Ophthal. Physiol. Opt. 2010 30: 611–617
Material and lighting hues of object colour
Rumi Tokunaga and Alexander D. Logvinenko
Department of Vision Sciences, Glasgow Caledonian University, City Campus, Cowcaddens Road,
Glasgow, G4 0BA, UK
Abstract
Observers can easily differentiate between a pigmented stain and the white surface that it lies on.
The same applies for a colour shadow cast upon the same surface. Although the difference between
these two kinds of colour appearance (referred to as material and lighting hues) is self-evident even
for inexperienced observers, it is not one that has been captured by any colour appearance model
thus far. We report here on an experiment supplying evidence for the dissociation of these two types
of hue in the perceptual space. The stimulus display consisted of two identical sets of Munsell papers
illuminated independently by yellow, neutral, and blue lights. Dissimilarities between all the paper/
light pairs were ranked by five trichromatic observers, and then analysed by using non-metric
multidimensional scaling (MDS). In the MDS output configuration, the Munsell papers lit by the same
light made a closed configuration retaining the same order as in the Munsell book. The paper
configurations for the yellow and blue lights were displaced transversally and in parallel to each other,
with that of the neutral light located in between. The direction of the shift is interpreted as the yellowblue lighting dimension. We show that the yellow-blue lighting dimension cannot be reduced to that of
the reflected light.
Keywords: chromatic illumination, dissimilarity judgements, human colour vision, multidimensional
scaling, Munsell papers, object colour
Introduction
The colour of an evenly illuminated object can be
described in terms of three dimensions. While the
terminology varies, these are usually named as hue,
chroma, and lightness (Evans, 1974; Wyszecki and
Stiles, 1982; Brainard, 2003; Foster, 2008). However,
three dimensions are enough to describe the object
colours only when illumination is constant and homogeneous. We refer to these three dimensions as the
material dimensions of object colour (Tokunaga and
Logvinenko, 2010). When illumination varies across the
scene, one needs more dimensions to describe the object
colour appearance (Hunt, 1977; Fairchild, 2005). We
refer to the dimensions of object colour which mainly
Received: 5 September 2009
Revised form: 15 January 2010
Accepted: 24 January 2010
Correspondence and reprint requests to: Rumi Tokunaga.
Tel. & Fax: +81 (0)22 217 5469.
E-mail address: rumi.tokunaga@riec.tohoku.ac.jp
correlate with illumination variation as lighting dimensions of object colour (Tokunaga and Logvinenko, 2010).
The existence of lighting dimensions is supported by the
fact that observers can readily distinguish a change in a
scene produced by a material change from that made by
an illumination change (Craven and Foster, 1992;
Foster et al., 2001; Kingdom, 2008). Actually, it is a
matter of everyday experience that we are capable of
distinguishing coloured shadows from pigmented areas
(e.g., stains). The question is whether this ability is due
to cognitive inferring, or is mediated by additional
dimensions of the object colour manifold. Some evidence obtained when studying the colour transformations induced by pseudoscopic depth reversal, testifies in
favour of the latter (Logvinenko, 2009a).
Logvinenko and Maloney (2006) showed that for
achromatic objects an achromatic lighting dimension
emerged when the light intensity varied. Specifically, as
anticipated by Katz (1935) nearly a century ago, a single
lightness dimension was proven not to be enough to
describe the colour appearance of achromatic papers
illuminated by an achromatic light source of a different
intensity. Using multidimensional scaling, achromatic
ª 2010 The Authors, Ophthalmic and Physiological Optics ª 2010 The College of Optometrists
doi: 10.1111/j.1475-1313.2010.00733.x
612
Ophthal. Physiol. Opt. 2010 30: No. 5
colours were found to constitute a two-dimensional
manifold with the dimensions Ôsurface-lightnessÕ and
Ôsurface-brightnessÕ: relating to difference in surface
albedo and to difference in illumination intensity respectively. Surface-brightness was shown to be independent
from brightness of reflected light (Logvinenko, 2005;
Logvinenko and Maloney, 2006). The existence of the
surface-brightness dimension was confirmed in the
chromatic domain as well (Tokunaga et al., 2008).
