Color

Color comprises three attributes:

  • Hue is the attribute of color that distinguishes red from green, purple from yellow, etc.
  • Saturation is the richness of color, and is determined by the extent to which it is chromatic (i.e. comprises essentially hues such as purple, red and yellow) rather than achromatic (i.e. comprising black, white or grey). Mixture of the achromatic reduces the saturation. Zero saturation is completely grey.
  • Lightness is the brightness of light intensity in the scene.

While hue was believed to drive color preference, a study by Jacobs & Hustmyer (1974) showed that color saturation was more important than hue.

The perception of color and color preferences is found here.

Studies of color by Dr A.Lothian

In Lothian, 2000, the color of scenes was examined by classifying within each scene, the colors for each type of feature (Table 1) For example, the colors registered for dry pastures and straw crops were grouped and listed next to their scenes in which these features occur.

Table 1 Frequency of colors in scenes
Table 1

Note: The number of colors discriminated by the computer numbers tens of thousands. The colors summarized here are simplified into the major colors, e.g. red, blue.

 

SA 6Scene with many colors

SA 7Scene with few colors

Typically, the scenes comprised the ground color (either straw or green depending on season), the greens of trees and shrubs or of other vegetation, a light blue of distant vegetation, and the blues of the sky. Scenes near the coast or river or with other water features include the color of water. Scenes in mountains, on the coast or along the River Murray may include orange or straw-colored rockfaces.

spectrumFigure 1 Color spectrum

An analysis was carried out of the colors in the scenes using a scanner which numbered the hues (Table 2). The resultant mean hues are summarized in Figure 2. This shows that the colors did not cover the full range – there were no blues, indigo-violet or pink hues (i.e. >140).

Table 2 Color Spectrum Chart

HuesNote: These are the colors in the scenes using a scanner which numbered the hues.

Fig 2Lothian, 2000          Figure 2 Spectrum scale of major hues for each feature

Figure 2 indicates the following:

  • The arid dunes (hue 28.7) were yellow but closer to the red-orange end of the spectrum than beach sand and dunes (hue 45), which were far more yellow.
  • Sea scenes (hue 128) were light blue but bluer than other water bodies (Murray Valley, dams, mound springs) which at 110 were rated as green.
  • There is a marked contrast between the straw and green colored crops and pastures – respectively hues 30 and 77. A similar contrast exists between the straw dry ground (hue 31) and green ground (hue 75).
  • Hills and vegetation in the distant (respectively hues 108.5 and 108) were virtually identical. Vegetation in the foreground and middle ground of a scene (hue 91) became more distinctly bluer in the distance (hue 108), a characteristic of the Australian landscape.

Participants evaluated the scenes for their dominant hue and the author checked to ensure that un-nominated colors – brown, white and grey, were considered.

The number of scenes represented by each hue is summarized by Table 2. The green and yellow hues were the most dominant, together accounting for 63% of the total. This partly reflects the large proportion of agricultural scenes, specifically the cereal growing scenes of straw or green colors. Figure 3 show the average ratings for scenes on the basis of their dominant colors.

Table 2 Frequency and rating of dominant hues

Table 2

Fig 3Lothian, 2000. Note: Scale exaggerated.
Figure 3 Color preferences for South Australian landscapes

The reason for the blue scenes being the most preferred is not hard to find as all but one of the 26 scenes include significant areas of water including of the coast, Murray Valley or inland waters. Sixteen of the scenes were of the coast, which is the most highly rated landscape in South Australia, and a further nine were of inland waters. Thus the question is whether the preference for the blue color reinforces the preference for water environments, i.e. is a lead factor, or whether the content of the scene follows and the color is the more important. Given the evidence, the first hypothesis is the more likely, that the color does not determine the preference but serves to reinforce the preferences for water.

