Visual masking

From Scholarpedia
Bruno G. Breitmeyer and Haluk Ogmen (2007), Scholarpedia, 2(7):3330. doi:10.4249/scholarpedia.3330 revision #182339 [link to/cite this article]
Jump to: navigation, search
Post-publication activity

Curator: Haluk Ogmen

Visual masking is the reduction or elimination of the visibility of one brief (≤ 50 ms) stimulus, called the “target”, by the presentation of a second brief stimulus, called the “mask”. Introduced near the end of the 19th and beginning of the 20th century (Exner, 1868; McDougal, 1904; Sherrington, 1897; Stigler, 1911) and extensively studied since then, masking, an interesting phenomenon in its own right, is a useful tool for exploring the dynamics of visual information processing (Breitmeyer & Öğmen, 2006). As a technique for studying the dynamic and microgenetic aspects of vision, masking rests on the following assumptions:

  • an interval, on the order of a few tens to two or three hundreds of milliseconds, is required from onset of a stimulus to its measurable effects on behavior or its conscious awareness;
  • the information conveyed by a stimulus is actively processed during this interval;
  • the processing can occur in several specialized, multi-level visual pathways;
  • the responses to the mask and the target can interact at specifiable levels of processing.

The simplest type of a mask consists of a large, spatially uniform increase or decrease of light. Such masks have been used to study rapid or transient light and dark adaptation (Crawford, 1947; Sperling, 1965). Because processes of light and dark adaptation are primarily retinal in nature (Riggs & Wooten, 1972; Werblin, 1971), masking by light is strong when the target and mask are presented to the same eye(s), but weak or absent when presented to separate eyes (Battersby & Wagman, 1962).

In the past four decades, most research efforts have shifted to studying post-retinal processes of visual masking, particularly those located at cortical levels. To accomplish this, the mask (and target) must consist of spatial patterns whose contours are defined by luminance or wavelength differences relative to the background on which they are presented. The outcomes of pattern masking studies depend on the choice of display, viewing condition, and the stimulus, timing, and task parameters.

Display parameters include the luminance and wavelength of the background on which stimuli are presented. Stimulus parameters include the shape, size, luminance, wavelength, retinal eccentricity, number, and degree of spatial overlap of the target and mask patterns.

Timing parameters include the duration of the target and mask stimuli as well as the time interval -- most commonly expressed in terms of stimulus onset asynchrony (SOA) -- separating the onsets of the target and mask stimuli. Forward masking occurs when the mask’s onset precedes the target’s onset; backward masking occurs when the mask’s onset follows the target’s onset. The special condition where the target and mask onsets are synchronous is called simultaneous masking.

Viewing conditions include monocular, binocular, or dichoptic viewing of the target and mask stimuli. In the former two conditions, the target and mask are presented to the same eye(s); in the latter condition, to separate eyes. The latter condition thus can best isolate postretinal, cortical levels of processing.

Task parameters define the criterion content (i.e., the particular information) used to respond to the target. Relying on subjective criteria, observers may be asked to make visibility ratings of the target, or luminance matching of the target to a comparison stimulus. With more objective criteria, the observer is asked to respond to some aspect of the target such as its brightness, color, or shape in a multi-alternative, forced-choice discrimination task. By systematically varying the SOA and one or more of the other parameters, one can infer from the results how, and at what stages of visual processing, the responses to the target and mask interact.

Figure 1: (a) Typical target and mask patterns (top) and the depiction of the percepts they generate (bottom) (b) Examples of paracontrast and metacontrast stimuli and a description of the attendant phenomenology.


Typical Methods, Findings, and Phenomenology of Visual Masking

Figure 1 shows several variants of pattern masking depending on the shape of the stimuli and the time interval separating their onsets. Although some forward masking effects are of theoretical interest, backward pattern has held center stage in studying the microgenesis of visual perception.

The left side of Figure 1a illustrates a target‑mask combination that is used in a pattern‑masking procedure termed masking by (random) noise or noise masking (Kinsbourne & Warrington, 1962). Here the mask is a collection of small squares that are randomly black or white (these are known as random dots). The target is an alphabetic letter. The mask elements, which spatially overlap the target, are designed to bear little, if any, structural relationship to the target contours. In this case, there is little masking as can be seen even when the mask and target are shown simultaneously.

The right side of Figure 1a illustrates a target‑mask combination that is used in a pattern‑masking procedure termed masking by structure or structure masking. Here the mask is a collection of short lines of different orientations. The target is another alphabetic letter. Now the mask elements are designed to resemble the contour properties (e.g., orientation, curvature) of the target. In this case, there is a lot of masking as can again be seen even when the mask and target are shown simultaneously.

Figure 1b shows a typical example of nonoverlapping target and mask stimuli used in forward paracontrast and backward metacontrast masking. Paracontrast and metacontrast are specific instantiations of structure masking, because contiguity and similarity of contours are shared between the target and mask.