Moreover, it was recently shown that an additional
chromatic lighting dimension emerged when the illumination chromaticity varied (Tokunaga and Logvinenko,
2010). In fact, this study was an extension of the work of
Logvinenko and Maloney (2006) to the chromatic
domain. The major difference between the two studies
was that Tokunaga and Logvinenko (2010) varied both
papers and lights in the yellow-blue chromatic dimension, whereas Logvinenko and Maloney (2006) used the
achromatic dimension. The results obtained by Tokunaga and Logvinenko (2010) were similar to those
obtained by Logvinenko and Maloney (2006).
Specifically, the Munsell papers lit by three different
lights were represented as three slightly curved contours
approximately parallel to each other in the dissimilarity
space.
Figure 1. Experimental set-up and a stimulus display. A digital
projector (DP) provided independent illumination of the six fields of
the stimulus display (SD). Seven Munsell papers were presented
against a white background covered by a black random-dot design,
inside fields #1 and #2. The fields #3 and #4 were used to present a
pair of Munsell papers of standard dissimilarity (100). The fields #5
and #6 were illuminated in such a way that there were always two
fields illuminated by each of three lights. A computer (PC) randomly
flashed on a pair of light emitting diodes (one in the field #1, and one
in the field #2) indicating the pair of Munsell papers to be evaluated.
An observer entered the dissimilarity rank for the pair pressing a
button on the response box (RB).
It must be noted that in both previous studies the
chromaticity of the papers was equal to that of the
lights. Also in each study only two hues were engaged:
achromatic hues in Logvinenko and Maloney (2006),
and yellow and blue hues in Tokunaga and Logvinenko
(2010). Here we report on an experiment in which, along
with yellow and blue, Munsell papers of various other
hues were used. The rationale was to make sure that
evidence for the existence of the chromatic lighting
dimension of object colour could also be obtained whilst
using more than two hues.
Methods
The experimental set-up was similar to that used in our
previous experiments (Tokunaga et al., 2008; Tokunaga and Logvinenko, 2010) as shown in Figure 1. The
stimulus display was divided into six equal rectangular
fields covered by white paper with a random-dot
pattern. Illumination of each field was adjusted independently by using a digital projector (MT1050; NEC
Display Solutions Ltd, Tokyo, Japan). Two fields (#1
and #2) were used to present stimulus papers: Munsell
papers of various hues and maximal (for each hue)
chroma (5R4/14, 5YR7/12, 5Y8/12, 5G6/10, 10BG5/8,
5PB5/12 and 10P5/12). The angular size of each paper
was approximately 4. Two identical sets of these seven
papers were simultaneously displayed in each field.
Figure 2. The stimuli presented in the CIE 1976 uniform chromaticity diagram. The symbol shape depicts the Munsell paper:
circle = 5R4/14, triangle (up) = 5YR7/12, triangle (right pointing) = 5Y8/12, square = 5G6/10, star = 10BG5/8, triangle (down) =
5PB5/12, diamond = 10P5/12. The colour of the lines indicates the
colour of illumination: yellow, neutral (black line), and blue.
ª 2010 The Authors, Ophthalmic and Physiological Optics ª 2010 The College of Optometrists
Material and lighting hues: R. Tokunaga and A. D. Logvinenko
Illumination of each field was equi-illuminant (60 lux),
and its chromaticity independently varied from session
to session at three levels: neutral, yellow, and blue.
Their CIE 1931 x,y-chromaticity coordinates were
(0.28, 0.34), (0.38, 0.44), and (0.16, 0.08). Figure 2
shows the CIE 1976 uÕvÕ-coordinates of the light
reflected from all the Munsell papers under each of
the three illuminations.
The observerÕs task was to evaluate dissimilarity
between a pair of papers (one from each of two fields).
Red light-emitting diodes (3 mm in diameter) set up next
to each of the papers indicated which pair was to be
assessed. A yellow paper (5Y8/12) under the yellow
illumination and a blue paper (5PB5/12) under the blue
illumination were mounted in two other fields (#3 and
#4) as a standard of dissimilarity (Figure 1). They were
present throughout the experiment within the observerÕs
field of view. The remaining two fields (#5 and #6) were
void of Munsell papers, and were used to balance the
overall illumination in the experimental display. They
were illuminated so that there were always two displays
illuminated by each of three lights (Figure 1). This
measure was taken to reduce chromatic adaptation and
to keep the global adaptation state of the observers as
constant as possible throughout the experiment.
