Open Access
Article  |   August 2023
Afterimage duration depends on how deeply invisible stimuli were suppressed
Author Affiliations
  • Motomi Shimizu
    Department of Psychology, Graduate School of Humanities, Chiba University, Inage-ku, Chiba-shi, Chiba, Japan
    [email protected]
  • Eiji Kimura
    Department of Psychology, Graduate School of Humanities, Chiba University, Inage-ku, Chiba-shi, Chiba, Japan
    [email protected]
Journal of Vision August 2023, Vol.23, 1. doi:https://doi.org/10.1167/jov.23.8.1
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Motomi Shimizu, Eiji Kimura; Afterimage duration depends on how deeply invisible stimuli were suppressed. Journal of Vision 2023;23(8):1. https://doi.org/10.1167/jov.23.8.1.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Quantifying visual responses to stimuli that are outside of awareness is a critical task for the study of visual consciousness. The current study psychophysically investigated whether afterimages reflect visual responses to stimuli that are not consciously visible throughout adaptation due to interocular suppression. A Gabor adaptor was presented to one eye of the observer, and a counterphase-flickering Gabor suppressor was presented to the other eye, thereby rendering the adaptor invisible during adaptation. To manipulate the depth of the suppression of the invisible adaptor, we varied the orientation difference between the adaptor and suppressor. We found that, even though the adaptor was not visible during adaptation, the afterimage duration varied depending on the orientation selectivity of interocular suppression. The duration was the shortest when the orientations of the adaptor and suppressor were identical and lengthened when the orientation differences increased. This finding could not be explained by confounding factors such as potential changes in contrast sensitivity that were caused by the suppressor. Our findings suggest that the magnitude of visual responses to stimuli suppressed below the threshold of awareness can be measured using the afterimage duration. Afterimages could be an effective tool for quantifying visual responses, irrespective of observers’ conscious awareness of a presented stimulus.

