blelb blelb visual perception lab
   

Experiments with Spot 01


In blelbspot 01, you can allocate to every colour field 6 different shades of grey including black and white or 6 different primary colours. The cross-hairs can be moved on the right-hand diagram to alter the proportions of the four fields of a square picture module. Counting the various zoom levels, billions of different images can be created using this tool. The exceptional features of some of these images are commented on below :

Experiments with grey tones

Unambiguous transparency (pictures 1 and 2)
Even with grey tones, interesting phenomena can be observed. In perception research, we talk of the transparency effect (1) which can be systematically examined by means of patterns of stripes. If we choose four different shades of grey for the four elements, 24 different configurations become possible. In the case of 8 configurations, we can say precisely which stripes appear to run at the front of the picture and which are located, partially covered, in the middle-ground. In pictures 1 and 2, perception of transparency is unambiguous.
Unambiguous transparency Unambiguous transparency Ambiguous transparency No transparency
1 2 3 4
 
Ambiguous transparency (picture 3 )
In the case of 8 further configurations, we can speak of ambiguous transparency, that is to say, as in picture 3, there are two possible interpretations which are not perceived with the same degree of ease.
No transparency (picture 4)
In the case of the remaining 8 configurations, no transparency is evident; there are no fields which allow a background to shimmer through. Although we can see horizontal and vertical stripes, the elements at the intersections are opaque like the black squares in picture 4.
Criteria
  • Unambiguous transparency occurs
     1. when the two lightest elements of the module are next to each other and
     2. the darkest borders on the lightest.
  • Ambiguous transparency occurs
     when only the first of these conditions is met.
  • There is no transparency
     when the first condition is not fulfilled.
The geometry of the stripes, which can be changed in blelbspot 01 by moving the crosshairs, has no influence on these different categories. Research findings relating to the conditions governing transparency can be found in various works including those of Metelli (2).

Experiments with colours

Metelli's Assumption
Fabio Metelli (2) developed a theory about how the properties of the elements are split up in the transparency range. This only applies very roughly in the experiments shown here. According to Metelli, the colour properties of overlapping are governed by the proportional addition of colour from the properties of the foreground and the opaque middle-ground. Metelli determines the hypothetical degree of mixing algebraically from the light values of the individual colour fields. We believe that the blue, green and red content of the individual mixed colours should be investigated separately and how they work together in the transparency range explored. In addition to that, it has been shown that the shape and the structural organisation of the elements play a role. It is also necessary to take the Retinex Theory of Edwin H. Land (3) into account. In any case, the following colour illusion already shows that Metelli's assumption cannot be valid in all situations.
Transparenz-Effekt
5
Transparency effect (pictures 5 - 8)
If we apply the theory of transparency to the grey values of the colours used in pictures 5 - 8, a pattern of stripes demonstrating ambiguous transparency is produced. It is possible to see either the vertical stripes in cyan or the horizontal ones in red in the foreground. In contrast to the grey value rules, however, we can force light horizontal or vertical stripes into the foreground. The use of colours increases the complexity of the perceptual experience.
Colour illusions
The neutral grey squares appear to have a reddish hue.
Transparency effect The effect of scaling The effect of scaling
6 7 8
 
The effect of scaling
Pictures 6 through 8 demonstrate a constant doubling of scale. It becomes increasingly difficult to identify the borders of the structures as straight equidistant lines. The geometrical/optical illusions change. All neighbouring fields have an influence on one another. If we observe picture 8 from a greater distance, it soon becomes possible to see an integral grey value which can be predicted using the rules of proportional colour addition.
 

Experiments with colours and proportions

Organisation of foreground, middle-ground and background (pictures 9 and 10)
Picture 9 shows narrow horizontal stripes in the foreground which are strongly influenced by their larger neighbouring elements. When looked at from close up, the grey fields demonstrate an intensive red hue; when looked at from further away, however, the grey appears bluish. When we see black as the background, the red columns appear in the middle-ground and as a result of the horizontal stripes in front of them demonstrate rhythmical colour distortions and constriction. When we see red as the background, the black columns in the middle-ground demonstrate concave distortion. With black in the middle-ground, there is no colour distortion.
 
geometrical illusion
9
The geometrical illusion, which jumps from the red objects to the black ones when background and middle-ground are interchanged, can be better understood in the light of the findings of Gaetano Kanizsa (5). Kanizsa shows that when one colour is positioned over another, the border always belongs to the object in front. Picture 9 shows for the first time that with spatial organisation that can be interpreted in various ways an illusion of the border can also tilt; the aforementioned border distortion changes the sign.
Simultaneous contrast
10
Simultaneous contrast (picture 10)
Picture 10 shows white or black columns with distorted borders in the middle-ground and opaque red stripes in the foreground which appear light red over the black and dark red over the white. This colour illusion is a special case of simultaneous contrast.
 
Size illusion
11
Size illusion (picture 11)
What can be seen here are either vertical white stripes with blue inclusions in the foreground or horizontal blue stripes in the background. What is striking here is the size illusion; the group of white stripes seems wider than the group of blue stripes. Light objects on a black background appear to expand, while dark objects on a light background appear to shrink.
 
