MATERIALS AND METHODS

The study was carried out at the University of Szczecin and the University of Life Sciences in Poznań, Poland, from January, 2019 to February, 2020.

Experimental material: The material for this research was commercial Gouda cheese (purchased in a local shop in Poznań).

Physicochemical characteristics of cheese: In order to define the characteristics of cheese, the content of water, fat and protein were marked.

The total solid content in cheese was determined according to ISO²¹. The test sample was weighed, mixed with dried sand and kept in a drying oven set at 102°C for 3 h. Following this, the sample was cooled and weighed. The procedure was repeated until the difference between two subsequent weights was 0.0002 g or less. The total solids content was expressed as percentage by mass:

Water content (%) = 100-The total solid

The fat content in cheese was determined using the ISO²² method. The test sample was digested with hydrochloric acid and then ethanol was added. The acid-ethanolic solution was extracted with diethyl ether, light petroleum and the solvents were removed. The mass of the substances extracted was determined. The described method was based on the Schmidt-Bądzyński-Ratzlaff principle²².

The protein content of cheese was determined by measuring total nitrogen using the ISO²³ method. To convert the nitrogen reading to cheese protein, the nitrogen reading was multiplied by the factor of 6.38.

Sample preparation: The Gouda cheese was cut by means of an electric cutter into slices 1.7 mm thick. The cheese slices were square. The side length of the square was 10 cm (Fig. 1). In the research, the thickness of a layer was increased from 1.7 (1 slice) to 13.6 mm (8 slices). The color was always measured on the surface of slice one. Each subsequent slice was placed at the bottom and the color was measured to determine the effect of sample thickness on color. The temperature of samples was 20°C.

Fig. 1:	One slice of cheese (1.7 mm) on the white and black background (white and black paper)

Devices and color system: The color analyses were conducted using a X-Rite SP60 spectrophotometer (Grand Rapids, Michigan, USA) and a portable spectrophotometer spectro-pen (Dr. Lange, Düsseldorf, Germany). The measurements by X-Rite SP60 were recorded by using: illuminant D65, standard observer 10°, geometry d/8, measuring area 8 mm, SPEX (Specular Component Excluded) setting. In spectro-pen the measurements were recorded by using: illuminant D65/10°, geometry 45°/0°(circular) and 5 mm measuring area. The color was measured in CIE L*a*b* system. This color model permits an objective color representation and its use is essential for applications where the results must match those of human perception²⁴. In the CIE L*a*b* space, color is described by three components:

The colors can be described by a combination of their lightness, green/redness and blue/yellowness. CIE L*a*b* is similar to CIE LCh (L: Lightness, C: Chroma/saturation, h: Hue). The difference between them is in the different coordinate systems used to describe the two spaces: the CIE L*a*b* space is described in Cartesian coordinates, while the CIE LCh space in cylindrical coordinates. The relationships between their respective coordinates are therefore as follows²⁵:

For color measurement the X-Rite SP60 was used and the results for L*, a*, b*, C, h were obtained on the device display. In the Spectro-pen-L*, a* and b* were obtained.

Background: White and black paper were used as background. The values of L*, a*, b* for black and white papers measured with X-Rite SP60 were: 35.55, 0.01, -1.20 and 91.65, 1.10, -6.95 (±0.01), respectively and for Spectro-pen: 26.1, 0.9, -1.5 and 92.5, -0.6, -1.4 (±0.1), respectively.

In the case of L*a*b* space, the difference between two colors is calculated by the formula:

ΔE L*a*b* = [(ΔL*)² + (Δa*)² + (Δb*)²]^1/2

A standard observer sees the difference in color as follows:

Statistical analysis: Four cheeses were taken and each sample was analyzed thirteen times. Means and standard deviations were calculated using Microsoft Excel spreadsheet from the Microsoft Office package. Data collected experimentally were analyzed statistically using the t-statistics. A two-sample t-test was used to determine if two population means were equal (test at a significance level of p<0.05). All calculations were performed in the Statistical 12.0 software package.

RESULTS AND DISCUSSION

The Gouda cheese used as material for the experiment was characterized by 44.4±0.5% water, 25.0±0.2% protein and 26.5±0.3% fat content.

Table 1:	Influence of the background color on L* a* and b* coordinates
a, bDifferent letters in a row denote statistically significant differences between results, separately for each instrument (p<0.05), L*: From black to white, a*: From green to red, b*: From blue to yellow

Table 2:	The difference between colors of cheese for white and black background

In the Jo et al.⁶ study, Gouda cheeses with water content of 41.4-46.3% and fat 26.6-28.5% were characterized by L* = 83.5-84.5, a* from -1.43 to 3.27 and b* between 30.7-34.6.

