Lightness constancy and illumination discounting

Contrary to the implication of the term "lightness constancy", asymmetric lightness matching has never been found to be perfect unless the scene is highly articulated (i.e., contains a number of different reflectances). Also, lightness constancy has been found to vary for different observers, and an effect of instruction (lightness vs. brightness) has been reported. The elusiveness of lightness constancy presents a great challenge to visual science; we revisit these issues in the following experiment, which involved 44 observers in total. The stimuli consisted of a large sheet of black paper with a rectangular spotlight projected onto the lower half and 40 squares of various shades of grey printed on the upper half. The luminance ratio at the edge of the spotlight was 25, while that of the squares varied from 2 to 16. Three different instructions were given to observers: They were asked to find a square in the upper half that (i) looked as if it was made of the same paper as that on which the spotlight fell (lightness match), (ii) had the same luminance contrast as the spotlight edge (contrast match), or (iii) had the same brightness as the spotlight (brightness match). Observers made 10 matches of each of the three types. Great interindividual variability was found for all three types of matches. In particular, the individual Brunswik ratios were found to vary over a broad range (from .47 to .85). That is, lightness matches were found to be far from veridical. Contrast matches were also found to be inaccurate, being on average, underestimated by a factor of 3.4. Articulation was found to essentially affect not only lightness, but contrast and brightness matches as well. No difference was found between the lightness and luminance contrast matches. While the brightness matches significantly differed from the other matches, the difference was small. Furthermore, the brightness matches were found to be subject to the same interindividual variability and the same effect of articulation. This leads to the conclusion that inexperienced observers are unable to estimate both the brightness and the luminance contrast of the light reflected from real objects lit by real lights. None of our observers perceived illumination edges purely as illumination edges: A partial Gelb effect ("partial illumination discounting") always took place. The lightness inconstancy in our experiment resulted from this partial illumination discounting. We propose an account of our results based on the two-dimensionality of achromatic colour. We argue that large interindividual variations and the effect of articulation are caused by the large ambiguity of luminance ratios in the stimulus displays used in laboratory conditions.     

Lightness constancy: Ratio invariance and luminance profile

The term simultaneous lightness constancy describes the capacity of the visual system to perceive equal reflecting surfaces as having the same lightness despite lying in different illumination fields. In some cases, however, a lightness constancy failure occurs; that is, equal reflecting surfaces appear different in lightness when differently illuminated. An open question is whether the luminance profile of the illumination edges affects simultaneous lightness constancy even when the ratio invariance property of the illumination edges is preserved. To explore this issue, we ran two experiments by using bipartite illumination displays. Both the luminance profile of an illumination edge and the luminance ratio amplitude between the illumination fields were manipulated. Results revealed that the simultaneous lightness constancy increases when the luminance profile of the illumination edge is gradual (rather than sharp) and homogeneous (rather than inhomogeneous), whereas it decreases when the luminance ratio between the illumination fields is enlarged. The results are interpreted according to the layer decomposition schema, stating that the visual system splits the luminance into perceived lightness and apparent illumination components. We suggest that illumination edges having gradual and homogeneous luminance profiles facilitate the luminance decomposition process, whereas wide luminance ratios impede it.     

Measuring the meter: on the constancy of lightness scales seen against different backgrounds

The constancy of a 16-step achromatic Munsell scale was tested with regards to background variations in two experiments. In experiment 1 three groups of observers were asked to find lightness matches for targets in simultaneous lightness displays by using a 16-step achromatic Munsell scale placed on a white, black, or white-black checkered background. In experiment 2, a yellow-blue checkered background and a green-red checkered background replaced Munsell scales on the black and on the white backgrounds. Significant effects of scale background on matches were found only in experiment 1, suggesting that background luminance is a crucial factor in the overall appearance of the scale. The lack of significant differences in experiment 2, however, may stand for an overall robustness of the scale with respect to background luminance changes occurring within certain luminance ranges.     