Introduction
When dissimilar images are presented to the left and right eyes, interocular suppression is induced, in which one of the images becomes perceptually dominant while the other is suppressed. Interocular suppression has been examined in various dichoptic presentation techniques, such as binocular rivalry and continuous flash suppression. When dichoptic images are equally potent in the competition for dominance, a binocular rivalry is observed, wherein the percept alternates between monocular images every few seconds (Blake & Logothetis, 2002). A more robust and stable suppression can be induced using a recently developed continuous flash suppression (CFS) technique (Tsuchiya & Koch, 2005). In CFS, a dynamic, high-contrast, and contour-rich pattern presented to one eye becomes exclusively dominant, rendering a lower-contrast stimulus presented to the other eye invisible for an extended time, possibly up to several minutes (Tsuchiya & Koch, 2005). 
The steady and prolonged suppression produced by CFS can be critical for investigating the visual mechanisms that mediate interocular suppression and unconscious visual perception in general. For example, Tsuchiya and Koch (2005) demonstrated that CFS-induced suppression could significantly interfere with the afterimage formation. They showed that, when an adapter (e.g., a Gabor patch) presented to one eye was suppressed and rendered invisible by a dynamic Mondrian pattern presented to the other eye, the negative afterimage of the adapter could be reduced by half in terms of perceived contrast. Afterimage formation had been considered a retinal process (Brindley, 1962; Loomis, 1972; Loomis, 1978), and previous studies using binocular rivalry suggested that interocular suppression did not affect the strength of afterimage (Lack, 1978). Tsuchiya and Koch (2005) argued that the afterimage formation is at least partially mediated by cortical processing (see also Gilroy & Blake, 2005; Shimojo, Kamitani, & Nishida, 2001; Suzuki & Grabowecky, 2003) and that more robust and stable suppression caused by CFS revealed the cortical contribution. 
The afterimage strength, defined in terms of perceived contrast or duration, can be a good measure to investigate the effects of consciousness on visual processing. It is well known that a longer and more vivid afterimage can be induced after the offset of an adaptor with higher contrast or longer duration (Georgeson & Turner, 1985; Kelly & Martinez-Uriegas, 1993; McLelland, Baker, Ahmed, & Bair, 2010). Thus, the afterimage strength can be used to estimate the magnitude of the visual responses to the adaptor (Suzuki & Grabowecky, 2003). This application of the afterimage is found to be preserved even if the adaptor is suppressed from awareness due to interocular suppression induced by CFS (Tsuchiya & Koch, 2005; van Boxtel, Tsuchiya, & Koch, 2010) or binocular rivalry (Gilroy & Blake, 2005). Tsuchiya and Koch (2005) showed that afterimage strength positively correlated with the visibility of the adaptor by manipulating the spatial frequency and contrast of the adaptor as well as the contrast of the suppressor. The less the adaptor was visible during the adaptation, the more reduced the resulting afterimage became. One advantage of the afterimage study is that it does not require an additional probe to examine the effects of the suppressed stimulus. 
As in Tsuchiya and Koch (2005), previous studies investigating the effects of interocular suppression have examined the correlation between the degree of adaptor visibility and the strength of the resulting afterimage (Gilroy & Blake, 2005; van Boxtel et al., 2010). However, the afterimage strength can also be used as a measure of unconscious visual processing. Using binocular rivalry, Gilroy and Blake (2005) showed that, even in the complete absence of visual awareness, the afterimage remains for a longer period when the adaptor has a higher physical contrast. This finding suggests that the visual response to the adaptor can vary in magnitude, even though it is below the threshold of awareness, and that the differential visual response can be reflected in the afterimage duration. Another study also showed that the afterimage duration decreased with the physical contrast of the suppressor in CFS (van Boxtel et al., 2010). 
This study investigated whether the afterimage duration reflects the orientation selectivity of interocular suppression, even when observers cannot detect the adaptor at all. Strong interocular suppression was produced by presenting a counterphase-flickering Gabor stimulus with full contrast (suppressor) to one eye. The resulting suppression was so strong that an adapting Gabor stimulus with full contrast (adaptor) presented to the other eye was rendered invisible for at least several seconds, irrespective of their relative orientation. Several seconds were sufficient to induce a visible afterimage of the adaptor. In our experiments, the physical contrast of both the suppressor and adaptor was kept constant, and only their relative orientation was varied to manipulate the depth of interocular suppression. 