Contradiction to the grey scale rule
12
Contradiction to the grey scale rule on transparency (picture 12)
According to Metelli's theory the narrow horizontal stripes should appear to be transparent. In fact, they appear to be opaque. In this picture, the spatial organisation has four possible interpretations; the vertical yellow bands or the vertical white ones can be in the foreground, or alternatively the horizontal black stripes or horizontal yellow bands are experienced at the front. In the first case, the background seems white, in the second yellow, in the third yellow and in the fourth black. The border illusions follow these changes; the horizontal black stripes seem wider than the vertical white ones.
 
Bamboo effect
13
Bamboo effect (picture 13)
Border illusions dominate in picture 13; the brain is not capable of seeing the borders of the vertical bands as straight lines. This distortion is caused by the black inclusions. The yellow rectangles tend to pull in and appear concave; the horizontal borders expand and bend out slightly, the black squares are distorted into knots. The bamboo effect can probably not be explained by the established size illusion.
Fluttering hearts
14
Fluttering hearts (picture 14)
This effect occurs when the colours used in the figures and the ground have approximately the same degree of lightness and are decoded by the colour receptors on the retina at different speeds. Blue objects on a red ground begin to dance when the light is dim, and when the picture is shaken slightly their geometry is registered by the retina out-of-phase.
Sir Charles Wheatestone and Sir David Brewster published this effect in 1844. This phenomenon is known in art history under the name of "fluttering hearts" thanks to the work of Marcel Duchamp (7).
In addition, angle distortion and wavy shaped configuration of the objects occur. If the figures and ground are exchanged, the geometry of the red grid in the foreground is stabilised.
 
Neon effect
15
Neon effect (picture 15)
In picture 15, the cyan line segments on the black grid seem to be surrounded by a soft fluorescent shimmer, which spreads out into the intermediate white spaces. This illusion was first discovered by the Dutch psychologist, van Tuijl in 1975 (6). In addition, the intermediate white spaces take on a greyish-reddish hue.
Modified Hermann grid
16
Modified Hermann grid (picture 16)
The optical effects at the intersections of a regular grid have been known since 1870. In the Hermann grid, light or dark points appear which are assumed to be an accompanying effect of neuronal image processing. The receptors triggered by the intersection points are subject to lateral inhibitions from four sides, while all the other points of the grid are affected from only two sides. In Picture 16, the intersection points are coloured white rather than the yellow of the classic Hermann grid. This leads to the occurrence of a new colour effect. Around the spot on which the eye is fixed, the intersections remain white. Away from this point they flash lilac; as the eye roves over the picture, the coloured areas move, too. When the distance from which the picture is viewed is increased, their number multiplies. Lateral inhibition is only removed in the area of the fovea centralis. Only those intersection points which are not in the fovea centralis area remain unaffected by the illusion.
 
Horizontal vertical effect (pictures 17 and 18)
Here, you can see two different pictures. Compare their features with regard to colour, shape and spatial layering. In fact, these two pictures are identical apart from having been rotated through 90 degrees. Both the colour and geometrical illusions probably stem from a preferred reading direction and from the slight astigmatism that virtually everybody has.
 
Horizontal vertical effect 1  Horizontal vertical effect 2
17  18
 
Bezold effect (pictures 19 and 20)
What is striking here is Wilhelm von Bezold's (4) colour illusion, which is also referred to as the assimilation effect, and which still remains a neurological puzzle today. Both pictures contain the same colours but their different positions relative to one another lead to extreme changes in colour. The same red is perceived as red-violet in one picture and as orange-red in the other. Blue and yellow change their hues less because - in this case - only two neighbouring elements are different.
The results of an in-depth investigation of the Bezold effect together with a parameter study and effect optimisation have been published (8).
 
Bezold effect 1  Bezold effect 2
19  20

Bibliography:
1) Watanabe, T. & Cavanagh, P. (1993)
Surface decomposition accompanying the perception of transparency.
Spacial Vision, 7, 95-111.
2) Metelli, F. (1974)
The Perception of transparency.
Scientific American 230(4), 90-98
3) Land, E. H. (1977)
The Retinex theory of color vision.
Scientific American 237, 108-128.
4) Bezold, W. v. (1874)
Die Farbenlehre. (Braunschweig: Westermann).
5) Kanisza, G. (1979)
Organisation in Vision. Essays on Gestalt Perception, Praeger.
6) van Tujil, H. (1975)
A new visual illusion: Neonlike color spreading and complementary color induction between subjective contours.
Acta Psychologica, 39, 441-445.
7) Knuchel, H. und Nänni, J. (1996)
Seesaw.
Lars Müller Publishers, CH-5401 Baden, ISBN 3-906700-73-9
8) Knuchel, H. und Nänni, J. (1991)
BLUE. YELLOW. RED. Color-Anagrams
Lars Müller Publishers, CH-5401 Baden, ISBN 3-906 700-41-0
 
03/16/2002