Metacontrast masking does not change with different display technologies: A comparison of CRT and LCD monitors

Abstract

Most vision labs have had to replace the formerly dominant CRT screens with LCDs and several studies have investigated whether changing the display type leads to changes in perceptual phenomena, since fundamental properties of the stimulation, e.g., the transition time between frames, differ between these different display technologies. While many phenomena have proven robust, Kihara et al. (2010) reported different metacontrast masking functions on LCDs compared to CRTs. This difference poses a challenge for the integration of new LCD-based findings with the established knowledge from studies with CRTs and requires theoretical accounts that consider the effects of different display types. However, before further conclusions can be drawn, the basic findings should be secured. Therefore, we tried to reproduce the display type effect by comparing metacontrast masking on an LCD and a CRT in two experiments. Our approach differs from the previous study by increasing the power and reliability of the measurements and carefully matching the two display types. In addition to display type, we varied target-mask stimulus-onset asynchrony (SOA) and stimulus-background polarity. Regardless of display type and polarity, we found the typical type-B masking functions. Evidence for a SOA-dependent display type effect in the black-on-white polarity condition from Experiment 1 was not replicated in Experiment 2. Overall, the results indicate that metacontrast masking effects on objective and subjective measurements, i.e., discriminatory sensitivity and phenomenological reports, do not vary significantly with display technologies. This lack of display effects is discussed in the context of current theories of metacontrast masking.

Keywords: CRT; Consciousness; Display technologies; LCD; Visual masking.

Fig. 1

Exemplary metacontrast masking functions of type A (A) and type B (B). Note. Masking functions are shown in terms of measurements of the discriminatory sensitivity d′ as a function of stimulus onset asynchrony (SOA). The figures are based on unpublished individual data from two different individuals and are shown only to illustrate the descriptions in the text

Fig. 2

Luminance signal change of a single pixel in a simple stimulus sequence. Note. Left column depicts a change from black to white to black; right column depicts a change from white to black to white. First row shows a corresponding stimulus sequence on the screen and the location of the measured pixels (red circle). Second and third rows give recordings of the actually measured luminance of the CRT and LCD devices used in the current study, respectively. Time is arbitrarily locked to stimulus onset. See text for further explanation and Supplementary Material S1 for details on the method for measuring the luminance signal

Fig. 3

Requested and actual signal changes for LCD and CRT displays. Note. A Requested stimulus onsets and offsets for a target presented for 20 ms (orange) followed by a mask with a SOA of 20 ms and a duration of 120 ms (blue). Note that the offset of the mask (duration of 120 ms) lies outside the displayed time interval. B–E Normalized luminance signals for the target-mask sequence in (A) as measured at the screen surface (orange and blue lines in the top panel, respectively) for all four display type × polarity conditions. The colored ribbons below each panel depict the polarity-dependent signal threshold state of the respective pixel at a given time. Thresholds were based on the ISO 9241–305 standard for response times. Full colored areas represent phases of super-threshold stimulus signal; white areas represent phases of supra-threshold stimulus signal, and pastel-colored areas represent transitional phases. Note that in the condition black on white (C and E), stimuli are ON when the luminance is low. Also note that for black on white stimuli on a CRT (E), considerations about the (non-displayed) background signal are crucial for interpretations about stimulus visibility (see text). Time is locked to target onset. Gray vertical lines mark frame boundaries of 10 ms. See text for further explanation and the method section and Supplementary Material S2 for details on the measurements

Fig. 4

Illustration of the luminance and spectral profiles of the two displays. Note. A Luminance of black and white areas in cd/m² for CRT (upper row) and LCD (lower row) for two polarity conditions: white stimuli on black background (left) and for black stimuli on white background (right). Each of the four boxes represents the respective visible display area and image configuration. Values within and around the stimulus patch represent the mean of 10 measurements at the respective location with standard deviation given in brackets. Overall means of the background color are given in the lower left corner of each display. Ambient luminance measured from distance of 1 m are given in the upper right corner of each display. B Spectral profiles of CRT (upper panel) and LCD (lower panel). Lines represent averages across 10 measurements, with standard deviations represented by a lighter shading of the same color