These masking methods can yield a variety of functions relating masking magnitude to the SOA separating the target and mask. Figure 2 shows idealized monotonic and nonmonotonic, U-shaped, backward masking functions. On the Y axis is shown the visibility of the target. On the X axis is shown SOA in milliseconds: at zero, onset of mask (M) and target (T) are simultaneous. At negative values, onset of the mask precedes that of the target, yielding paracontrast. At positive values onset of the target precedes that of the mask, yielding metacontrast. In the monotonic case, masking is strongest (i.e., target visibility is lowest), when the target and mask are presented simultaneously and decreases as their SOA increases. In the U-shaped case, masking is weak at target-mask simultaneity, strongest at intermediate SOAs (30-80 ms) and weak again at large SOAs (> 150 ms). Monotonic functions can be obtained with either noise or structure masks when the stimuli are presented to one eye (monocular viewing) or to both (binocular viewing) and when the energy (duration × luminance contrast) of the mask to that of the target (M/T ratio) is larger than 1.0 (e.g., 3.0). Structure masking tends to be U-shaped a) when the M/T ratio is less than 1.0 (e.g., 0.4) or b) when target and mask are presented to separate eyes (dichoptic viewing) even when the M/T ratio energy ratio is larger than 1.0.

Figure 2: Depiction of generalized paracontrast and metacontrast functions. The X axis is the stimulus onset asynchrony (SOA) and the Y axis represents a measure of target visibility such as perceived brightness.

This is important for three reasons. First, dichoptic interactions occur at cortical levels of processing, levels on which object perception (while initiated by activity at retinal and other precortical levels of processing) ultimately also depends. Second, like masking by light, most of the backward noise and structure masking at short SOAs is due to the precortical, energy-dependent process of spatiotemporal integration (Turvey, 1973). Third, the other process, responsible for maximally suppressing or disrupting the visibility of the target at intermediate SOAs – called masking by interruption, occurs at cortical levels. Several past and current network models of backward masking by interruption rely on some form of neural inhibition, which in all cases reduces the strength of the neural activity generated by the target (Bridgeman, 1971; Francis, 1997; Öğmen, 1993; Weisstein, 1968). Recent findings indicate that such inhibition may interrupt reentrant (feedback) rather than feedforward activity in the cortical target-processing pathway (c.f. Breitmeyer, 2007).

With noise or structure masks and monocular or binocular viewing, forward masking, like backward masking, increases as the M/T ratio increases and decreases or is eliminated when the target and mask are presented to separate eyes (Breitmeyer, 1984). As shown in Figure 2, paracontrast (spatially nonoverlapping forward masking) produces a complicated function relating target visibility to SOA (Breitmeyer et al., 2006). Beginning at a paracontrast SOA of about 500 ms, as the leading mask’s onset approaches that of the target, the target visibility gradually decreases below its no-mask, baseline level, reaching a local minimum at a paracontrast SOA of -120 to -180 ms. At still smaller SOA magnitudes, target visibility increases, reaching a maximum above the baseline level at an SOA of -40 to -60 ms, before again decreasing and attaining a minimum at an SOA of -10 to -20 ms. This multiphasic nonmonotonic paracontrast masking function has been attributed to complex interactions among the following mask-produced effects: i) a long-latency, long-duration cortical inhibition (Berman et al., 1991; Nelson, 1991), ii) a short-latency, short-duration inhibition (Benardete & Kaplan, 1997; Connors, Malenka & Silva, 1988) and a cortical facilitation resulting from activation of the thalamic and midbrain reticular systems by the preceding mask (Bachmann, 1994).

Applications and Uses of Masking

The study of visual masking is informative in several ways. (1) The phenomenon of backward pattern masking, particularly the counter-intuitive finding that the mask can severely impede the visibility of the target even though the target is presented first, is interesting in its own right. Several competing models as well as qualitative explanations of this phenomenon have been proposed in the last two decades, and testing them of course requires the study of pattern masking (see Breitmeyer & Öğmen, 2006, Chapters 4 & 5). (2) Visual processing is a dynamic, temporally evolving phenomenon (VanRullen & Thorpe, 2001), and pattern masking can be useful for investigating the temporal sequence and levels of cortical feedforward and re-entrant (Lamme et al., 2002) information processing required for recognition of stimuli ranging from simple geometric forms to faces and complex scenes. (3) Higher-level visuo-cognitive processes such as perceptual grouping, visual imagery, and selective attention can modulate visual masking (Breitmeyer & Öğmen, 2006, Chapter 7). Moreover, masking plays a significant role in studies of the temporal parameters characterizing visual attention, especially in studies of the attentional blink in which rapid serial visual presentations (RSVP) of stimuli are used. (4) Visual masking has been used recently to explore its relation to masking by transcranial magnetic stimulation (TMS) (Breitmeyer et al., 2004).<review> And what was found?. </review> (5) Both visual and TMS masking are important methodological tools that are currently used to explore visual awareness and its implications for the controversial field of “subliminal perception”. The fact that information rendered unavailable to conscious report due to visual pattern masking can nonetheless influence a variety of motor, cognitive and emotional processes has been repeatedly established in recent years (Breitmeyer & Öğmen, 2006; Chapter 8). (6) Visual masking has been and continues to be used to study certain clinical anomalies related to vision and brain function, such as amblyopia, closed head injury, Parkinson’s disease, developmental dyslexia, mania, and schizophrenia as well as to nonclinical, specific-subject populations. Studies of visual masking may therefore provide a better understanding of perceptual anomalies and markers in any of these subject populations (Breitmeyer & Öğmen, 2006; Chapter 9).