Means and standard deviations for the color parameters as a function of thickness for white and black background are shown in Table 1.

The background significantly affected all parameters values up to two slices of cheese (3.4 mm).

No differences in L* coordinate were found (between white and black background) for the thickness equal to and more than 5.1 mm (for both devices), in coordinate a* for the thickness equal to 11.9 mm and 10.2 (for X-Rite and Spectro-pen, respectively) and in b* coordinate, for the thickness equal to and more than 8.5 mm.

In color evaluation, an untrained human eye would not see the difference in the influence of background on cheese color already for 3 slices of cheese, i.e., from the layer thickness of 5.1 mm (ΔE<2) (Table 2). That is why an objective, instrumental evaluation of cheeses is so important.

The background had no statistically significant effect on chroma from 8.5 mm cheese thickness and on hue from 11.9 mm cheese thickness (Table 3).

Because of the pioneering character of research, there exists no dairy products literature on the impact of background colour or sample thickness on the color coordinates in the CIE L*a*b* system.

Sandusky and Heath¹⁴ reported that background color significantly affected the meat color when the 0.5 cm thick broiler breast meat slice was used, whereas only a few differences were found by testing a thicker meat sample (1.0 and 1.5 cm). Also, in the study by Bianchi and Fletcher²⁶, the main effect of the background color was significant only for 1 cm thick samples. The background color did not affect meat color measurements for thicker samples.

Sánchez-Zapata et al.¹³ found that when white paper, yellow plastic, black plastic or aluminium foil were used as background, the pâté sample thickness had no effect on the color measurement.

Table 3:	Influence of the background color on chroma and hue (X-Rite)
C: Chroma, h: Hue, a,bDifferent letters in a row denote statistically significant differences between results (p<0.05)

Table 4:	Correlation coefficient and significance levels (p) between the color coordinates and thickness of cheese (white background)
L*: From black to white, a*: From green to red, b*: From blue to yellow

Table 5:	Correlation coefficient and significance levels (p) between the color coordinates and thickness of cheese samples (black background)
L*: From black to white, a*: From green to red, b*: From blue to yellow

In the case of tuna pâté, sample thickness must be taken into consideration only when white ceramic or white plastic materials are used as background.

American Meat Science Association in meat color measurement guidelines recommend using samples that are at least 12-15 mm thick, stacking wafer-type samples and evaluating them against a white background or stacking product²⁷.

As sample thickness increased from 1.7 mm-13.6 cm for white background, the cheese had lower L* and b*, but higher a*. For black background, the thicker the cheese sample, the lighter the cheese, for X-Rite. For Specto-pen, L* was at a similar level (68.54 vs. 68.10), whereas a* and b* increased their values.

Table 4 presented Pearson’s correlation coefficient between the coordinates and the thickness of the cheese. Except for a* coordinate, all of the measured color coordinates showed strong, negative and statistically significant correlation. Data in Tables 5 presented analogous results obtained on black background. Except for L* parameter, all of the measured color coordinates showed strong, positive and statistically significant correlation (except for b* parameter obtained from Spectro-pen).

When comparing the results of color measurement received from two different pieces of equipment, statistically significant differences were obtained for all thicknesses of the cheese samples, for both white and black backgrounds. Baardseth et al.²⁸ found that the variations from one instrument (for color measurement) to another are systematic and can be described by linear regression. The regression can be used to compare color values expected from one instrument with those obtained from another.

The results obtained in this research for the tested cheese coincided with earlier scientific contributions, even though these concerned different materials. Researchers confirmed that the colour of surface, sample thickness and the type of equipment used for color measurement affect color discriminants. These results are important in evaluating and reporting cheese color.

CONCLUSION

The research showed that a minimum sample thickness of 11.9 mm can be considered satisfactory for the analyzed Gouda cheese, which can eliminate the influence of the background color. Considering different cheese composition and texture, in laboratory practice, it is recommended to take a minimum 2 cm thick sample for the evaluation of cheese color.

SIGNIFICANCE STATEMENT

This research revealed the need for using minimum cheese thickness for marking its color with an aim of eliminating background color impact on the results. The results of the research are important contributions and can be used both in scientific studies as well as in laboratory practice. This study will aid dairy researchers and controllers in obtaining objective and comparable results for cheese colors.