Luminance gradient can break background-independent lightness constancy

A display with a luminance gradient was shown to induce a strong lightness illusion (Logvinenko, 1999 Perception 28 803-816). However, a 3-D cardboard model of this display was found to produce a much weaker illusion (less than half that in the pictorial version) despite the fact that its retinal image is practically the same. This is in line with the hypothesis that simultaneous lightness contrast is solely a phenomenon of pictorial perception (Logvinenko et al, 2002 Perception 31 73-82). The residual lightness illusion in the 3-D model can be accounted for by the fact that this model is a hybrid display. Specifically, while it is a real object, a pictorial representation (of the illumination gradient) is superimposed on it. Thus, lightness in the 3-D display is a compromise between two opposite tendencies: the background-independent lightness constancy and the lightness illusory shift induced by the luminance gradient.     

The influence of depth segmentation on colour constancy

In real scenes, surfaces in different depth planes often differ in the luminance and chromatic content of their illumination. Scene segmentation is therefore an important issue when considering the compensation of illumination changes in our visual perception (lightness and colour constancy). Chromatic adaptation is an important sensory component of colour constancy and has been shown to be linked to the two-dimensional spatial structure of a scene (Werner, 2003 Vision Research 43 1611 - 1623). Here, the question is posed whether this cooperation also extends to the organisation of a scene in depth. The influence of depth on colour constancy was tested by introducing stereo disparity, whereby the test patch and background were perceived in either the same or one of five different depth planes (1.9-57 min of arc). There were no additional cues to depth such as shadows or specular highlights. For consistent illumination changes, colour constancy was reduced when the test patch and background were separated in depth, indicating a reduction of contextual influences. An interaction was found between the influences of stereo depth and spatial frequency on colour constancy. In the case of an inconsistent illumination change, colour constancy was reduced if the test patch and background were in the same depth plane (2-D condition), but not if they were separated in depth (3-D condition). Furthermore, colour constancy was slightly better in the 3-D inconsistent condition than in the 2-D inconsistent condition. It is concluded that depth segmentation supports colour constancy in scenes with inconsistent illumination changes. Processes of depth segmentation are implemented at an early sensory stage of colour constancy, and they define visual regions within which the effects of illuminant changes are discounted for separately. The results support recent models that posit such implementation of scene segmentation in colour constancy.     

Illumination change at a depth edge can reduce lightness constancy

Lightness constancy requires that a surface retain its lightness not only when the illumination is changed but also when the surface is moved from one background to another. Occlusion of one surface by another frequently results in a retinal juxtaposition of patches under different illuminations. At such edges, retinal luminance ratios can be much higher than in scenes with a single illumination. We demonstrate that such retinal adjacencies can produce failures of lightness constancy. We argue that they are responsible for departures from perfect lightness constancy in two prior experiments that examined the effects of depth relations on lightness constancy.     

The classification and integration of edges as critical to the perception of reflectance and illumination

A pattern of luminances equivalent to that of a traditional simultaneous lightness display (two equal gray squares, one on a white background and the other on an adjacent black background) was presented to observers under two conditions, and matches were obtained for both perceived reflectance and perceived illumination level of the squares and their backgrounds. In one condition, the edge dividing the two backgrounds was made to appear as the boundary between a white and a black surface, as in the traditional pattern. The squares then were perceived as almost the same shade of middle gray. In the other condition, a context was supplied that made the edge between the backgrounds appear as the boundary between two illumination levels, causing one square to appear black and the other white. These results were interpreted as a problem for local ratio theories, local edge theories, and lateral inhibition explanations of lightness constancy, but as support for the concepts of edge classification, edge integration, and the retinal image as a dual image.     

Infant lightness perception: do 4-month-old infants follow Wallach's ratio rule?

When adults view a test disk embedded in a higher-luminance surround, the perceived lightness of the disk is largely determined by the surround-to-disk (S/D) luminance ratio (Wallach's ratio rule). Performance of 4-month-old infants tested with a forced-choice novelty-preference technique was consistent with predictions based on Wallach's ratio rule. This result suggests that the ability to extract and maintain information about local luminance ratios is present early in infancy. This ability is likely to contribute to the development of lightness constancy.     

