November 12, 2022 / Editor

Colour Perception & Colour Attributes


By definition, colour can be described as the visual attribute of an object that primarily results from the way a particular object emits or reflect light. Literature reports that the human eye has the ability to distinguish nearly up to ten million colours (Leong, 2006). When it comes to the appearance of a colour, three essential attributes come into play, and these are hue, chroma/saturation, and lastly, value which is also known as lightness. Hue is the quality of colour as determined by its dominant wavelength, while saturation, on the other hand, depicts the vividness of colour (Kuehni, 2003). Lightness remains to be highly associated with the degree or variation of brightness. Colour attributes have been reported to play a crucial role when it comes to colour perception. Understanding how these attributes influence colour perception is thus essential in understanding how human beings make colour decisions. Based on this background, this context thus ventured into literature to investigate and discuss the most important colour attribute in colour perception and colour decisions.


Literature indicates that colour-related stimuli is processed by the occipital lobe in the brain (Racey et al., 2019). Even though it is also responsible for the overall visual perception processing, it can also be noted that this processing heavily depends on the attributes of the light as received by eye receptors. Even though light has a number of attributes such as form, colour, direction, and so on, this context was interested in colour. Colour itself has its own attributes which were named above as hue, chroma, and value. This context was thus heavily interested to know just how these attributes of colour influence colour perception as processed in the occipital lobe. To answer this question, a number of studies were thus provided and thoroughly investigated for insights related to this research question.


Interested to understand the ability of individuals to distinguish colour attributes, thirty-six glossy pair samples from the Munsell atlas were observed in two experiments. In the first experiment, samples of each pair to be distinguished were made to differ in one of the colour attributes, while in the second experiment, samples involved an identical colour attribute (Melgosa et al., 2000). All 36-pair samples were distributed into four groups in each experiment, where each group had nine pair samples (Melgosa et al., 2000).

  • In this first experiment, three pair samples varied only in value, three other pair samples varied only in chroma, and lastly, three other pair samples varied only in hue. 
  • In the second experiment, the research was not interested in the ability of the participants to identify which samples had a varying colour attribute but rather an identical one (Melgosa et al., 2000). As a result, three samples pair had the same value, three other sample pairs had the same chroma, and lastly, three other sample pairs had the same hue. Participants in this study were simply asked to observe and answer which colour attribute was different in the first experiment one, and which colour attribute was identical in the second experiment. 

It is important to note that participants in this study were not informed about the design of the study, hence were not aware which colour attributed was changing or identical in the subsequent experiment (Melgosa et al., 2000).

Based on the nature of this study, there was no time limit, and even though all nine-sample pairs in a group were simultaneously displayed, the observers judged each sample pair one at a time. It is also worth mentioning that the participants of this study consisted of a total of forty observers who had been determined to have normal colour vision. To also understand whether having prior knowledge of colorimetry would influence their ability to identify the tested colour attributes, half of the observers in these experiments were experienced and had extensive knowledge of colorimetry fundamentals (Melgosa et al., 2000).

The results of this study indicated that the overall ability of participants to distinguish colour attributes was low. In the first experiment, participants were only able to identify 60.2 % of the sample pairs whose attributes of interest differed (Melgosa et al., 2000). In the second experiment, on the other hand, participants were only able to identify 50.6% of the pairs that shared an identical attribute (Melgosa et al., 2000). When it comes to the comparison of the three attributes, chroma was reported to be the least successfully identified attribute in both experiments. Hue, on the other hand, was the most successfully and easily identified attribute in both experiments.

Similar to this study, Kuehni (2003) also acknowledged that it was easier for an individual to identify a differing attribute between two colours rather than when that attribute was shared. Despite the fact that hue was reported to be the most important attribute in colour perception in this study, it was in value that small differences were easily identified. In hue, only large differences were easily identified when compared to smaller differences (Melgosa et al., 2000).

Zhang & Montag (2006) took Melgosa’s experiment a notch higher with the aim of establishing how well people could use different colour attributes. This study, also had two experiments, and in the first experiment, the main outcome being assessed was how fast and accurate a participant/observer could use the three different sets of colour controls (RGB; lightness(L), Chroma(C), Hue(H); and Lightness(L), redness/greenness(r/g), yellowness/blueness(y/b)) to match the colour of colour patches displayed on a screen (Zhang & Montag, 2006). The second experiment in this study took a different approach whereby participants were presented with colour pairs and were required to judge the similarity or difference of these pairs using the LCH control method or the L, r/g, y/b control method named above.