Fig. 5

Stimuli and trial sequence of the experiment. Note. A Congruent (upper row) and incongruent (lower row) combinations of target and mask stimuli for the two polarity conditions black stimuli on a white background (left) and white stimuli on black background (right). B Sequence of events in a single trial in the condition with black stimuli on a white background. Note that the computer recorded a response only during the last interval titled “until answer”

Fig. 6

Signal detection measures across SOA, display type, and polarity conditions for Experiment 1. Note. Dotted lines represent sensitivity values from CRT, solid lines from LCD. Black lines and points depict sensitivity values of black-on-white polarity, gray lines and points depict white-on-black polarity. A Averaged masking functions. B Individual masking functions, sorted by visually assessed observer type. Numbers in the upper right corner of each panel are added for ease of reference. Panels 1–5 show plain type-A observers. Panels 6–10 show participants with ambiguous observer type. Panels 11-–16 show plain type-B observers. Red numbers indicate participants with inconsistent masking types across conditions. C Averaged mask biases; positive values reflect a tendency to respond according to the shape of the mask, negative values reflect a tendency to respond contrary to the shape of the mask. D Individual mask biases, sorted in the same way as in (B). Red numbers indicate participants with inconsistent masking types across conditions (see B). All error bars represent within-subject standard error of the mean following Loftus and Masson (1994)

Fig. 7

Phenomenological reports in Experiment 1. Note. Each panel depicts whether or not a participant responded affirmatively to a question regarding an experience of the corresponding type of percept in the current session. Affirmation of seeing a specific type of percept is marked by a square. The color and pattern of the square indicate the display type and polarity. Negation of a type of percept is depicted by a white square. Black squares represent confirmation of a type of percept in a black-on-white polarity session, and gray squares in a white-on-black polarity session. Squares with a striped pattern represent confirmation of a type of percept in a CRT session, and homogeneous squares in an LCD session

Fig. 8

Signal detection measures across SOA and display type for Experiment 2. Note. Dotted lines represent sensitivity values from CRT, solid lines represent sensitivity measures from LCD. Stimuli were presented with the black-on-white polarity only. A Averaged masking functions. B Individual masking functions, sorted by visually assessed observer type. Numbers in the upper right corner of each panel are added for ease of reference. Panels 1–7 show plain type-A observers. Panels 8 and 9 show participants with ambiguous observer type. Panels 10–24 show plain type-B observers. Red numbers indicate participants with inconsistent masking types across conditions. C Averaged mask biases; positive values reflect a tendency to respond according to the mask shape, negative values reflect a tendency to respond contrary to the mask shape. D Individual mask biases, sorted in the same way as B. Red numbers indicate participants with inconsistent masking types across conditions. All error bars represent within-subject standard errors of the mean, following Loftus and Masson (1994)

Fig. 9

Phenomenological reports in Experiment 2. Note. Each panel depicts self-reported availability of one of the seven types of percept for each of the N = 24 participants. Confirmation of having seen a type of percept is marked by a square of the respective conditions pattern, negation by leaving that area white. Black squares with a striped pattern represent confirmation of a percept in a CRT session and pattern-less squares in an LCD session. The polarity was black-on-white for both display types

Fig. 10

Signal detection measures across SOA and display type for the unified sample. Note. Dotted lines represent sensitivity values from CRT, solid lines from LCD. Only data for the B/W condition are considered here. A Averaged masking functions. B Averaged mask biases; positive values reflect a tendency to respond according to the mask shape, negative values reflect a tendency to respond contrary to the mask shape. All error bars represent within-subject standard error of the mean, following Loftus and Masson (1994)