Contrast constancy revisited: The perceived contrast of sinusoidal gratings above threshold

The contrast sensitivity function of the human visual system, measured with sinusoidal luminance gratings, has an inverted U shape with a peak around 2–4?c/deg. Above threshold, it is thought that luminance gratings of equal physical contrasts but of distinguishably different spatial frequencies are all perceived as having similar contrasts, a phenomenon that has been termed contrast constancy. However, when suprathreshold contrast matches were measured for pairs of luminance gratings whose spatial frequencies were indistinguishable, the matching curves were not flat and followed a similar inverted U shape form as the contrast sensitivity function at threshold. It was therefore suggested that contrast constancy may only be revealed when matching the contrasts of clearly distinguishable spatial frequencies. Here, observers matched the perceived contrasts of suprathreshold luminance gratings of similar but visibly different spatial frequencies between 0.25 and 16?c/deg. The results show that, much like the contrast sensitivity function at threshold, observers are more sensitive to intermediate spatial frequencies (1–6?c/deg) than they are to either higher or to lower spatial frequencies. This tuning is evident when matching reference contrasts of 30–80%, implying a significant role in everyday vision. To demonstrate that these results were not due to local adaptation, the experiment was repeated with shorter stimulus duration, producing the same results. The extent of departure from contrast constancy found in the present study is compared to previously reported suprathreshold measurements. The results are also discussed with consideration to limitations with display apparatus such as monitor blur.     

Mesopic lightness,brightness, and brightness contrast

At mesopic mean luminances, a fixed luminance contrast produces less brightness contrast than it does at photopic luminances. This suggests that lightnesses of surfaces might also be altered at low luminances. I measured lightness, brightness, and brightness contrast in CRT simulations of achromatic paper patchworks. The illuminance of the standard pattern was fixed, producing 0.12,1.2, or 12 cd/m². The illuminance on the test pattern was varied in a lightness constancy paradigm. Constant brightness contrast required more luminance contrast at lower mean luminances. Failures of lightness constancy occurred at the lowest mean luminances, but they were minor in comparison with the loss of brightness contrast in the same pattern. These results have implications for imaging applications. Often, image content falls in both the photopic and the mesopic ranges. Our results indicate that brightness contrast may decrease substantially in low-luminance regions without large changes of surface lightness.     

Lightness of an object under two illumination levels

Anchoring theory (Gilchrist et al, 1999 Psychological Review 106 795-834) predicts a wide range of lightness errors, including failures of constancy in multi-illumination scenes and a long list of well-known lightness illusions seen under homogeneous illumination. Lightness values are computed both locally and globally and then averaged together. Local values are computed within a given region of homogeneous illumination. Thus, for an object that extends through two different illumination levels, anchoring theory produces two values, one for the patch in brighter illumination and one for the patch in dimmer illumination. Observers can give matches for these patches separately, but they can also give a single match for the whole object. Anchoring theory in its current form is unable to predict these object matches. We report eight experiments in which we studied the relationship between patch matches and object matches. The results show that the object match represents a compromise between the match for the patch in the field of highest illumination and the patch in the largest field of illumination. These two principles are parallel to the rules found for anchoring lightness: highest luminance rule and area rule.     

Perceived area and the luminosity threshold.

Observers made forced-choice opaque/luminous responses to targets of varying luminance and varying size presented (1) on the wall of a laboratory, (2) as a disk within an annulus, and (3) embedded within a Mondrian array presented within a vision tunnel. Lightness matches were also made for nearby opaque surfaces. The results show that the threshold luminance value at which a target begins to appear self-luminous increases with its size, defined as perceived size, not retinal size. More generally, the larger the target, the more an increase in its luminance induces grayness/blackness into the surround and the less it induces luminosity into the target, and vice versa. Corresponding to this luminosity/grayness tradeoff, there appears to be an invariant: Across a wide variety of conditions, a target begins to appear luminous when its luminance is about 1.7 times that of a surface that would appear white in the same illumination. These results show that the luminosity threshold behaves like a surface lightness value--the maximum lightness value, in fact--and is subject to the same laws of anchoring (such as the area rule proposed by Li & Gilchrist, 1999) as surface lightness.     

