Masking with minimal contours: Selective inhibition with low spatial frequencies

Abstract

Paracontrast and metacontrast magnitudes were measured in a target identification task. A particular class of illusory contours is described that did not mask in the paracontrast condition but did show a large metacontrast magnitude. The discontinuity in the masking function is interpreted in terms of the Fourier decomposition of the visual scene that is performed by cells selectively responsive to discrete bands of spatial frequencies. The class of contours that we describe can only mask through inhibition generated by their low spatial frequencies. These results are consistent with recent models of masking based on two independent modes of inhibition-within sustained visual channels, and between sustained and transient visual channels.One of the most important developments in visual psychophysics has been the discovery and classification of cells that are selectively responsive to different bands of spatial frequencies. Psychophysical and electrophysiological measurements (for reviews, see Breitmeyer, 1984;De Valois & De Valois, 1980;Graham, 1981;Shapley & Lennie, 1985; chap. 4) give evidence that the visual scene is represented in early vision in terms of a coarse Fourier decomposition, and that different pieces of the spectrum are processed by independent channels. Recently these ideas have been introduced into a theoretical framework for visual pattern masking (Breitmeyer, 1984;Breitmeyer & Ganz, 1976).Pattern masking is a phenomenon in which perceived contrast or clarity of contour for one stimulus decreases as a result of the presentation of a second stimulus. In paracontrast (mask precedes target) and metacontrast (target precedes mask) masking, there is no overlap between mask and target contour. In this paradigm, it is generally possible to identify that a target has been shown, even when the target suffers contrast reduction and blurring (Breitmeyer, 1984). The loss of contour detail and contrast without complete elimination of stimulus detection suggests that masking operates by differentially affecting the processing of the high-spatial-frequency content of the stimulus. The central idea of Breitmeyer and Ganz's (1976) theory is that there are two independent modes of inhibition acting on the channels that process high spatial frequencies. According to this theory, paracontrast masking is achieved primarily by intrachannel inhibition caused by an antagonistic center-surround interaction within the The authors wish to thank Jack McArdle for several helpful discussions concerning the experimental design and analysis of the data. We also want to acknowledge conversations with Dennis Proffitt and the advice and encouragement he offered. D.G. was supported by NIH Grant HD0703tH>1 during the course of this work. Address correspondence to David Gilden, Department of Psychology, Gilmer Hall, University of Virginia, Charlottesville, VA 22903-2477. receptive field of a single cell. It is theorized that metaeontrast masking is produced by center-surround interactions (as in paracontrast) and by an addition...

Role of eccentricity and size in the global precedence effect.

“…They found nonsignificant differences for response times (RTs) to the small letters in the control and experimental conditions. This showed that the slower RTs to the small letters in the compound stimuli did not result from masking by neighboring elements, as suggested by some previous studies (e.g., Gilden, MacDonald, & Lasaga, 1988; Johnson, 1986; Pomerantz, 1981; Wenderoth & Beh, 1977).…”

“…Nonsignificant differences for RTs to the small letters in the control and experimental conditions ruled out the possible effect of masking which may increase the reaction time to any local properties by the other neighbouring elements. This effect has been suggested in some studies (e.g., Gilden, MacDonald, & Lasaga, 1988;Johnson, 1986;Pomerantz, 1981;Wenderoth & Beh, 1977). For example, Johnson (1986) proposed that his subjects were able to recognise the presence or absence of a letter faster when that letter or a nontarget letter (e.g., B or L) was presented individually than in a condition in which a redundant (repeated) display of them was presented (e.g., BBBBB or LLLLL).…”

Comment on "Competition for consciousness among visual events: The psychophysics of reentrant visual processes" (Di Lollo, Enns & Rensink, 2000).

“…However, as Enns and Di Lollo (1997) and Di noted, and as Figure 1A indicates, strong masking can occur when the mask is physically very small (e.g., four dots). This was not a new finding, as some masking from sparse masks was found in studies by Werner (1935), Sherrick andDember (1970), andGilden, Mac-Donald, andLasaga (1988), but in combination with other properties it suggested to Enns and Di Lollo (1997) a new type of masking.…”

“…This kind of mask was used to produce the data in Figure 1a. This was not an entirely new finding as some masking from sparse masks occurred in studies by Werner (1935), Sherrick and Dember (1970) and Gilden, MacDonald and Lasaga (1988); but in combination with other properties it suggested to Enns and Di Lollo (1997) a new type of masking.…”

Metacontrast masking is processed before grapheme–color synesthesia

“…Evidence of metacontrast has been found in single cells of V1 in both the cat (Bridgeman, 1975) and the monkey (Bridgeman, 1980). However, metacontrast can also be obtained using illusory or subjective contours (Gilden, MacDonald, & Lasaga, 1988), and area V2 is the first processing site for subjective contours (Petry, 1987). If metacontrast were to be unaffected by synesthesia, we could therefore further narrow the seat of grapheme-color synesthesia as being beyond the level of V2.…”