Previous studies showed that the orientation difference between the Gabor or grating stimuli presented to the two eyes is an important factor in determining the depth of the interocular suppression. Orientation selectivity has been demonstrated in binocular rivalry (Baker & Graf, 2009; Ling & Blake, 2009; Stuit, Cass, Paffen, & Alais, 2009) and, more recently, in CFS (Han & Alais, 2018; Han, Blake, & Alais, 2018; Yang & Blake, 2012). Thus, we assumed that the depth of the interocular suppression could vary with the relative orientation between the suppressor and adaptor. 
Experiment 1
Experiment 1 investigated whether interocular suppression affects afterimage formation in an orientation-selective fashion. Suppressing and adapting Gabor stimuli (suppressor and adaptor, respectively) were presented to different eyes, and their relative orientation was manipulated. The adaptor was made invisible for the entire period of adaptation by means of interocular suppression. Then, the duration of the afterimage induced by the adaptor was measured as a function of the orientation difference. 
Methods
Observers
Two groups of observers were independently recruited for Experiments 1a and 1b. In all, nine observers participated in Experiment 1a, and 10 participated in Experiment 1b. One author (MS) participated in both experiments. All observers had normal or corrected-to-normal visual acuity. All observers, except for MS, were naïve regarding the purpose of the experiments. Before the main experiment, each observer's dominant eye was determined by measuring perceptual dominance time in conventional binocular rivalry (rivalry with orthogonal Gabor patches, tested for 1 minute in five trials). Before the experiment, the observers who participated in this and the following experiments provided informed consent after a thorough explanation of all procedures. The experiments were conducted in accordance with the tenets of the Declaration of Helsinki. 
The sample size and the number of trials were examined based on the subsampling simulation (Baker, Vilidaite, Lygo, Smith, Flack, Gouws, & Andrews, 2021) using the data from our preliminary experiment (six observers and 20 trials for five orientation difference conditions and 80 trials for the control [no suppression] condition). We estimated an effect size of ηp2 = 0.67 for the main effect of the orientation difference. Then, we conducted an analysis of variance (ANOVA) on a randomly subsampled dataset 10,000 times and counted the number of times the main effect was statistically significant. The results confirmed that the statistical power exceeded 90% with a sample size of four observers and 10 trials for each condition. While taking these results into account, we followed laboratory practices and aimed to recruit 10 observers in each experiment. We set the number of trials to 20 to collect a sufficient number of trials, even when the observers reported insufficient suppression on some trials. 
Apparatus
In this study, we used two sets of equipment in Experiments 1a and 1b. In Experiment 1a, the stimuli were generated using a VSG2/5 graphics card with a 15-bit color depth (Cambridge Research Systems, Ltd., Kent, UK) and MATLAB (MathWorks, Natick, MA). The stimuli were displayed on a 21-inch color cathode-ray tube (CRT) monitor (GDM-F500R; Sony Corporation, Tokyo, Japan) that had a spatial resolution of 1280 × 962 pixels and a refresh rate of 80 Hz. The observer's responses were collected using a CT3 response box (Cambridge Research Systems). In Experiment 1b, the stimuli were generated using MATLAB and Psychophysics Toolbox 3 (Brainard, 1997; Kleiner, Brainard, & Pelli, 2007; Pelli, 1997). The stimuli were displayed on a 19-inch color CRT monitor (T766; EIZO Corporation, Ishikawa, Japan) that had a spatial resolution of 1280 × 1024 pixels and a refresh rate of 85 Hz. The intensity of each phosphor was manipulated with 8-bit resolution. For each monitor, spectroradiometric calibration was performed on three phosphors of each monitor with a CS-1000 spectroradiometer (Konica Minolta, Tokyo, Japan) and an LS-100 luminance meter (Konica Minolta). The stimuli for the left and right eye were presented side by side on the CRT monitor. The observers viewed the stimuli through a mirror stereoscope that was placed in front of their eyes so that the left-side and right-side stimuli on the monitor were projected to the left and right eye, respectively. The observer's head was stabilized using a chin rest to maintain an optical distance of 92 cm (Experiment 1a) or 93 cm (Experiment 1b). The experiments were conducted in a dark room. 
Stimuli
The suppressor used was a dynamic Gabor patch that was presented to the observer's dominant eye (Figure 1). The Gabor patch was the product of a sinusoidal grating (∼2.5 cpd, 100% Michelson contrast, and mean luminance of 18.0 cd/m2) and a Gaussian window (σ = 0.22°). The polarity of the suppressor abruptly reversed approximately 10 times per second, corresponding to counterphase flickering at 5 Hz (Experiment 1a) or 4.7 Hz (Experiment 1b). The orientation and phase of the suppressor were randomly selected for each trial to reduce the effects of the spatial anisotropy and local phase adaptation. The adaptor, which was presented to the observer's non-dominant eye, was a static Gabor patch, whose spatial properties were the same as those of the suppressor, apart from orientation. The orientation of the adaptor relative to the suppressor was determined according to the orientation difference condition described later in the Conditions section. 
Figure 1.
 