Perceived area and the luminosity threshold

Observers made forced-choice opaque/luminous responses to targets of varying luminance and varying size presented (1) on the wall of a laboratory, (2) as a disk within an annulus, and (3) embedded within a Mondrian array presented within a vision tunnel. Lightness matches were also made for nearby opaque surfaces. The results show that the threshold luminance value at which a target begins to appear self-luminous increases with its size, defined as perceived size, not retinal size. More generally, the larger the target, the more an increase in its luminance induces grayness/blackness into the surround and the less it induces luminosity into the target, and vice versa. Corresponding to this luminosity/grayness tradeoff, there appears to be an invariant: Across a wide variety of conditions, a target begins to appear luminous when its luminance is about 1.7 times that of a surface that would appear white in the same illumination. These results show that the luminosity threshold behaves like a surface lightness value—the maximum lightness value, in fact—and is subject to the same laws of anchoring (such as the area rule proposed by Li & Gilchrist, 1999) as surface lightness.     

Response priming driven by local contrast, not subjective brightness

We demonstrate qualitative dissociations of brightness processing in visuomotor priming and conscious vision. Speeded keypress responses to the brighter of two luminance targets were performed in the presence of preceding dark and bright primes (clearly visible and flanking the targets) whose apparent brightness values were enhanced or attenuated by a visual illusion. Response times to the targets were greatly affected by consistent versus inconsistent arrangements of the primes, relative to the targets (response priming). Priming effects could systematically contradict subjective brightness matches, such that one prime could appear brighter than the other but could prime as if it were darker. Systematic variation of the illusion showed that response-priming effects depended only on local flanker-background contrast, not on the subjective appearance of the flankers. Our findings suggest that speeded motor responses, as opposed to conscious perceptual judgments, access an early phase of lightness and brightness processing prior to full lightness constancy.     

Lightness constancy: a direct test of the illumination-estimation hypothesis

Many models of color constancy assume that the visual system estimates the scene illuminant and uses this estimate to determine an object's color appearance. A version of this illumination-estimation hypothesis, in which the illuminant estimate is associated with the explicitly perceived illuminant, was tested. Observers made appearance matches between two experimental chambers. Observers adjusted the illumination in one chamber to match that in the other and then adjusted a test patch in one chamber to match the surface lightness of a patch in the other. The illumination-estimation hypothesis, as formulated here, predicted that after both matches the luminances of the light reflected from the test patches would be identical. The data contradict this prediction. A second experiment showed that manipulating the immediate surround of a test patch can affect perceived lightness without affecting perceived illumination. This finding also falsifies the illumination-estimation hypothesis.     

T-junctions,apparent depth,and perceived lightness contrast

Observers performed lightness matches for physically equivalent gray targets of a simultaneous lightness contrast display and displays in which both targets were on the same background. Targets either shared a common line-texture pattern with their respective backgrounds or did not. Results indicate that when targets share a line-texture pattern with their respective backgrounds, a contrast effect is obtained. However, when the target's pattern is different than the background's pattern, perceived contrast is significantly reduced and the target appears as a separate 3-D entity. This result applies to both vertically and horizontally oriented displays, to targets that are increments or decrements, and to line-texture patterns that are black or white. Line patterns that are shared by targets and backgrounds result in T-junctions that provide occlusion information. We conclude that targets and backgrounds perceived to be on separate planes because of T-junctions are less likely to be perceptually grouped together and that their luminance values are less likely to be compared with one another.     

