*3.4. ColorScent*

As seen in previous studies, the amount of color that can be expressed is very limited, because the intensity of fragrance perception is poor. Scent acts as a good trigger for memory and emotion, because it can mediate the exploration of works of art in terms of general memory or emotion. Because smells and memories are connected in the brain, memories can be recalled by smells, and smells can sometimes be evoked by memories.

De Valk et al. [107] conducted a study of odor–color associations in three distinct cultures: the Maniq, Thai, and Dutch. These groups represent a spectrum in terms of how important olfaction is in the culture and language. For example, the Maniq and Thai have elaborate vocabularies of abstract smell terms, whereas the Dutch have a relatively impoverished language for olfaction that often refers to the source of an odor instead of the scent itself (e.g., it smells like banana). Participants were tested with a range of odors and asked to associate each with a color. They also found that across cultures, when participants used source-based terms (i.e., words naming odor objects, such as "banana"), their color choices reflected the color of the source more often than when they used abstract smell terms such as "musty". This suggests that language plays an important mediating role in odor–color associations [107].

Gilbert et al. [108] confirmed that humans have a mechanism that unconsciously associates specific scents with specific colors. For example, aldehyde C-16 and methyl anthranilate are pink; bergamot oil is yellow; caramel lactone and star anise oil are brown; cinnamic aldehyde is red; and civet artificial, 2-ethyl fenchol, galbanum oil, lavender oil, neroli oil, olibanum oil, and pine oil are reminiscent of green. Repeated experiments that produced similar results demonstrated that those results were not random. Also, of the 13 scents, civet was rated as the darkest, and bergamot oil, aldehyde c-16, and cinnamic aldehyde were rated as the lightest [108].

Kemp et al. [109] found that the color lightness of a color was perceived to correlate with the density of a fragrance and that strong scents are associated with dark colors.

Li et al. [110] developed the ColorOdor, an interactive device that helps the visually impaired identify colors. In this method, the camera attached to the glasses worn by the user recognizes the color, and the Arduino controls the piezoelectric transducer system through Bluetooth to vaporize the liquid scent associated with the color. Although culture plays a role in color–odor connection, user research showed color–odor mappings are (white, lily), (black, ink), (red, rose), (yellow, lemon), (green, camphor leaves), (blue, blueberry), and (purple, lavender). When a blind person touches the "white" part of the picture with a finger, it sends a signal to the fragrance generator so that "lily" is emitted. The visually impaired who knows that white and lily scent are related can know that the part is white through lily scent. However, this study was not intended to allow the visually impaired to appreciate works of art, and there is a limitation that the association of used fragrance and color was not based on scientific experiments and results, but mostly due to subjective selection of researchers. Nevertheless, the attempts of those who used scent to convey color to the visually impaired offers us many implications [110].

Lee et al. [89] assumed that each scent has its own unconscious relationship with color and concept, which the researchers called color directivity and concept directivity, respectively. Through experiments, they found specific scents with color directivity and concept directivity and then used those scents to successfully deliver information about the colors used in artworks to the visually impaired. Another study on the transmission of color information using scent found a scent with consistent color and concept orientation and applied it to tactile paper, allowing the visually impaired to actively explore and be immersed in the form and color of an artwork. Instead of understanding art appreciation as an educational technique that unilaterally conveys the authority and interpretation of third parties, such projects invite the visually impaired to directly experience artwork through their own senses. In this case, a special ink scent was applied to the surface of swell paper. After visually impaired students were exposed to the work, their degree of comprehension was measured in terms of the accuracy of their scent–color recognition, a usability evaluation, and an impression interview. In [89], scent is released when a user rubs their finger over the area where the scent was applied to the painting in Tactile Color Book. Unlike people with congenital blindness, people with an acquired visual impairment retain the concept of color. When they smell something, they naturally associate it with color through the memories associated with it. Sight and hearing do not have the same powerful recall ability as smell. For the prototype, the scents of menthol, orange, and pine were chosen to express blue, orange, and green, and the scents of rose, lemon, grape, and chocolate were chosen to express red, yellow, purple and brown, respectively. When smelling a particular scent, sensitivity decreases by 2.5% every second, with 70% disappearing within one minute. However, even under adaptive fatigue conditions, other odors can be identified. On the paintings, the scents were arranged at intervals of 3 cm or more to prevent mixing. The paintings in Tactile Color Book use the same perfumes used in aromatherapy. In their experiments, color-concept directivities were found for the scent of oranges (orange color), chocolate (dark brown), mint (blue), and pine (green). Orange scent shows high saturation orange color directivity, and it shows concept directivity of bright, extroverted, and strong stimulus. Chocolate scent showed brownish color directivity with low color lightness, and showed concept directivity of roundness, lowness, warmth, and introversion. Pine scent and menthol scent appeared in turquoise color in terms of color directivity, but concept directivity menthol was found to have a greater association with the concept of coolness. The rest of the scents used in the experiment did not show distinctly significant and consistent characteristics in color directivity and concept directivity. Based on the results of the experiment, the scent of orange is associated with orange, menthol is blue, pine is green, and chocolate is brown colors, respectively, as shown in Table 5. Orange has been shown to be bright, extroverted, and associated with strong stimuli (angled form, high notes). This property is also generally applicable when describing the characteristics of yellow and red. Considering that orange is a blend of these two colors, one can give orange a universality that encompasses all three colors. Menthol and pine were similar in color directivity; the concept of coolness was considered to be about 22% higher in menthol than pine. Therefore, menthol was designated as blue, the color associated with coolness, and pine was designated as green, respectively. Using this scent–color association, the scent coding color is used in tactile textbooks for the visually impaired to appreciate artworks. The test result showed a high color recognition rate, and a positive result was obtained that induces subjective immersion in the artwork appreciation experience of the visually impaired. Among visually impaired students, the color recognition accuracy was 94.3%, and the usability evaluation score averaged 70 points. In the interviews, students said that the system was intuitive and easy to learn. In addition, the students said that the scents allowed them to better understand the content of the artwork. In the other scents, except for the above four scents, rose, phoenix, apricot, strawberry, lemon, and apple, as a result of the experiment, no significant color orientation or conceptual orientation could be observed visually or numerically [89].