Five normal trichromatic observers took part in the
experiment. All of them except one (the co-author RT)
were naı̈ve as to the purpose of the experiment. They sat
2 m away from the stimulus display, the size of which
was 102 · 93 cm (29 · 26). Viewing was binocular.
The experimental room was semi-darkened, so that
there must have been cues which indicated that these
were real papers lit by real lights.
The experiment was divided into six sessions. In each
session a fixed pair of lights was used. A session
consisted of 49 trials in which all possible 7 · 7 pairs
of papers were evaluated. In each trial a pair of papers
(one in the field #1, and one in the field #2 in Figure 1)
was indicated randomly by the light-emitting diodes.
Observers were instructed to estimate the dissimilarity
(a)
613
between the papers as compared to the standard pair
(fields #3 and #4 in Figure 1) with a number, taking the
standard dissimilarity as 100. Each session was repeated
six times for each observer.
Results
The averaged (across subjects and repetitions) dissimilarities were analysed using a non-metric multidimensional scaling algorithm (Cox and Cox, 2001). Figure 3
presents a three dimensional MDS solution for the
average dissimilarity matrix, that is, such a configuration in the 3D space that the distances between the
points are, in general, in the same order as the
dissimilarities between the corresponding stimulus pairs
(paper/light). The stress for the output configuration in
Figure 3, that is, an index showing the relative proportion of the mismatches between the distances and the
dissimilarities (for a formal definition of stress see e.g.,
Cox and Cox, 2001, p. 64–68) was 0.03. In Figure 3a the
yellow line connects the symbols denoting the seven
Munsell papers illuminated by the yellow light. Likewise, the blue and black lines correspond to the blue and
neutral illuminations. The colour of the symbol indicates the colour of the paper. As one can see, for each of
the illuminations, the seven papers form a closed
contour (referred to as a hue contour), and the order
of papers within each contour is in keeping with the
order of papers in the Munsell book. So it is with the
orders of the circular configurations in the chromaticity
diagram (Figure 2).
There is, however, an essential difference between
Figures 2 and 3. Whereas the three hue contours in
Figure 2 are displaced within the same chromaticity
plane (recall that the stimuli were illuminated by equiilluminant lights), the hue contours in Figure 3a are
transversally shifted. Figure 4 shows the relationship
between the dissimilarity judgements (Figure 3) and the
corresponding distances in the CIE 1976 uniform
chromaticity diagram. The top row indicates that the
(b)
Figure 3. Two different views of the output configuration produced by the non-metric MDS algorithm. Each point represents a Munsell paper
illuminated by a particular light. Notations are the same as in Figure 2. (a) The markers corresponding to the papers lit by the same light are
connected by the lines of the same colour. (b) The markers corresponding to the same paper are connected by lines. See Appendix.
ª 2010 The Authors, Ophthalmic and Physiological Optics ª 2010 The College of Optometrists
614
Ophthal. Physiol. Opt. 2010 30: No. 5
Figure 4. Dissimilarities vs chromaticity differences. In each plate
the vertical axis is the dissimilarity between a pair of Munsell papers
illuminated separately by one of the three lights: yellow (Y), neutral
(N), or blue (B); the horizontal axis is the chromaticity difference
between the reflected lights evaluated in terms of the CIE 1976
uniform chromaticity diagram. Dissimilarities (respectively, chromaticity differences) were normalised by the dissimilarity (respectively,
the chromaticity difference) of the pair (5Y8/12; 5PB5/12) being used
as the standard (see Figure 1), produced under the lighting conditions in question (i.e., B-N, Y-B, etc.) for each observer individually.
The correlation between dissimilarity and chromaticity difference is
significant (p < 0.05) in each panel, R denoting the correlation
coefficient. The bottom row of plots display a selection from the
middle row that consist of pairs of papers which were identical.