Schematic diagram of the stimulus sequence in Experiments 1a and 1b. A suppressor was presented to the observer's dominant eye (DE, left column), and an adaptor was presented to the non-dominant eye (NDE, middle column). A negative afterimage of the adaptor was seen on the blank display after the offset of the dichoptic stimuli. A typical percept is illustrated in the right column of the figure. A random noise mask was presented after the observer's response only in Experiment 1b.
Figure 1.
 
Schematic diagram of the stimulus sequence in Experiments 1a and 1b. A suppressor was presented to the observer's dominant eye (DE, left column), and an adaptor was presented to the non-dominant eye (NDE, middle column). A negative afterimage of the adaptor was seen on the blank display after the offset of the dichoptic stimuli. A typical percept is illustrated in the right column of the figure. A random noise mask was presented after the observer's response only in Experiment 1b.
Both the suppressor and adaptor were presented at the center of a uniform circular achromatic background (x = 0.313, y = 0.329, 18.0 cd/m2), subtending 2.3° (Experiment 1a) or 4.6° (Experiment 1b) in diameter, displayed on the left and right sides of the monitor. The two backgrounds were presented to each eye to support binocular fusion. The observers were instructed to fixate on the center of the background during the trial. 
In Experiment 1b, at the end of each trial, a dynamic achromatic mask was presented for 1 second to prevent carryover effects across consecutive trials. This mask consisted of an array of pixels with randomly determined luminances. It subtended 1.62° × 1.62° and was refreshed every frame (85 Hz). 
Procedures
The same procedure for presenting the stimuli was used in both Experiments 1a and 1b. A central fixation dot (0.1° in diameter), together with a beep sound, indicated that a trial was ready to begin. The observer's key press initiated the stimulus sequence. The fixation dot disappeared 1 second after the key press, and then the suppressor was presented alone to establish its perceptual dominance (Figure 1). The adaptor was gradually introduced according to a Gaussian temporal profile from 0% to 100% over 0.5 second, and it was maintained at that maximal value for 2.5 seconds, resulting in a total exposure time of 3 seconds. Although the adaptor had high maximal contrast, this type of dichoptic presentation generally resulted in exclusive dominance of the suppressor throughout the presentation. The abrupt offset of the stimuli usually generated a negative afterimage of the adaptor. The offset was perceptually salient, and it allowed the observer to distinguish the physically presented stimuli from the afterimage. We confirmed that, when the suppressor was presented alone, it did not induce any afterimage at all in a preliminary experiment. 
The observer was instructed to press one of the three buttons according to his or her percept. If an afterimage was seen, the first button was to be pressed for as long as the afterimage was visible. If, however, no afterimage was seen, the second button was to be pressed. If the suppression was not strong enough, and at least one portion of the adaptor was perceived during the stimulus presentation, the observer reported it using the third button. After the observer's response, a dynamic random noise mask was presented for 1 second to reduce residual aftereffects in Experiment 1b (Figure 1). The trials for which the observer reported insufficient suppression were discarded and were not used for analysis of the afterimage duration. The percentage of the discarded trials averaged across different observers was 11.3% (variance = 12.7%), ranging from a minimum of 0.0% to a maximum of 45.0%. The percentage of the discarded trials did not change systematically as a function of the orientation difference between the suppressor and adaptor (see Supplementary Material). 
At the beginning of the experiment, the observers were dark adapted for at least 5 minutes and then pre-adapted to the circular background for another 2 minutes. Observers were allowed as many practice trials as they needed to familiarize themselves with the stimuli and the task before the main experiment. 
Conditions
In the suppression condition, the relative orientation between the suppressor and adaptor was varied from 0° (identically oriented) to 90° (orthogonally oriented) in five steps (i.e., 0°, 22.5°, 45°, 67.5°, and 90°). The absolute orientation of the suppressor was randomly selected for every trial. The five orientation conditions were tested 20 times in a pseudorandom order for each observer. The duration of the afterimage was also measured in the no-suppression (control) condition, and this was used as a baseline. The procedure of presenting the adaptor in the control condition was identical to that used in the suppression condition, except for the absence of the suppressor. Thus, the adaptor was visible to the observer in the control condition for the entire adaptation period. In Experiment 1a, the suppression and control conditions were conducted in different blocks; the control condition, which was also administered in 20 trials, was tested before and after the suppression condition. In Experiment 1b, all conditions, including the control condition, were randomly interleaved within a single block and repeated 20 times. 
Data analysis
The data in this and the following experiments were analyzed with both frequentist statistics using ANOVA and Bayesian statistics using linear mixed-effects models (Baayen, Davidson, & Bates, 2008). The latter analytical approach is particularly advantageous for handling repeated measures, as it avoids the potential pitfalls of data aggregation. Aggregation may result in a loss of information and potentially mask individual differences, leading to a skewed representation of the underlying reality. By utilizing all trials for each observer, the mixed-effects model provides a more accurate and reliable estimation of the effects under investigation. 
We followed the method detailed in Hesselmann, Darcy, Rothkirch, and Sterzer (2018) for reporting Bayes factors (BFs). Initially, we extracted the model exhibiting the highest BF (i.e., the best model) and assigned its BF as one. Subsequently, we recalibrated the BFs of all other models in relation to this best model; consequently, the BF of each model indicates the extent to which the data are more consistent with the best model than with the model being considered. We have made our data and R code in this and the following experiments available at an online repository (https://osf.io/sn2ap/). 
Results
Before the analysis, one observer in Experiment 1b was excluded from the analysis because this observer never reported the afterimage in the control condition, and the afterimage duration was quite short (<0.3 second) even when the afterimage was reported in the suppression condition. In the analysis of Experiment 1a, the results of the two trial blocks for the control condition were combined before averaging. 
The mean afterimage duration is shown in Figure 2 as a function of the orientation difference between the suppressor and adaptor. The results were analyzed in a mixed ANOVA comprised of one between-subjects factor (Experiment 1a vs. 1b) and one within-subjects factor (five orientation differences and control) using the anovakun function in R (R Foundation for Statistical Computing, Vienna, Austria) (Iseki, 2023). Mauchly's test was used for the sphericity assumption. If the assumption was not met, the Greenhouse–Geisser correction was applied. The analysis revealed a significant main effect of orientation difference, F(3.01, 45.2) = 17.96, p < 0.001, ηp2 = 0.54, but the main effect of experiment and the interaction were not significant, F(1, 15) = 0.51, p = 0.49, ηp2 = 0.03; F(3.01, 45.2) = 0.81, p = 0.49, ηp2 = 0.05, respectively. A post hoc analysis for multiple comparisons with Bonferroni correction (α = 0.05/15) showed that the afterimage duration in the 0° condition was shorter than that in the 45°, 67.5°, 90°, or control conditions, t(15) = 5.54, p < 0.001; t(15) = 7.25, p < 0.001; t(15) = 7.27, p < 0.001; t(15) = 3.56, p = 0.003, respectively. Notably, a significant difference from the control condition was only observed in the 0° condition, t(15) = 3.56, p = 0.003. 
Figure 2.
 
