Chinese (Simplified)EnglishFrenchItalianPortugueseSpanish
|

Ocean Optics
Worldwide Headquarters
Largo, Florida, USA

+1 727-733-2447

[email protected]

Sales, Service
& Support Facility
Duiven, The Netherlands

EU:
+31 26-319-0500
FR:
+33 442-386-588

[email protected]

Ocean Optics GmbH Sales,
Service & Support Facility
Ostfildern, Germany

+49 711-34-16-96-0

[email protected]

Sales Support
for the
United Kingdom

+44 1865-819922

[email protected]

Sales, Service
& Support Facilities
Shanghai, PRC – Beijing, PRC

+86 21-6295-6600

[email protected]

Close
Home > News & Events > Controlling the Color Consistency of Beverages Using Spectral Sensing

Controlling the Color Consistency of Beverages Using Spectral Sensing

by Miriam Mowatt, Applications Scientist

Introduction

Spark Accessories - Whiskey MeasurementColor is a crucial factor in the quality assurance of food and beverages. Customers strongly associate color with food quality, and consistent color helps to ensure brand integrity. For food and beverage producers, consistency is vital.

This article will show how transmission spectroscopy can be applied to achieve fine control of beverage color. Using a spectral sensor, measurements were taken of whiskey samples. Natural food colorants were added to simulate minor variations in the color such as might be seen from batch to batch of whiskey. Color is notoriously fickle and depends on the background light, the sample, and too often the subjectivity of human judgment. Spectroscopic color measurements can objectively and accurately distinguish between samples, even when our eyes cannot.

 

Measuring the Color of Beverages: Whiskey

Measuring color is challenging, as the way our eyes detect color is not directly related to easily measured parameters such as light intensity. Remember the recent furor in social media where no one could agree if a photographed dress was blue and black or white and gold? Color is dependent on the sample, on the illuminant and on subjective human perception. Spectral sensing, however, makes color measurement quantifiable, repeatable and consistent. Measuring the amount of light across the entire visible wavelength range brings real advantages when compared to traditional colorimetry due to the higher resolution and wider bandwidth of the measurement

In food quality control, consistency of product color is crucial. To control color, it is vital that the measurement is repeatable from system to system. Typically, a number of identical measurement setups are used across multiple locations or factories, so reliable system to system measurement is important.

To test for repeatability, the color of a single whiskey sample was measured using three Spark spectral sensors. The Spark is a lightweight and compact spectral sensor, ideal for use in handheld devices or integration onto production lines. The spirit was measured in a cuvette and illuminated with a white light LED. The cuvette was directly attached to the sensor using a dedicated cuvette holder, which also contained the LED (this was a prototype sampling setup). Reference and dark measurements were repeated for each sample.

The color was calculated in xyz color coordinates. This is a measurement system used to quantify color in a 3D coordinate space. The xyz parameters can be derived from the commonly used CIE 1931 XYZ system such that x = X/(X + Y + Z), and y = Y/(X + Y + Z).  The projection of xy is displayed in the CIE 1931 chromaticity diagram in Figure 1. The color measured by each Spark unit is shown on a comparable xy plane in Figure 1. The results show that each Spark measured the color within 0.001 units, demonstrating that the Spark is able to achieve highly repeatable results between units.

Figure 1: Whiskey color measured by Ocean Optics’ Spark spectral sensor (right). The measurement of the same whiskey was repeated with three Spark units and is displayed using xy color coordinates.

Figure 1: Whiskey color measured by Ocean Optics’ Spark spectral sensor (right). The measurement of the same whiskey was repeated with three Spark units and is displayed using xy color coordinates.

Measuring Color Differences

To simulate an example of a color consistency control test, three samples of a single whiskey were measured over a series of small color changes. Caramel, a natural food colorant used in the industry, was added in a series of increasing concentrations to each sample. The color of each sample was compared with a reference whiskey sample containing no extra caramel other than the caramel already present from the samples’ manufacturers.

To quantify the color difference between samples, delta E was calculated. Delta E is a measure of color difference. It is difficult to reliably determine color by eye, since it is often subjective and can look different depending on ambient lighting conditions. Delta E can be used to quantify a color difference against a reference. It is defined using the color coordinates CIE L* a* b*, which is another commonly used color coordinate system. Each parameter refers to the following: lightness or intensity for L*, red/ green chromaticity for a*, and yellow/blue chromaticity for b*. The calculation for delta E is the following:

ΔE = √(L*2 – L*1) + (a*2 – a*1) + (b*2 – b*1)

A delta E value between 1 and 3 is typically undetectable by eye. A delta E of 1 is often referred to as the lowest detectable limit for someone who is trained in color matching. For a perceptible color difference for the average person, a delta E of 2 or 3 is often used (1), (2).

As you can see in Figure 2, the whiskey samples used for this study are very close to each other in color, with differences nearly impossible to distinguish by eye. While a delta E of 1 is the absolute lowest limit of perceivable difference, a delta E of up to 10 appears close to identical for most people.

Figure 2: Whiskey samples containing different concentrations of additional caramel for color adulteration are aligned in a series with the lowest concentration on the left and the highest on the right.

Figure 2: Whiskey samples containing different concentrations of additional caramel for color adulteration are aligned in a series with the lowest concentration on the left and the highest on the right.

The delta E data calculated for these measurements are displayed in Figure 3. The graph clearly shows a linear trend between color and concentration of the additional caramel colorant. The samples measured reach a delta E of approximately 10, and go down to 1 or just below. This demonstrates that minor differences in color can be detected and quantitatively identified using spectral sensing. It is also clear from the linear trend of the data that it is possible to determine exactly how much additional caramel is in each sample. This is a powerful technique for identifying variation in the production of beverages, even below the limit of human perception. Beverage color can be identified accurately and rapidly using the Spark. This enables food and beverage producers to rapidly and reliably monitor the color and consistency of their product during or after production.

Figure 3: Using the Spark spectral sensor, we measured the color difference, delta E, between a whiskey sample and the same sample with caramel added as a color adulterant. The results show a linear trend between added caramel concentration and delta E.

Figure 3: Using the Spark spectral sensor, we measured the color difference, delta E, between a whiskey sample and the same sample with caramel added as a color adulterant. The results show a linear trend between added caramel concentration and delta E.

Conclusion

Spectral sensing is a powerful tool for food and beverage color measurement. Repeatable and reliable, spectral techniques can be used to detect minor changes in color at or below the limits of human perception. Here, the Spark spectral sensor was used to measure the transmissive color of whiskey samples. The ability to detect small shifts in color was demonstrated with a Delta E measurement. Ensuring color consistency is an important aspect of food and beverage quality. With modern spectral sensing tools, color measurements are reliable, quick and easy to do.

Works Cited

  1. Threshold and suprathreshold perceptual color differences. Witzel, R., Burnham, R., Onley, J. 1973, Optical Society of America.
  2. A re-determination of the trichromatic coeffients of the spectral colours. Wright, W. D. 1928, Optical Society of America.