Hence, they show how the dissimilarities, produced only by a
difference in illumination, correlate with the corresponding chromaticity differences.
distances along each contour in Figure 3 are more or less
in line with the chromaticity differences between the
corresponding colour stimuli. However, as we can see in
the middle row in Figure 4, the distances between the
various points of different hue contours in Figure 3a
correlate rather poorly with the chromaticity differences
between the corresponding colour stimuli.
Consider, for example, the yellow paper 5Y8/12 under
the blue light and the blue paper 5PB5/12 under the
neutral light. They have practically the same chromaticity (Figure 2). Yet, they were judged at 60% of the
maximal dissimilarity (which was 104% of the standard,
and observed between papers 5PB5/12 and 5R4/14
under the blue and yellow lights respectively). This
indicates that observers did not rest their judgments
upon the chromaticity of the light reflected from these
papers. In other words, the dissimilarity between these
papers renders a different dimension.
As argued elsewhere (Tokunaga and Logvinenko,
2010), this dimension is unlikely to be brightness. More
specifically, although the yellow and blue lights might
have differed from the neutral light in brightness, it is
highly unlikely that the hue contours are displaced with
respect to each other because of the possible brightness
difference. Indeed, if this were the case then in our
previous experiment, in which the same three illuminants and only yellow, neutral, and blue Munsell papers
were used, the obtained (i.e., lighting) shift should have
been observed in a direction collinear with that along
which the papers were arranged. However, as mentioned
above, the lighting shift was transversal to the material
hue contours. Hence, we believe that this new dimension
cannot be reduced to either the brightness or chromaticity of the reflected light. In line with the terminology
put forward previously (Tokunaga and Logvinenko,
2010) it will be referred to as lighting hue. Thus, the
material hue varies along the hue contours, the lighting
hue in a transverse direction.
The Friedman test (a two-way non-parametric ANOVA) performed for a subset of data obtained only for
sessions where both fields were equally illuminated,
showed a significant effect of illumination between hues
(v22 ¼ 85:3; p<0:001). This means that the three hue
contours in Figure 3 are significantly different from each
other. To look into this difference we calculated the
centroid of each hue contour, and the radial distances
from each centroid to all the seven points of each
contour (Figure 5). The distances for the hue contour
obtained under the blue illumination were found to be
smaller than for the neutral and yellow illuminations. It
follows that the hue contour for the blue illumination is
somewhat shrunken. This is similar to the lightness
continuum shrinking observed for darker lights by
Logvinenko and Maloney (2006).
Figure 5. Shrinking of the blue hue contour. The vertical axis
represents the distance from each point in Figure 3 to the corresponding centroid. The colour of line represents the colour of light.
ª 2010 The Authors, Ophthalmic and Physiological Optics ª 2010 The College of Optometrists
Material and lighting hues: R. Tokunaga and A. D. Logvinenko
Figure 6. Location of the colour stimuli in the CIELAB space.
Notations are the same as in Figure 2. The yellow cross, black plus
sign, and blue circle (which overlap) indicate the white surface under
yellow, neutral and blue illumination respectively.
Discussion
Multidimensional scaling of Munsell papers lit by
chromatic lights shows that papers of different illumination are dissociated into separate layers in the
dissimilarity space. For example, the papers under blue
light are significantly dislocated from the papers under
yellow and neutral lights. Therefore, contrary to an old
but still popular belief that illumination is discounted by
the visual system so as to compute object colour (e.g.,
Kaiser and Boynton, 1996; Whittle, 2003), illumination
is represented in the perceptual space. Specifically,
illumination is represented with what we call lighting
dimensions of the object colour manifold (e.g., surface
brightness and lighting hue).
There is every indication that the lighting dimensions
cannot be reduced to the colour dimensions of reflected
light. Indeed, as shown in our experiment, differently
illuminated papers reflecting nearly metameric lights
were judged as very dissimilar. Moreover, the lighting
dimensions are unlikely to be deduced from the tristimulus coefficients of the reflected light. For example,
CIELAB fails to predict a dissociation of hue contours
as regards a lighting hue similar to that observed in
Figure 3. The CIELAB prediction is displayed in
Figure 6. The three hue contours as predicted by
CIELAB are intertwined with each other. There is no
systematic shift between the hue contours produced for
different illuminations in Figure 6.
