The bread was made using a formulation described previously [1, 15]. Briefly, bread made using 250 g of commercial wheat flour (Camellia; Nisshin Foods, Manufacturer No. 2623798, Lot No. D, Tokyo, Japan) as the basic ingredient (100%) consisted of butter (4%), sugar (6.8%), salt (2%), water (72%), and dried yeast (1.12%) [1]. To avoid variation in individual kneading styles, the bread was made using an automatic bread-making machine (SD-RBM; Panasonic, Osaka, Japan) set to mode number 11 [1]. At 2 h after baking [15, 16] (to allow the bread to cool to room temperature), a 25-mm-thick slice of bread was cut using a commercial breadknife (AC-0059; Kaijirushi, Tokyo, Japan).

The same crumb was illuminated with five different light sources (F, TL₈₃, TL₈₄, CWF, D₆₅) in a color assessment cabinet (P60; Huanyu, Shenzhen, China) (5,400 mm × 7,100 mm × 5,500 mm) [1]. A UV light source was added, and the visual palatability and visual firmness of the crumb were compared under different illuminations. The UV light source emits purple light and a faint blue glow, although the main wavelength is not in the visible light spectrum. The color temperatures (K) of the five light sources were: F, 2,731; TL₈₃, 3,056; TL₈₄, 4,057; CWF, 4,126; and D₆₅, 6,529 [1]. The illumination intensities (lx) were: F, 612; TL₈₃, 1,771; TL₈₄, 2,395; CWF, 1,221; and D₆₅, 2,575 [1].

A commercial digital camera with a GR lens (focal length = 18.3 mm, equivalent focal length approximately 28–35 mm; GRIII; RICOH, Tokyo, Japan) was used to generate approximately 24.24 million-pixel images as dng files [1]. The identity of the white bread photographs distributed to participants was confirmed statistically.

Sensory evaluation of the visual palatability and visual firmness of the crumb of white bread was performed using printed images of the crumb of bread baked with the typical formulation outlined above [1]. The visual palatability and visual firmness of the crumb were assessed using images in dng format under six illuminations, including UV light. The images of the crumb of the bread were printed using a commercial laser printer (LBP-841C; Canon, Tokyo, Japan). Twenty-seven participants with a mean ± standard deviation age of 20.2 ± 0.4 years were recruited from among students at Hirosaki University. Palatability and firmness were scored on a seven-point scale (−3, −2, −1, 0, +1, +2, +3) [1, 15] using an image of white bread taken under D₆₅ illumination as a reference [1]. Higher scores indicated greater palatability or greater visual firmness. Based on randomization principles, only untrained participants were recruited for the visual sensory evaluations [15, 17]. The arithmetic means of the visual palatability and visual firmness scores were calculated. The experimental procedure was explained thoroughly to all participants, and they all provided written informed consent.

This study was performed to enhance the reliability of evaluations of the visual palatability of bread crumb images by using duplicate tests.

A method for sensory evaluation of the visual palatability and visual firmness of white bread was developed using images presented on a screen with a projector (EB-W06; Seiko Epson, Suwa, Japan). This measurement method was based on a procedure using printed images of the crumb of the bread taken under various illumination light sources, including UV light, but instead used images projected on a screen. The participants assessed images of white bread crumb presented simultaneously on the screen. The image slide consisted of an image of the crumb of white bread photographed under illumination with the D₆₅ light source on the upper side and the “experimental crumb image” on the lower side. All participants assessed visual palatability and visual firmness using the seven-point scale outlined above. This procedure was also explained to all participants, and they all provided written informed consent.

The reliability of the evaluations of the visual palatability of bread crumb using images presented on a screen was confirmed using duplicate tests.

One-way analysis of variance (ANOVA) with Tukey’s post hoc test was conducted using KaleidaGraph (ver. 5.04, Synergy Software, Reading, PA, USA) to compare the scores among the bread samples photographed under various illumination conditions. In all analyses, P < 0.05 was taken to indicate statistical significance.

Due to the lack of color temperature data for the UV light source, only the effects of the other five light sources on visual palatability are plotted in Figure 1. Sato [1] investigated the correlation between color temperature and apparent palatability scores, and reported a linear regression line with a coefficient of determination (R²) of 0.730, while R² for the 27 participants in this study was 0.764. Compared with previous results [1], only two plot positions—between CWF and TL₈₄—were changed. Visual palatability scores were lower under the F light source than under the TL₈₃ light source. The highest palatability score was for the crumb of bread photographed under TL₈₃, and the lowest was for the bread crumb photographed under D₆₅; this result was reproduced in at least two experiments.

To analyze factors affecting the visual palatability of white bread crumb, the correlation between the palatability and visual firmness of the crumb under illumination with various light sources, including UV light as a blue and purple color light, are shown in Figure 2. The linear regression line had an R² value of 0.865, indicating high reproducibility of the relation between visual palatability and visual firmness even with the addition of scores for sensory attributes assessed under the UV light source. A linear relation between the visual palatability and visual firmness of the crumb of white bread was seen (R² = 0.813) on the omission of the visual sensory scores under the UV light source [1]. As stated above, visual sensory scores were obtained for the crumb illuminated under illumination with six light sources, including UV light, and the number of participants was increased to 27 untrained students in the retest. Indeed, a better relation between these sensory attribute scores was obtained with the addition of the scores obtained under the UV light source. The participants selected the softer bread even when looking only at the surface of the crumb.

The visual palatability scores of the crumb for the 13 participants in the first test were consistent with the sensory results of the 27 participants in the retest (Figure 3). The linear regression line had an R² value of 0.954.

The palatability scores for printed photographs corresponded to those of images presented on a screen using a projector, and the linear regression line had an R² value of 0.905 (Figure 4). The correlation between visual palatability and visual firmness had an R² value of 0.813 when the UV light source was excluded [1], but R² = 0.917 when using a projector and with the addition of the UV light source as a blue and purple light source (Figure 5).

To ensure that the photographs distributed to participants were identical, a color analysis of five photographs selected at random was conducted. Copies of photographs of the crumb taken under the D₆₅ light source were compared after printing, and the L^*a^*b^* space values were analyzed using a color reader (CR-13; Konica Minolta, Tokyo, Japan) (Table 1). Although there were no significant differences in L^* and a^*, the b^* values of the crumb photographs were significantly different for the second image. A retest of this color analysis for the second image yielded the same result. The lower b^* values in the second image may have been due to insufficient yellow ink. Practically, a difference of 1–2 points in scores reflects almost the same visual assessment. Based on general sensory evaluation theory, food samples distributed to participants should be identical, with the same qualities, as stated above. Thus, the identity of the white bread photographs distributed to five participants was confirmed in this study. In addition, the sensory evaluation scores for the photograph taken under all light sources were acquired in a blinded manner. Only the image under the D₆₅ light source was compared to itself. Therefore, a photograph of the crumb acquired under the D₆₅ light source may have a visual sensory palatability score of 0, even if taken in a blinded manner. The plots of the crumb photographed under the D₆₅ light source showed sensory palatability scores of around 0 (Figures 1–3), indicating the reliability of this sensory evaluation method. The sensory evaluation scores—i.e., the palatability and visual firmness scores—may vary within about ± 1 points, and the variation in color among printed images might influence the sensory evaluation results.