019_19COrange hue dominant

SA 9Green hue dominant

SA 8Yellow hue dominant

SA 10Blue hue dominant

Of the 13 scenes of orange, the next most preferred color, six were of arid ranges and a further four of other arid areas (gibber, saltbush, dunes). The high rating of arid ranges is reflected in the high rating of this hue. The relatively low ranking of scenes with dominant red hues is surprising given the usually high rating of red. Six of the eleven scenes were of arid ranges, which were highly rated but the remaining five included gibber and arid plains, which were among the lowest rated scenes.

The low rating of the scenes with yellow and green as the dominant hues reflects their preponderance in scenes of cereals and pasture, which are of flat terrain and generally low rated. The bottom rated set of scenes in which brown is the dominant hue included samphires, vineyard, dairy flats and a waterhole, all low rated. The reason for brown being so disliked may be due to its association with death, e.g. brown crops, dead leaves. The ratings of the scenes in which blue was the dominant color were 2.34 or 47% higher than those with brown scenes, the lowest rated color.

In the study of the Lake District (Lothian, 2013b), the presence of seasonal flowers in fields enhanced landscape preferences by an average of 10% (Figure 4).

Lakes 5 103CScene with yellow flowers rated 5.79

Lakes 6 104CSame scene without flowers rated 4.84  Lothian, 2013b
Figure 4 Influence of the presence of color on landscape preferences, Lake District, UK

Other studies of color

Bishop (1997) studied the use of computer-based color analysis software in examining color in the landscape and found it practical.

In her study of agrarian areas in Western Norway, Strumse (1994) found that scenes with flowers rated highly – one was 4.56 (1- 5 scale) and was the sixth preferred out of 60 scenes. The average was 4.34.

The color of trees could, according to Kaufman & Lohr (2004), indicate nutrient deficiency in the trees by appearing yellow. To test this, they examined whether people have distinct preferences for tree color. Vibrant green can indicate a healthy nutrient-rich plant with high sugar and carbohydrate content, whereas yellow foliage might indicate an unhealthy nutrient deficient plant. On the basis of a finding that macaques monkeys select leaves which correspond to high nutrient levels, and that these monkeys have the same color vision as humans, the researchers asked participants to rate a range of tree colors. The results indicated that all green and red trees, regardless of their hue and saturation were regarded positively while brown and orange-brown hues were regarded negatively (Figure 5).

Fig 4Kaufman & Lohr, 2004    Figure 5 Preference for selected tree canopy colors

Subsequently, Lohr (2007) reported that while all tree colors were calming, that the green of a healthy tree was more calming than the other canopy colors including other greens, reds and yellows. Yellow, which is associated with severe nutrient deficiency, was the least calming.

IMG_8218CThe appeal of colorful flowers

The idea that color preferences derive from signals in nature such as the color of a flower to attract insects or to be cautious near a bright-colored wasp has been current for some time. Palmer & Schloss (2010) tested this and found people’s likes corresponded with liked objects such as blue with clear skies and clean water, whereas disliked colors were also associated with disliked objects such as brown with death and rotten food. They termed this the ecological valence theory of color (see also Palmer et al, 2012).

Color summary

In the author’s study the scenes were scanned and their colors analyzed by a computer program for hue, saturation and lightness. The number of colors per scene averaged 4.5, with a range from two to eight. Colors which were largely absent were the reds and the indigo/violet, at opposite ends of the color spectrum. Analysis was undertaken of the colors of scenic features, e.g. river cliffs, trees & vegetation. Among the findings were that arid dunes were redder than the yellow/white coastal dunes, sea scenes were light blue, and vegetation in the distance took on a blue tinge. Analysis of color preferences was based on an assessment of the dominant color in each scene. This found that the colors were preferred in the following order of preference: blue, orange, indigo/violet, red, green, grey, yellow, and lastly brown. These vary quite markedly from the hues found in the theme: The perception of color, to be favored (in descending order): blue, red, green, violet, orange, yellow.

Other studies have found that color influences landscape preferences both positively and negatively depending on the hues and saturation of the color. Research has shown that green and red are the most preferred. Such colors may be indicative environmental cues of favorable food while the South Australian preferences may reflect the colors of water and the outback.