This is hardly surprising because, first of all, CIELAB
(as well as more recent colour appearance models
approved by the CIE, such as the CIECAM02) is
applicable only for single-illuminant scenes. They are
not supposed to be used for scenes with multiple lights,
such as that of our experiment. Secondly, while
CIECAM02 duplicates some colour dimensions (i.e.,
brightness vs lightness, colourfulness vs chroma), it
retains hue as a single dimension. It follows that within
the conceptual framework on which CIELAB and
615
CIECAM02 are based, there is no room for two types
of hue. Admittedly, there are two ÔkindsÕ of hue in
CIECAM02. However, these are merely two ways of
representing what is actually the same entity, rather than
actually two sorts of hue, which we are proposing the
existence of in the present paper.
Although lighting hue is obviously related to the
illumination, it is not the hue of the ambient illumination. Assuming that our observers saw the colour of
objects and the colour of their illumination separately
and independently, (as suggested by some scientists, e.g.,
Mausfeld, 1998; MacLeod, 2003), the dissimilarity
between the pairs of identical papers would be approximately equal. This is because the dissimilarity would be
determined only by the difference in illumination. In
other words, the symbols in the bottom row graphs in
Figure 4 would be arranged horizontally at the same
level. However, this is definitely not the case. The
Friedman test performed for this subset of data showed
the effect of paper to be highly significant (v26 ¼
131:0; p<0:001). Thus, dissimilarity between paper/light
pairs is not separable, that is, it cannot be split into that
dissimilarity between papers, and that between lights.
Therefore, it would be a mistake to treat lighting hue as
the illumination hue.
The above discussion leads to the following question:
are lighting and material hues different colour dimensions, or are they just two sides of the same coin? In
other words, is there a qualitative difference between the
hue difference produced by the neutral paper lit by
yellow and the same paper lit by blue light, on the one
hand; and the hue difference produced by the yellow
paper and blue paper under day light? Our data suggests
that lighting and material hues are qualitatively different
types of experience. Indeed, when the yellow paper lit by
the blue light is observed, we experience strong lighting
blue and strong material yellow in the same place and at
the same time. Therefore, the major dogma of colour
vision that yellow and blue cannot be simultaneously
experienced (Hering, 1920/1964) should be restricted to
hues of the same type: i.e., either material or lighting
ones. For example, material yellow and material blue
cannot be simultaneously experienced in the same place
under normal viewing conditions.
The coexistence of lighting blue and material yellow
hues can be understood if we assume that each illumination generates its own three-dimensional material
colour manifold. Recall that the colour solids induced
by different illumination are geometrically different,
that is, they take different shapes (Logvinenko, 2009b).
Therefore, our assumption is that different colour solids
induce different three-dimensional manifolds of object
colours. In a multiple light scene, as in our experiment,
observers are able to distinguish areas of object
colours belonging to different object colour manifolds.
ª 2010 The Authors, Ophthalmic and Physiological Optics ª 2010 The College of Optometrists
616
Ophthal. Physiol. Opt. 2010 30: No. 5
Moreover, they are able to order these areas as they can
order self-luminous objects. Thus, in a multiple-light
variegated scene there are two colour orders: material
and lighting. Therefore, the distinction between the
material and lighting colour dimensions is nothing more
than the distinction between the material and lighting
colour orders. In fact, there is just a single set of hues,
but a two-dimensional one. There are two colour orders,
acting on this set: the material order and the lighting
order. For instance, a circular order along the material
hue contours in Figure 3a illustrates one order, whereas
the same points in Figure 3b are ordered in the transverse direction that illustrates another order.
Although both the material and the lighting dimensions were found to contribute to the dissimilarity
judgements, contribution from the lighting dimensions
was found to be considerably less effective than that
from the material dimensions. Indeed, the separation
between the three hue contours in Figure 3a is much
smaller than their size. For example, the shift between
the yellow and blue contours in the CIE uniform
chromaticity diagram (i.e., the distance between the
neutral points under the yellow and blue illuminations)
is 1.38 times as much as the distance between the most
distant papers 5R4/14 and 10P5/12. On the contrary, the
shift between the yellow and blue hue contours (i.e., the
distance between the corresponding centroids) in Figure 3a is only 31% of the largest dissimilarity between
papers under the blue light (papers 5R4/14 and 5PB5/
12). Therefore, the lighting hue difference is discounted
by a factor of 4.4. A similar lighting discounting was
found in related studies (Tokunaga and Logvinenko,
2010; Logvinenko and Maloney, 2006).
