrophotometer vs. Colorimeter: What’s the Difference?, Aug-6, 2020, link
Christine H. Scaman, Chapter 43 Spectroscopy Basics, in Handbook of Food Science & Technology, link