Testing the coplanar ratio hypothesis of lightness perception

What determines an object's lightness remains unclear, but it is generally thought that the ratios of its luminance to the luminance of other objects in a scene play a crucial role because these ratios allow the relative reflectance of each object to be estimated, providing all the objects are under the same illumination. Because objects that lie in the same plane are typically illuminated equally, it has been suggested that it is the luminance ratios between coplanar objects that primarily determine lightness (Gilchrist, 1977 Science 195 185-187; Gilchrist et al, 1999 Psychological Review 106 795-834). An alternative hypothesis is that perceived illumination differences can affect lightness directly. As the studies that provided evidence for the coplanar ratio hypothesis always varied the illumination and the coplanar relationships simultaneously, it is unclear which hypothesis is correct. I measured the influence of each factor separately and found that the perceived illumination differences have a greater effect on lightness.     

Impossible shadows and lightness constancy

The intersection between an illumination and a reflectance edge is characterised by the 'ratio-invariant' property, that is the luminance ratio of the regions under different illumination remains the same. In a CRT experiment, we shaped two areas, one surrounding the other, and simulated an illumination edge dividing them in two frames of illumination. The portion of the illumination edge standing on the surrounding area (labelled contextual background) was the contextual edge, while the portion standing on the enclosed area (labelled mediating background) was the mediating edge. On the mediating background, there were two patches, one per illumination frame. Observers were asked to adjust the luminance of the patch in bright illumination to equate the lightness of the other. We compared conditions in which the luminance ratio at the contextual edge could be (i) equal (possible shadow), or (ii) larger (impossible shadow) than that at the mediating edge. In addition, we manipulated the reflectance of the backgrounds. It could be higher for the contextual than for the mediating background; or, vice versa, lower for the contextual than for the mediating background. Results reveal that lightness constancy significantly increases when: (i) the luminance ratio at the contextual edge is larger than that at the mediating edge creating an impossible shadow, and (ii) the reflectance of the contextual background is lower than that of the mediating one. We interpret our results according to the albedo hypothesis, and suggest that the scission process is facilitated when the luminance ratio at the contextual edge is larger than that at the mediating edge and/or the reflectance of the including area is lower than that of the included one. This occurs even if the ratio-invariant property is violated.     

Perceived lightness, but not brightness, of achromatic surfaces depends on perceived depth information

Three experiments were conducted in an attempt to replicate and clarify Gilchrist's (1977, 1980) experiments on the effects of depth information on judgments of achromatic surface color. Gilchrist found that coplanarity, and not retinal adjacency, was the dominant factor in determining achromatic color matches. Because such matches can be made on the basis of either brightness or lightness, we obtained judgments of both qualities. Stereopsis was added to enhance the perceived depth effect of Gilchrist's display, which was otherwise simulated closely on a high-resolution CRT. The results for lightness followed the same pattern as those of Gilchrist, but were smaller in magnitude. This discrepancy may reflect reduced extraneous lighting effects in our displays. Our results therefore agree with related studies in suggesting that lightness matches are based on relationships among coplanar surfaces. Brightness matches, however, were not influenced by perceived depth.     

Hess and Pretori revisited: Resolution of some old contradictions

An early experiment by Hess and Pretori (1894) was replicated and modified in an attempt to determine why they failed to find the ratio principle later discovered by Wallach (1948). Separating the two surround-infield patterns by darkness made very little difference. However, allowing the observer to adjust the infield luminance (as in Wallach) rather than the surround luminance (as in Hess & Pretori) revealed some startling effects. At surround:infield luminance ratios greater than approximately 100∶1, there is no ratio effect; all infields appear equal and totally dark. Converging evidence is presented that Hess and Pretori’s data in this region actually represent surround-matching by the observers. Nor are ratio effects found with increments (infield brighter than surround). When free to match either infield luminances or ratios (by controlling infield luminance), observers match luminances. For decrements with ratios between 1∶1 and approximately 100∶1, lightness constancy and the ratio principle hold.