As pointed out elsewhere (Tokunaga and Logvinenko, 2010), the existence of the lighting dimensions
alongside the material dimensions of object colours
addresses the paradox familiar to everyone who has
dealt with objects under chromatic illumination. On the
one hand, chromatic light makes objects change their
colour appearance. On the other, in a sense, the objectsÕ
colour appearance remains relatively constant despite
the change of the illumination. Although such objectcolour constancy has been intensively studied (e.g.,
Katz, 1935; Pokorny et al.,1991; Smithson, 2005; Foster, 2008; Brainard, 2009), it still remains unclear why
object-colour constancy has never been registered perfect, and what exactly it is that remains constant when
illumination changes. The imperfection of object-colour
constancy manifests itself in the impossibility of achieving an asymmetric (i.e., across illuminants) colour
match, a well known phenomenon in literature on
colour constancy (e.g., Brainard et al., 1997; Foster,
2003; Logvinenko and Maloney, 2006). An asymmetric
colour match is indeed impossible, because the difference in illumination results in a difference in lighting
dimensions; for instance, in lighting hue – as in our
experiment. However, it might happen that despite the
difference in the lighting dimensions, the colour of the
objects will be the same with respect to the material
dimensions. In this case one can say that a sort of objectcolour constancy takes place. However, asymmetric
colour matching is not quite the appropriate method of
measuring the constancy of the material dimensions of
object colour. One needs more elaborate techniques for
this purpose.
To summarise, the multidimensional analysis of
Munsell papers lit by various chromatic lights shows
that we experience two types of hues: material hues and
lighting hues. This paves the way for a bold idea – that
the object colour manifold can be envisaged as a tridimensional bundle of the traditional (material) threedimensional object colour manifolds.
Acknowledgement
Supported by EPSRC research grant EP/C010353/1
(AL).
References
Brainard, D. H. (2003) Color appearance and color difference
specification. In: The Science of Color, 2nd edn (ed. S. K.
Shevell), Optical Society of America, Washington D.C, pp.
191–216.
Brainard, D. H. (2009) Color constancy. In: The Sage
Encyclopedia of Percept (ed. B. Goldstein), Sage Publications, Los Angeles, pp. 253–257.
Brainard, D. H., Brunt, W. A. and Speigle, J. M. (1997)
Colour constancy in the nearly natural image. I. Asymmetric
matches. J. Opt. Soc. Am.A. 14, 2091–2110.
Cox, T. F. and Cox, M. A. A. (2001) Multidimensional scaling,
2nd edn, Chapman & Hall/CRC, Boca Raton, FL.
Craven, B.J. and Foster, D.H. (1992) An operational approach
to colour constancy. Vision Res. 32, 1359–1366.
Evans, R. M. (1974) The Perception of Color. Wiley, New
York.
Fairchild, M. D. (2005) Color Appearance Models, 2nd edn,
Wiley, Chichester, UK.
Foster, D.H. (2003) Does colour constancy exist? Trends Cogn.
Sci. 7, 439–443.
Foster, D.H. (2008) Color Appearance. In: The Senses: A
Comprehensive Reference, Vol. 2, Vision II (Volume Eds T.
Albright and R. H. Masland; Series Eds. A. I. Basbaum, A.
Kaneko, G. M. Shephard and G. Westheimer), Academic
Press, San Diego, pp. 119–132.
Foster, D.H., Nascimento, S. M. C., Amano, K., Arend, L.,
Linnell, K. J., Nieves, J. L., Plet, S. and Foster, J. S. (2001)
Parallel detection of violations of color constancy. Proc.
Natl Acad. Sci. USA 98, 8151–8156.
Hering, E. (1920/1964) Outlines of a Theory of the Light Sense
(translated by Hurvich, L. M. and D. Jameson), Harvard
University Press, Cambridge, MA.