In the actual experiment, two patches in the first experiment were displayed on the centre of a computer screen while separated by a distance of about .5cm. Using three sliders, participants were then required to match the two colour patches to be identical as much as possible. Seventeen out of the twenty-four participants were experienced, while seven were naïve in the first experiment. In the second experiment, eighteen observers were experienced, while thirteen were naïve (Zhang & Montag, 2006).

Results of the first experiment indicated that expert participants/observers had higher accuracy in matching the patch pairs using the LCH sliders than when using those of RGB. For the naïve participants, the use of both LCH and the L, r/g, y/b sliders achieved better matching accuracy when compared to those of the RGB (Zhang & Montag, 2006). The matching accuracy of both naïve and expert participants was not statically different when using the L, r/g, y/b sliders. The average time taken to complete these tasks was also measured, whereby the average time for using RGB was 66s, LCH was 52 s, and the L, r/g, y/b sliders was 57s (Zhang & Montag, 2006).

In the second experiment, results revealed that the performance of judging different and similar attributes was different for every observer. Even so, the identification of different attributes had a higher performance when compared to the identification of common attributes. This had been equally reported by Melgosa in the previous study. Additionally, the LCH sliders were reported to have been significantly easier to use when compared to those of L, r/g and y/b. Judgments with hue were better when compared to those of r/g and y/b, and lastly, experts depicted better judgment performance when compared to naïve observers (Zhang & Montag, 2006).

The implications of these results are enormous when it comes to answering the research question of interest. Similar to the study conducted by Melgosa, (Zhang & Montag, 2006) also supported that in colour perception, hue was the most important colour attribute. This was followed by lightness and chroma came last in both studies. Melgosa et al. (2000), had also acknowledged that colour discrimination was not an innate ability and that familiarity with global colour space could improve performance. The results of Zhang and Montag supported the statement since experienced observers had better colour attribute identification performance.


Colour Decisions

The other essential dimension of this discussion was to understand how colour attributes also influenced colour choices. In the preceding discussion, emphasis was placed on colour perception. This section, however, sought to examine which colour attribute, i.e., hue, chroma, and value, was the most important when making colour decisions. Melgosa et al. (2000) highlighted that hue was the most easily perceived attribute while value was the least perceived in colour perception. Is there a possibility that these results are maintained even when it comes to colour decisions? 

It is important to note that Melgosa et al. (2000) had acknowledged that his study was not purely perceptive and that it included both the aspects of perception and cognition. These aspects arose from the fact that his study utilized a large size of colour differences, whereby cognitive factors and familiarity with the global colour space could have significantly influenced the results presented in the above sections. Even though perception is a sensorial impression, cognition presents a more complex interaction between perception, learning, and reasoning. This is encountered in decision-making.

Literature indicates that blue remains to be one of the most preferred colours, and that lightness and saturation have a correlation to emotional responses (Gyu, 2014). This author also noted that even though studies using colour patches or papers have been fundamental in giving background and direction to research, in reality, colour is experienced through complex patterns that interact with not only people’s perceptions but also behaviour. This author also noted that previous research failed to account for confounding effects such as lightness, saturation, background colours, and light sources. Park thus conducted a study using physical model simulation to investigate colour preference in a real-life context. In this study, Gyu investigated the most preferred colours among children within hue families. He also investigated any existing correlations between those observed colour preferences with colour attributes (Gyu, 2014).

Thirty-three children from four schools in the state of Texas were recruited to participate in the study. Out of the total, thirty participants were boys, while three were girls. These children were between the age of seven years to eleven years hence corresponding to the concrete operational stage in Piaget’s theory of cognitive development. Models with a scale of 1:12 were constructed, where each had a single changeable side wall. A bed and a chair were placed inside the model rooms and all other features, such as light sources, windows size, room size, materials, etc., were identical. A garden picture was also installed to simulate a garden window in all room models. A fluorescent light was used in the interior, while an incandescent light was used to represent an exterior lighting source. Luminance levels were measured using a photometer and maintained during the entire experiment (Gyu, 2014).

In terms of procedure, participants were asked to select their three favourite coloured rooms in order and also their least preferred, coloured rooms in order. In the following session, selected rooms would be covered, and the interchangeable walls changed to other sets of colours within the hue family. This was continued until all five sessions had been completed. Based on the nature of the experiment, there was no time constraint, and children were allowed to spend as much time as they liked before coming up with a decision (Gyu, 2014).