Bachmann, T. (1994). Psychophysiology of Visual Masking: The Fine Structure of Conscious Experience. Commack, NY: Nova Science.

Battersby, W.S., & Wagman, I.H. (1962). Neural limitations of visual excitability. IV. Spatial determinants of retrochiasmal interaction. Am J Physiol 203, 359–365.

Berman, N.J., Douglas, R.J., Martin, K.A., & Whitteridge, D. (1991). Mechanisms of inhibition in cat visual cortex. J Physiol 440, 697–722.

Benardete, E.A., & Kaplan, E. (1997). The receptive field of the primate P retinal ganglion cell. I. Linear dynamics. Vis Neurosci 14, 169–85.

Breitmeyer, B. G. (2007). Visual masking: past accomplishments, present status, future developments. Adv Cogn Psychol 3.

Breitmeyer, B. G., & Öğmen, H. (2006). Visual Masking: Time Slices Through Conscious and Unconscious Vision. Oxford: Oxford University Press.

Breitmeyer, B.G., Ro, T., & Öğmen, H. (2004). A comparison of masking by visual and transcranial magnetic stimulation: implications for the study of conscious and unconscious visual processing. Conscious Cogn 13, 829–843.

Bridgeman, B. (1971). Metacontrast and lateral inhibition. Psychol Rev 78, 528–39.

Connors, B.W., Malenka, R.C., & Silva, L.R. (1988). Two inhibitory postsynaptic potentials, and GABAA and GABAB receptor-mediated responses in neocortex of rat and cat. J Physiol 406, 443-468.

Crawford, B.H. (1947). Visual adaptation in relation to brief conditioning stimuli. Proc R Soc Lond B 134, 283–302.

Exner, S. (1868). Über die zu einer Gesichtswahrnehmung nöthige Zeit. Wiener Sitzungsber Math-Naturwissensch Cl Kaiser Akad Wissensch 58 (Part 2), 601–632.

Francis, G. (1997). Cortical dynamics of lateral inhibition: metacontrast masking. Psychol Rev 104, 572–594.

Kinsbourne, M., & Warrington, E.K. (1962). The effect of an aftercoming random pattern on the perception of brief visual stimuli. Q J Exp Psychol 14, 223–234.

Lamme, V.A.F., Zipser, K., & Spekreijse, H. (2002). Masking interrupts figure–ground signals in V1. J Cogn Neurosci 14, 1044–1053.

McDougall, W. (1904). The sensations excited by a single momentary stimulation of the eye. Brit J Psychol 1, 78–113.

Nelson, S.B. (1991). Temporal interactions in the cat visual system. I: Orientation-selective suppression in the visual cortex. J Neurosci 11, 344–356.

Öğmen, H. (1993). A neural theory of retino-cortical dynamics. Neural Netw 6, 245–273.

Riggs, L. A., & Wooten, B. (1972). Electrical measures and physiological data on human vision. In Handbook of Sensory Physiology, Vol. VII/4. Visual Psychophysics (Eds. D. Jameson and L. H. Hurvich) pp. 690-731. New York: Springer.

Sherrington, C.S. (1897). On the reciprocal action in the retina as studied by means of some rotating discs. J Physiol 21, 33–54.

Sperling, G. (1965). Temporal and spatial visual masking. I. Masking by impulse flashes. J Opt Soc Am 55, 541–559.

Stigler, R. (1910). Chronotopische Studien über den Umgebungskontrast. Pflüs Arch Gesam Physiol 135, 365–435.

Turvey, M. (1973). On peripheral and central processes in vision: inferences from an information-processing analysis of masking with patterned stimuli. Psychol Rev 80, 1–52.

VanRullen, R., & Thorpe, S. J. (2001). The time course of visual processing: from early perception to decision making. J Cogn Neurosci 13, 454–61.

Weisstein, N. (1968). A Rashevsky–Landahl neural net: simulation of metacontrast. Psychol Rev 75, 494–521.

Werblin, F. (1971). Adaptation in the vertebrate retina: Intracellular recordings in Necturus. J Neurophysiol 34, 557-565.

Internal references

See Also

Personal tools

Focal areas