Visual color is closely related to perceptions. Consumer perception or purchase decision is made even prior to tasting food. Color is defined as the impact of wavelength in the visual spectrum from 390 to 760 nanometers (nm) of the human retina. Reflected light is perceived as color. To detect the color, either the human eye or the instrument used must be capable of recognizing the object and translate the stimuli into a perception of color.
Various manufacturers of colorimeters and spectrophotometers market their equipment as portable, benchtop, and in-process equipment. In food, effect of color is very important to “determine effects of raw ingredients to the finished product; its shelf life or changes due to processing; ensure suppliers are guaranteeing a consistent colored material; and determine if the final product meets internal quality standards established,” says Cliff Walsh, operations director at American Licorice Company.
Subjective Evaluation of Colors
Color in raw materials or in finished goods is important to a food processor. Easier said than done is the quick approach to check colors with the naked eye. “There are disadvantages associated with visual examination,” comments Ramon Navoa, the director of innovation at American
Licorice Company. “Judgment is influenced by lighting, visual deficiencies of the eye, or in a trained panel based on repeatability. All these affect variability,” he adds.
A subjective evaluation system can include matching the colors, Pantone color matching system, and actual photos of finished or raw materials. Another system called Munsell is used by the USDA. The color system divides hue into 100 equal divisions around the color circle. Applications include dairy products such as milk, cheese, egg yolk, beef fruits, and vegetables. Food manufacturers use Royal Horticultural Society’s color charts to standardize food colors. The Natural Color System Digital Atlas also has more than 1,950 colors that can be used to compare colors. Any visual examination or comparing of colors has inherent constraints and are product dependent.
Agricultural commodities may have batch to batch variation and getting a consistent supply may be more critical. For example, cinnamon’s flavor may be perceived to be meeting the aroma specification, but the color is variant based on the region the cinnamon is harvested, bark color, age of the bark, intentional contamination, and country of origin. Knowing variability exists with the color is hard to explain to a consumer who has expectations on the end-product’s visual appearance. Some naturally occurring colors also degrade based on exposure to heat, sunlight, processing conditions, and storage. This adds complexity to the color consistency expectations.
Artificial colors added to food have their inherent drawbacks. Colors are added to food to offset color loss due to light, air, extreme temperatures, storage, and moisture. Others use artificial colors to mask natural variations in color or enhance naturally occurring color. Artificial colors provide identity to the product, protect flavors and vitamins from damage, or are used for decorative purposes.
Colors, either natural or synthetic lakes or dyes, have inherent properties and applications. Applications dictate if the measurement of color becomes critical to monitor in process samples for color degradation or conformance to a standard. The product appearance may be a subjective phenomenon, but when it comes to color there are instruments available in the market. Many instrument manufacturers can provide assistance in providing equipment for specific applications.
Some naturally occurring colors degrade based on exposure to heat, sunlight, processing conditions, and storage.
The Instruments and Their Applications
Common colorimeters are Konica Minolta’s chroma meter, HunterLab colorimeters, and Hach Lange colorimeters. Colorimeters use sensors and simulate how a regular person views an object and quantifies the color differences between a standard and a production sample. Colorimeters employ three photocells as receptors, just like a human eye. The same wavelength is used to measure, and hence, measurement conditions do not change. A light source and a microprocessor convert colors to internationally accepted numeric values. Colorimeters feature a wide range of apertures and illumination for specific applications and various levels of data processing. They are good for measuring and comparing color differences between two specimens, strength determination, fastness determination, and shade sorting.
The best way to measure opaque liquids, solids, pastes, or powders is to use a 45/0 degree geometry instrument with a horizontal sample port. A liquid sample can be poured into the sample cup and measured. Blocks of cheese or slices of meat can be placed directly to the sample port aperture. With a circumferential illumination and a large measurement port, flakes, chips, and/or chocolate disks can be measured. A QC may have a standard target color that must be repeatedly manufactured by the production team. Colorimeters are ideal when the standard and measured batch are non-metameric, e.g. production batches. Natural colors such as chlorophyll, carotenoids, and anthocyanins can be measured in a colorimeter and quantify the pigments present in a food.
An inline color monitoring system mounted over a production line can give real-time data. Translucent samples will pose a concern and a “ring and disk” assembly is used to measure this type of sample. Brewed tea, for example, can be poured into the transmission compartment and a reading can be obtained.
The amounts of red, green, and blue needed to form any given color are called the “tristimulus” values, X, Y, and Z, respectively. The measurement is expressed in terms of X-Y-Z and the user can pinpoint the differences in lightness, chromaticity, and hue between the target and the sample. The color measurement taken in one location can be compared with another location or a different time in an internationally accepted terminology. This eliminates color perceptions and judgmental differences between technicians.