ª 2010 The Authors, Ophthalmic and Physiological Optics ª 2010 The College of Optometrists
Material and lighting hues: R. Tokunaga and A. D. Logvinenko
Hunt, R. W. G. (1977) The specifications of colour appearance. 1. Concepts and Terms. Color Res. Appl. 2, 55–68.
Kaiser, P. K. and Boynton, R. M. (1996) Human Color Vision,
Optical Society of America, Washington, DC.
Katz, D. (1935) The world of Colour, Kegan Paul Trench,
Trubner & Co., London.
Kingdom, F. (2008) Perceiving light versus material. Vision
Res. 48, 2090–2105.
Logvinenko, A. D. (2005) Does luminance contrast determine
lightness? Spat. Vis. 18(3), 337–346.
Logvinenko, A. D. (2009a) Pseudoscopic colour illusions. In:
Proceedings of ‘‘11th Congress of the International Colour
Association (AIC). 27 Sep – 2 October, Sydney, Australia.’’
(in press).
Logvinenko, A. D. (2009b) An object-colour space. J. Vis. 9,
1–23.
Logvinenko, A. D. and Maloney, L. T. (2006) The proximity
structure of achromatic colors and the impossibility of
asymmetric lightness matching. Percept. Psychophys. 68, 76–
83.
MacLeod, D. I. (2003) New dimensions in color perception.
Trends Cogn. Sci. 7, 97–99.
Mausfeld, R. (1998) Color perception: from Grassmann codes
to a dual code for object and illumination colors. In: Color
Vision (eds W. Backhaus, R. Kliegl and J. Werner), De
Gruyter, Berlin, pp. 219–250.
Pokorny, J., Shevell, S. K. and Smith, V. C. (1991) Colour
appearance and colour constancy. In: The Perception of
Colour, Vol 6, Vision & Visual Dysfunction (ed. P. Gouras),
Macmillan, Basingstoke, England, pp. 43–61.
Smithson, H. E. (2005) Sensory, computational and cognitive
components of human colour constancy. Philos. Trans. R.
Soc. Lond. B Biol. Sci. 360, 1329–1346.
Tokunaga, R. and Logvinenko, A. D. (2010) Material and
lighting dimensions of object colour. Vision Res. 50, 1740–
1747.
Tokunaga, R., Logvinenko, A. D. and Maloney, L. T. (2008)
Dissimilarity of yellow-blue surfaces under neutral light
sources differing in intensity: Separate contributions of light
intensity and chroma. Vis. Neurosci. 25, 395–398.
617
Whittle, P. (2003) Contrast colours. In: Colour Perception:
Mind and the Physical World (eds R. Mausfeld and D.
Heyer), Oxford University Press, Oxford, pp. 115–138.
Wyszecki, G. and Stiles, W. S. (1982) Color Science. Concepts
and Methods, Quantitative Data and Formulae, 2nd Edn,
John Wiley & Sons, New York.
Appendix
Table A1. The coordinates of the MDS output configurations plotted in Figure 3.
Dimension
Illumination
Munsell paper
1
2
3
Yellow
5R4/14
5YR7/12
5Y8/12
5G6/10
10BG5/8
5PB5/12
10P5/12
5R4/14
5YR7/12
5Y8/12
5G6/10
10BG5/8
5PB5/12
10P5/12
5R4/14
5YR7/12
5Y8/12
5G6/10
10BG5/8
5PB5/12
10P5/12
)16.523
)11.443
)8.252
8.427
7.448
7.738
)4.549
)15.826
)9.483
)8.306
8.594
8.337
9.111
)1.311
)7.941
)5.200
)4.545
9.373
11.608
13.997
8.746
)8.426
5.331
14.498
6.010
0.973
)4.684
)7.843
)9.291
4.735
15.570
5.984
1.097
)4.657
)7.350
)13.734
)0.659
10.201
2.457
0.266
)3.253
)7.225
4.954
1.962
)2.993
4.231
5.130
1.535
0.265
1.790
)0.913
)0.340
1.297
2.364
)3.933
)4.788
)1.984
)2.379
)2.949
0.285
)1.322
0.210
)2.423
Neutral
Blue
ª 2010 The Authors, Ophthalmic and Physiological Optics ª 2010 The College of Optometrists