Results of Experiments 1a and 1b. Two experiments were conducted with a basically identical procedure but different sets of equipment. The afterimage duration was plotted as a function of the orientation difference between the suppressor and adaptor (open symbols). The filled symbols represent the results in the control (no-suppression) condition. The black symbols and lines show the mean afterimage duration averaged across Experiments 1a and 1b. The error bars indicate ±1 SEM across different observers. The asterisks denote the conditions in which the afterimage duration was significantly longer than that in the 0° condition.
Figure 2.
 
Results of Experiments 1a and 1b. Two experiments were conducted with a basically identical procedure but different sets of equipment. The afterimage duration was plotted as a function of the orientation difference between the suppressor and adaptor (open symbols). The filled symbols represent the results in the control (no-suppression) condition. The black symbols and lines show the mean afterimage duration averaged across Experiments 1a and 1b. The error bars indicate ±1 SEM across different observers. The asterisks denote the conditions in which the afterimage duration was significantly longer than that in the 0° condition.
The afterimage durations were also analyzed by a linear mixed-effects model. In order to test the two main effects and their interaction, we compared models based on Bayes factors calculated using the BayesFactor package in R (Morey & Rouder, 2022). We calculated Bayesian linear mixed-effects models with a random effect of participants (participant ID [PID]) as a random intercept, as well as the fixed effects of Condition (five orientation differences and control) and Experiment (Experiment 1a vs. 1b). The fixed effects were the same as those used in the ANOVA. The results are summarized in Table 1. The model that predicted the data best (i.e., the best model) was the one with the two fixed effects of Condition and Experiment but without their interaction term. This model was favored over the interaction model (BF = 6.70, error = 2.9%). The other models were much less consistent with the data than the best model (BFs > 100). Multiple comparisons were performed using the emmeans package in R (Lenth, 2023). Though the overall findings were consistent with those obtained through the ANOVA, the linear mixed-effects model revealed a significant reduction of the afterimage duration in the control condition compared to those of the 67° condition, t(2134) = 4.742, p < 0.001, and the 90° condition, t(2134) = 6.188, p < 0.001. 
Table 1.
 
Bayes factor analysis of the afterimage duration using linear mixed-effects models. PID, participant ID.
Table 1.
 