The Commission Internationale de l’Eclairage (CIE) defined the color of an object on three primary stimuli: red (700 nm), green (546.1 nm), and blue (435.8 nm). Sometimes, tristimulus systems of representation of colors are not easily understood by the users in terms of object color. Other color scales, therefore, were developed to relate better to how we perceive color, simplifying the overall understanding.
A three-dimensional rectangular L, a, b, color space uses L (lightness) axis – zero is black and 100 is white; a (red to green) axis – positive values are red, negative values are green, and zero is neutral; and b (blue to yellow) – positive values are yellow, negative values are blue, and zero is neutral.
There are two popular L, a, b color scales in use today: Hunter L, a, b and CIE L*, a*, b*. They are similar in organization, but will have different numerical values. Hunter L, a, b and CIE L*, a*, b* scales are both mathematically derived from X, Y, and Z values. Hunter scale is over expanded in the blue region of color space, while CIE scale is over expanded in the yellow region. The current recommendation of CIE is to use L*, a*, b*.
A spectrocolorimeter is a hybrid instrument that gives data such as X, Y, and Z or CIE L* a* b* values. These are priced similar to the spectrophotometer. They are basically a spectrophotometer except that it does not output spectral data (%R) at various wavelengths. They are mostly QC lab type instruments.
Spectrophotometers measure light reflected, transmitted, or absorbed from a food product to a known standard. They have more sensors and measure spectral reflectance of an object at each wavelength on a visible spectrum continuum. They work best for liquid samples. A specimen is exposed to light and the reflected light waves are displayed as a curve on a graph. The size and shape of the curve is called a reflectance curve and is unique to each color.
Reflectance measurement (reflectance factor) is basically a reflectance of a food sample at a given wavelength compared to reflectance of the perfect diffuse white measured under the same exact conditions. The reflectance color measurements are more rapid. These are expressed as %R. If transparency of a dye solution is measured, it is denoted as %T. This quantity is equal to the percent of light at a given wavelength, transmitted through a thickness of 10 millimeters.
Choice of instrument depends on the food and the application type. Color discrimination threshold of the human eye greatly differs from the color differences defined by CIE. Using CIE values, color modeling has been developed for specific applications. Reflectance data can be reported as CIE L*a*b values: L – Light, a* – red, and b* – yellow.
Color Modeling in Fruits and Vegetables
Research attempts have been made to model color values. For example, vegetables when over-blanched can change to a green color. Depending on chlorophyll and chlorophyllide destruction, a generalized model for vegetables could be found. Chromatic changes of broccoli under modified atmosphere packaging at 20 degrees Celsius in perforated and unsealed polypropylene film packages for a storage period of 10 days indicated that using L*c*h* color space diagram, the modified atmosphere generated inside the perforated film packages with 4 macro-holes was the most suitable in maintaining the chromatic quality of the broccoli heads (Rai et al. 2009).
An important parameter of the postharvest life of tomatoes is color. One color model correlates the color level and biological age at harvest (Schouten et al. 2007). Data were analyzed using non-linear regression analysis and found that biological age of tomatoes can well be predicted at farmers’ level and can save a lot of postharvest losses. Interestingly, they also found a very good correlation between the color values and tomato firmness.
Precision of prediction using models having the parameters of a, b, and their product (a×b) was verified by sensory evaluation of 55 ripe mangoes. It was found that the fruits predicted to be mature could ripe with high-satisfied taste, while the ones predicted to be immature or over mature were mostly rejected by the panels (Jha et al. 2007). Hence, these mathematical relationships between ripeness, overall quality, and freshness index can be calculated.
The relationship between color parameters and anthocyanins of four sweet cherry cultivars using L*, a*, b*, chroma, and hue angle parameters (Berta et al. 2007) indicated that chromatic functions of chroma and hue correlate closely with the evolution of color and anthocyanin levels during storage of sweet cherries.
It was also shown that color measurements can be used to monitor pigment evolution and anthocyanin content of cherries.
The above paragraphs indicate that significant attempts have been made to model color values or combination thereof for prediction of various surface, as well as internal quality parameters, of various fruits and vegetables. However, very limited work on modeling of color values of other foods, such as food grain and oilseeds, are reported for prediction of their quality parameters. The coefficient of determination of these models may not always be as high as expected. In such cases, one may try to obtain the complete spectra of specimen instead of individual color values (L*, a*, b*, etc.) in the visible range of wavelength (400 to 700 nm) and develop models using the absorption or reflectance data.
Hue value—which identifies whether an object is red, yellow, green, or blue—research is underway and new equipment is being invented to address hue values. With more research underway and companies investing in color detection instrumentation, the visible color differences observed during stress of drought, heat, or other deficiencies or development of fruits will be possible in the near future. Subtle differences in color and purchasing decisions will be taken as a marketing advantage.
Dr. Veeramuthu works at American Licorioce Company as a senior QA manager and can be reached at 219-324-1464 or email@example.com.
References Furnished Upon Request