The results of this study revealed that some of the most preferred colours included: red (5R 6/8 and 7/8), yellow (5Y 9/8), green (5G 7/8), blue (5B 6/8), and lastly purple (5P 7/8) (Gyu, 2014). Analysis to examine correlation determined that saturation was the colour attribute highly correlated with the above colour preferences (Gyu, 2014). This was especially the case in red, green, blue, and purple hue families. In yellow preferences, nonetheless, lightness was the most correlated colour attribute. Even though previous studies have revealed that there is little to no differences in colour preferences based on gender, the results of this study revealed that girls preferred red and purple quite more when compared to their male counterparts (Gyu, 2014).

Tangkijviwat et al. (2008) also offered a number of insights into this matter. Similarly to Gyu (2014), this author stated that colour in day-to-day life cannot be perceived through the lens of just three-colour attributes. Tangkijviwat et al. (2008) noted that other modes, such as the unnatural object colour and light source colour modes, can also be taken into account when it comes to real-life colour perception. In this experiment, thirty-three colour chips from the Munsell notation were presented in different modes by adjusting the room illuminance and the colour chip room illuminance. Overall, results indicated that the colours which were most preferred were those that were brightest and highly saturated.  

This, in a way, supported the findings by Gyu (2014), that revealed saturation was the most highly correlated colour attribute in colour preferences. In his subsequent experiment that investigated perceived chromaticness, whiteness, and blackness, results showed all the above-named attributes played a role in determining the colour preference in different colour modes. Tangkijviwat et al. (2008) hence concluded that colour attributes cannot be the only dominant factor in the determination of colour preferences since the mode of colour appearance also matters.

Preliminary Assumptions

This context integrated a number of assumptions prior to an in-depth review of literature. Take, for instance, when describing colour appearance, it was noted that there is more to colour perception than just hue, saturation, and value. Other critical factors such as nuances, tones, tint, lightness, and shades also play a role in reality. This thus called for a wider approach to colour attributes in future studies.


Additionally, unlike the scales of colourfulness, i.e., saturation and lightness, hue is a qualitative perceptual dimension, thus, does not lend itself well when it comes to expressing changes in magnitude. Even though some studies assumed that colour discrimination was an innate ability, it was found to share both perceptive and cognitive aspects and, as a result, was influenced to a high degree by familiarity with colour. 

Most studies also assumed that there is little variation in colour perception between individual observers. This is usually not the case since there is a considerable variation in colour perception by individual observers (Kuehni, 2003). Finally, Kuehni (2003) also noted that individuals cannot operate as neutral colour measuring instruments since each conscious decision passes through some as yet unknown kinds of individual mental filters before judgment is delivered.


To sum up this context, hue was evaluated to be the most successfully and easily identified attribute, while chroma was the least in colour perception. In terms of colour decisions, findings revealed that chroma/saturation was the most important colour attribute. Regardless, studies had limitations, take, for instance, colour papers do not take into account the confounding effect of colour modes as influenced by factors such as background colours and light sources. Additionally, colour decisions were noted to be an outcome of cognition where each decision passes through some complex yet unknown kinds of individual mental filters before judgment is delivered. As a result, individuals cannot be used as absolute colour measuring instruments.



Gyu “Phillip” Park, J. (2014). Correlations between color attributes and children’s color preferences. Color Research & Application39(5), 452-462.

Kuehni, R. G. (2003). Color ordered: a survey of color systems from antiquity to the present. John Wiley & Sons.

Leong, J. (2006). Number of colors distinguishable by the human eye. The Physics Factbook—An Encyclopedia of Scientific Essays [online].

Melgosa, M., Rivas, M. J., Hita, E., & Viénot, F. (2000). Are we able to distinguish color attributes? Color Research & Application: Endorsed by Inter‐Society Color Council, The Colour Group (Great Britain), Canadian Society for Color, Color Science Association of Japan, Dutch Society for the Study of Color, The Swedish Colour Centre Foundation, Colour Society of Australia, Centre Français de la Couleur25(5), 356-367.

Racey, C., Franklin, A., & Bird, C. M. (2019). The processing of color preference in the brain. Neuroimage191, 529-536.

Tangkijviwat, U., Rattanakasamsuk, K., & Shinoda, H. (2008). Color preference affected by mode of color appearance. Color Research & Application: Endorsed by Inter‐Society Color Council, The Colour Group (Great Britain), Canadian Society for Color, Color Science Association of Japan, Dutch Society for the Study of Color, The Swedish Colour Centre Foundation, Colour Society of Australia, Centre Français de la Couleur35(1), 50-61. Zhang, H., & Montag, E. D. (2006). How well can people use different color attributes? Color Research & Application: Endorsed by Inter‐Society Color Council, The Colour Group (Great Britain), Canadian Society for Color, Color Science Association of Japan, Dutch Society for the Study of Color, The Swedish Colour Centre Foundation, Colour Society of Australia, Centre Français de la Couleur31(6