Bayes factor analysis of the afterimage duration using linear mixed-effects models. PID, participant ID.
Discussion
The results clearly showed that the orientation difference between the suppressor and adaptor affected the afterimage duration even when the adaptor was out of awareness. The duration was the longest when the orientation difference was 90°, and it decreased as the orientation difference decreased. It was the shortest when the orientation of the two stimuli was identical (0°). The essentially same results for the different sets of equipment (Experiments 1a and 1b) indicate the robustness of the findings. The observed orientation selectivity of the afterimage duration was consistent with the interpretation that the afterimage duration reflected the orientation selectivity of the interocular suppression. 
However, an alternative explanation is also possible—namely, that the highly visible suppressor produced an adaptation effect in the form of a reduction in contrast sensitivity in the contralateral as well as the ipsilateral eye. This effect then modulated the afterimage duration. It is well known that a reduction in contrast sensitivity follows the offset of a strong spatial pattern, and this effect partially transfers interocularly (Baker & Meese, 2012; Bjørklund & Magnussen, 1981). Moreover, a recent study found that this reduction in contrast sensitivity could affect the visibility of the afterimage (Brascamp, van Boxtel, Knapen, & Blake, 2010). Thus, it is conceivable that the suppressor reduced contrast sensitivity in the two eyes and decreased the duration of the afterimage that was induced by the contralateral adaptor. Furthermore, if the reduction was selective to the orientation difference between the suppressor and adaptor, it could explain the orientation selectivity of the afterimage duration. Then, the afterimage duration may not have depended on the orientation selectivity of the interocular suppression. We tested this interpretation in Experiment 2
Additionally, one aspect of the results shown in Figure 2 may require further examination. The afterimage duration in the suppression condition was not necessarily shorter than that in the control condition. We naturally predicted that the afterimage duration in the control condition would be the longest if the suppressor exerted a suppressive effect on the afterimage formation. The results appear to go against this prediction, although most of the differences in the afterimage duration between the suppression and control conditions were not statistically significant. We address this in Experiment 3
Experiment 2
In Experiment 2, we investigated how strongly a counterphase-flickering Gabor stimulus, the same as that used as the suppressor in Experiment 1, would influence the detection threshold of a low-contrast target stimulus that was presented subsequently to the contralateral eye. We also examined whether or not a possible reduction in contrast sensitivity was orientation selective. 
Methods
Observers
Ten observers, including one of the authors (MS), participated in Experiment 2. All observers had normal or corrected-to-normal visual acuity. To determine the sample size and the number of trials, we conducted a preliminary experiment with five observers using the same procedure as in Experiment 2 and found that the main effect of the stimulus condition (0°, 90°, and control) on the contrast threshold was not significant, F(2, 8) = 0.83, ηp2 = 0.17. Thus, we could not use the same subsampling simulation (Baker et al., 2021) as in Experiment 1. Instead, we used the simr package in R (Green & MacLeod, 2016) to perform a power analysis, which allowed us to manually set the effect size. We adopted a conservative approach by setting the effect size to 0.1 log units. This value was small compared to the step size of 0.05 log units used in the up–down method for threshold measurement. In fact, it might have been too small to account for the change in afterimage duration found in Experiment 1. Our simulation showed that a sample size of six observers and two repetitions of measurements would be needed to achieve 80% power. By increasing the sample size to ten observers, as in Experiment 1, the statistical power rose to approximately 97%. 
Apparatus
A set of equipment different from those used in Experiments 1a and 1b was utilized in Experiment 2. The stimuli were generated using a ViSaGe graphics card with a 14-bit color depth (Cambridge Research System) and MATLAB. The Psychophysics Toolbox 3 for MATLAB was also used in the phosphor calibrations and colorimetric calculations (Brainard, 1997; Kleiner et al., 2007; Pelli, 1997). The stimuli were displayed on a 21-inch color CRT monitor (GDM-F520; Sony Corporation, Tokyo, Japan) that had a spatial resolution of 1280 × 960 pixels and a refresh rate of 90 Hz. The observers viewed the stimuli through a mirror stereoscope, and the observer's head was stabilized using a chin and forehead rest to maintain an optical distance of 100 cm. 
Stimulus and conditions
The spatial and temporal properties of the dynamic stimulus were the same as those used in Experiment 1a. The target stimulus was a Gabor patch that was identical to the adaptor used in Experiments 1a and 1b. Its orientation was either identical (0°) or orthogonal (90°) to the dynamic stimulus. As in Experiments 1a and 1b, circular gray backgrounds were used, but their diameters were reduced to 3.9° here. 
Procedures
In each trial, after the observer pressed a key, the dynamic stimulus was presented on both sides of the fixation point in the observer's dominant eye for 3 seconds (Figure 3). The center of the dynamic stimuli was 1° away from the fixation point. Subsequently, the target was presented to the observer's non-dominant eye at the same location as the dynamic stimulus, either on the left or the right of the fixation point. The target contrast was linearly increased to the maximum over 250 ms, mimicking the appearance of a typical afterimage, and it was thereafter maintained at that value for 1 second. The target was followed by a dynamic mask (1.62° × 1.62°) for 1 second. The observer's task was to indicate to which side of the fixation point the target was presented. No feedback was given. 
Figure 3.
 
Schematic diagram of the stimulus sequence in Experiment 2. Two dynamic stimuli were presented to the dominant eye (DE, left column), and the target was presented to the non-dominant eye (NDE, middle column). A typical percept is illustrated in the right column of the figure.
Figure 3.
 
Schematic diagram of the stimulus sequence in Experiment 2. Two dynamic stimuli were presented to the dominant eye (DE, left column), and the target was presented to the non-dominant eye (NDE, middle column). A typical percept is illustrated in the right column of the figure.
Thresholds were estimated with the use of a double random staircase procedure. Two independent staircases were randomly executed in a single measurement. For each staircase, the target contrast was changed from well above or below the detection threshold that had been preliminarily estimated. From there, the contrast was varied in each trial, depending on the observer's response in the previous trial. An incorrect response increased the target contrast by 0.1 log units, and two successive correct responses reduced the contrast by 0.1 log units at first and then by 0.05 log units once an incorrect response was observed. The detection threshold was calculated by averaging over the last six of 10 reversals in a single staircase. The measurement was repeated twice for each condition in different sessions. The detection threshold was also measured in the control condition, where no dynamic stimulus was presented, and taken as a baseline. In total, four detection thresholds (two staircases × two repetitions) were averaged for each observer. 
Results and discussion
Figure 4 shows the mean threshold across different observers with individual data. We conducted a one-way repeated-measures ANOVA and found no significant effect of the stimulus condition (0°, 90°, and control) on the contrast threshold, F(2, 18) = 0.453, p = 0.64, ηp2 = 0.05. This result is not consistent with the account that pre-exposure to a strong dynamic stimulus in one eye affects the contrast threshold in the other. No interocular transfer of the adaptation effect has often been found when low spatial frequency gratings are used (Apthorp, Griffiths, Alais, & Cass, 2017; Baker & Meese, 2012; Cass, Johnson, Bex, & Alais, 2012). 
Figure 4.
 