**Table 5.** Color directivity, concept directivity, and color matching of each scent [89].

**Table 5.** *Cont.*

Nehmé et al. [111] investigated odor–color associations and based on the three tested populations like French–Lebanese–Taiwanese, odor–color associations could be affected by the function of odors in different countries. Culture induced experiences influence the perception of odors familiarity, which will affect the prevalence of either perceptive (intensity, irritancy, and hedonics) or semantic processing of these associations. According to Stevenson et al. [112], color brightness correlates with perceptual attributes of odors (odors that are more irritating, intense, and unpleasant are associated with brighter colors) and semantic attributes (more familiar and identifiable odors are associated with more saturated colors).

Maric et al. [113] presented their French participants with 16 odorants including caramel, cucumber, lavender, lemon, lime, mint chlorophyll, mirabelle plum (at low and high intensity), orange blossom, peppermint, rose, pineapple, shallot, smoked, violet, and wild strawberry. Two pairs of similar odorants (namely lemon/lime and peppermint/mint chlorophyll) were included in the hope of teasing out the finer nuances of odor–color matching. The participants had to pick one of 24 color patches varying in terms of their hue, saturation, and brightness for each odor. They also rated the odorants on 11-point scales in terms of their intensity, pleasantness, familiarity, and edibility. Significantly non-random color matches (with two to six of the colors) were reported for all of the odorants. Of interest, subsequent data analysis highlighted a significant positive relationship between the rated pleasantness of the odorants and both the lightness and saturation (chroma) of the color chosen, with more pleasant odors matched to colors with higher lightness and saturation. They [113] were chosen to cover of the four basic groups: flower (floral odors), sweet (fruits and candies odors), bad (smoked odors), and nature (plant odors). Fifteen food and floral natural aromas were selected as olfactory stimuli: lavender, orange blossom, rose, and violet as floral odors; caramel, mirabelle plum, pineapple, and wild strawberry as sweet odors; smoked as bad odor; cucumber, lemon, lime, mint chlorophyll, peppermint, and shallot as nature odors. Fourteen odors had the same aromatic intensity (fixed by our supplier). The same odor of mirabelle plum (a small yellow plum, specialty of the French region of Lorraine) was presented twice; that is to say with two different aromatic intensities (at low and high intensity). Sixteen olfactory stimuli were thus prepared by injecting 1 mL of each odorant into a small piece of carded cotton that was previously placed into a small opaque glass bottle. No salient visual cues were therefore available to participants [113].

Kim [114] showed that compared with the oriental and fresh families, the floral and woody families showed more distinguishable opposite patterns in both hue and tone parameters: the floral family with brighter warm colors, and the woody family with darker (or stronger) cool colors. The warm colors strongly evoked the floral family, while the cool colors did the fresh family. The brighter (darker) their lightness values become, the more the floral (woody) scents are associated.

Adams [115] showed that the odorants were assigned significantly different values on these line scales. Therefore, for example, the brightest and lightest odorants were lemon, apple, and peach, whereas the dimmest and darkest odorants were coffee, cinnamon, and chocolate (making one wonder whether these mappings may have been driven by the color of the source objects). As has been documented previously, robust shape associations were also documented in response to the odorants.

Russian composer Scriabin talked about fragrances synchronized with the lighting score that he had designed for his tone poem Prometheus: Poem of Fire [116].