Results of Experiment 2. The mean contrast threshold in the log unit is plotted for the suppression (open squares) and control (filled square) conditions. The thresholds for individual observers are also shown using gray circles. Error bars denote ±1 SEM across different observers.
Figure 4.
 
Results of Experiment 2. The mean contrast threshold in the log unit is plotted for the suppression (open squares) and control (filled square) conditions. The thresholds for individual observers are also shown using gray circles. Error bars denote ±1 SEM across different observers.
Consistent with this result, the Bayesian linear mixed-effects model analysis showed that the best model included only random effects (i.e., participants, “PID,” and daily sessions, “Day”) (Table 2). This model was favored over the model with a fixed effect of the stimulus condition, “Condition” (BF = 4.003), supporting the null hypothesis of no difference. 
Table 2.
 
Bayes factor analysis of the contrast threshold using linear mixed-effects models. PID, participant ID.
Table 2.
 
Bayes factor analysis of the contrast threshold using linear mixed-effects models. PID, participant ID.
Table 3.
 
Bayes factor analysis of the results of Experiment 3 using linear mixed-effects models. PID, participant ID; SF, spatial frequency.
Table 3.
 
Bayes factor analysis of the results of Experiment 3 using linear mixed-effects models. PID, participant ID; SF, spatial frequency.
Overall, it seems difficult to account for the changes in the afterimage duration found in Experiment 1 by assuming that dynamic high-contrast suppressor dichoptically reduced the contrast sensitivity in an orientation-selective fashion. 
Experiment 3
Experiment 3 aimed to more closely examine the finding in Experiment 1 that the afterimage duration was not necessarily shorter after the adaptor was suppressed. We investigated the effects of the spatial frequency of the suppressor and adaptor on interocular suppression. Previous studies have shown that spatial frequency is an important determinant of interocular suppression, and suppression is more prominent when the suppressor is composed of low spatial frequency (Yang & Blake, 2012). Moreover, other studies have shown that the effects of visual awareness on the afterimage duration depend on the spatial frequency of the stimulus (Brascamp et al., 2010; Tsuchiya & Koch, 2005; van Boxtel, 2017). Brascamp et al. (2010) specifically showed that awareness increased the afterimage duration at low spatial frequencies, and it decreased the duration at high frequencies. Thus, the duration of the afterimage induced by the unconscious adaptor may change with the spatial frequency of the adaptor. 
Another possible explanation for the finding in Experiment 1 could be the effect of attention. It has been demonstrated that attending to a stimulus can result in a weaker afterimage (Suzuki & Grabowecky, 2003; van Boxtel, 2017; van Boxtel et al., 2010). In the control condition of Experiment 1, the adaptor was visible and presumably drew the observer's attention, in contrast to the invisible adaptor in the suppression condition. Consequently, the attended adaptor in the control condition may have resulted in a shorter afterimage. If this is the primary cause of the finding in Experiment 1, the strength of the afterimage would not vary with spatial frequency, as shown in Brascamp et al. (2010). Experiment 3 tested these predictions by comparing the strength of afterimages produced by adaptors with different spatial frequencies in both the suppression and control conditions. 
Methods
Observers
In all, 10 observers participated in Experiment 3. The sample size for Experiment 3 was set to 10 observers to ensure consistency with Experiments 1 and 2. Following laboratory practices for calculating proportion, 100 repetitions were conducted for each condition. 
Apparatus
The equipment was the same as that used in Experiment 1b. 
Stimulus and conditions
The suppressor and adaptor were Gabor patches (100% contrast, σ = 0.22°), the same as those used in the previous experiments, apart from their spatial frequency. In Experiment 3, the spatial frequency of the suppressor and adaptor was the same and set to one of three values (1.0, 1.75, or 2.50 cpd). The orientation difference between the suppressor and adaptor was either 0° or 90°. 
As in Experiments 1 and 2, circular uniform backgrounds were presented as an aid for binocular fusion, but the following changes were made. The diameter of the background was set at 4.2°. The background was surrounded by a gray ring (outer edge, 5.0° in diameter; inner edge, 4.6°). A cross mark (0.2° × 0.2°) was always presented as a fixation pattern at the center of the background. 
Procedure
In each trial, the observer's key press initiated the stimulus sequence. After a blank period of 0.5 second, the suppressor was presented to the observer's dominant eye, and the two identical adaptors were presented to the other eye for 3 seconds (Figure 5). As was done in Experiments 1a and 1b, the onset of the adaptor was smoothed according to a Gaussian temporal profile from 0% to 100% over 0.5 second. The two adaptors were presented on both sides of the fixation cross. The center of the adaptors was 1° away from the fixation point. The suppressor was dichoptically overlaid in the same retinal location as either of the two adaptors. This presentation resulted in a perceptual experience wherein the dynamic and static stimuli were simultaneously seen side by side. The abrupt offset of the stimuli generated negative afterimages. The observer was asked to indicate which afterimage persisted longer after the afterimages faded away by pressing one of the two keys. The observer indicated by pressing another key when the adaptor to be suppressed was perceived during the presentation of the stimulus. The trials for which the observer reported insufficient suppression were discarded and were not used for the analysis. The average percentage of the discarded trials was 12.6% (variance = 11%), ranging from a minimum of 1.5% to a maximum of 36%. 
Figure 5.
 
Schematic diagram of the stimulus sequence in Experiment 3. A suppressor was presented to the observer's dominant eye (DE, left column), and two identical adaptors were presented to the non-dominant eye (NDE, middle column). Although one of the two adaptors was rendered invisible by the suppressor during the adaptation, two afterimages were perceived after the stimulus offset. A dynamic mask was presented following the observer's response. A typical percept is illustrated in the right column of the figure.
Figure 5.
 
Schematic diagram of the stimulus sequence in Experiment 3. A suppressor was presented to the observer's dominant eye (DE, left column), and two identical adaptors were presented to the non-dominant eye (NDE, middle column). Although one of the two adaptors was rendered invisible by the suppressor during the adaptation, two afterimages were perceived after the stimulus offset. A dynamic mask was presented following the observer's response. A typical percept is illustrated in the right column of the figure.
All combinations of an orientation difference (0° or 90°) and a spatial frequency (1.00, 1.75, or 2.50 cpd) were randomly tested. Each combination was repeated 100 times for each observer, but, due to a programming error during the experiment setup, it was repeated 50 times for one and 130 times for another observer. All measurements for each observer were usually completed in 2 days. 
Results and discussion
Figure 6 shows the percentage of trials in which observers reported that an afterimage induced by a visible adaptor lasted longer than that induced by an invisible one. We conducted a two-way, repeated-measures ANOVA and found significant main effects of orientation difference, F(1, 9) = 26.82, p < 0.001, ηp2 = 0.75, and spatial frequency, F(2, 18) = 6.93, p = 0.006, ηp2 = 0.44, as well as a significant interaction, F(2, 18) = 6.86, p = 0.006, ηp2 = 0.43. Post hoc analysis of the significant interaction revealed a significant simple main effect of orientation difference for all spatial frequencies (all p < 0.001). A simple main effect of spatial frequency was significant in the 90° condition, F(2, 18) = 8.90, p = 0.002, ηp2 = 0.50, but not in the 0° condition, F(2, 18) = 1.59, p = 0.231, ηp2 = 0.15. Multiple comparisons with Bonferroni correction for spatial frequency in the 90° condition showed that the percentage was significantly different between the lowest (1.0 cpd) and the highest (2.50 cpd) spatial frequencies, t(9) = 3.51, p = 0.007. 
Figure 6.
 
Results of Experiment 3. The percentage of trials in which the visible adaptor produced a longer afterimage than the invisible one is shown. Different symbols represent different spatial frequencies of the adaptor, as shown in the legend. The error bars denote 95% confidence intervals. The dashed line represents the 50% level, which indicates that the two afterimages were comparable in duration.
Figure 6.
 
Results of Experiment 3. The percentage of trials in which the visible adaptor produced a longer afterimage than the invisible one is shown. Different symbols represent different spatial frequencies of the adaptor, as shown in the legend. The error bars denote 95% confidence intervals. The dashed line represents the 50% level, which indicates that the two afterimages were comparable in duration.
The binary responses of the observers were subjected to analysis using a generalized linear mixed model (GLMM), where the logit function was employed as the link function. This allowed for the evaluation of the number of trials in which the afterimage of the visible adaptor persisted longer without the need to convert the number of trials into a proportion. Additionally, the Bayes factor of the GLMM was approximated using the method proposed by Wagenmakers (2007)
The Bayesian GLMM identified the main effect model with the interaction as the best model (Table 3). The interaction model was favored over the no-interaction model with BF = 8.17. The other models were much less consistent with the data than the best model (BFs > 100). Multiple comparisons with Bonferroni correction in the 90° condition showed that the percentage was significantly higher for the lowest spatial frequency than those for 1.75 cpd (z = 4.917, p < 0.001) and 2.5 cpd (z = 5.684, p < 0.001). 
The results of Experiment 3 with the highest spatial frequency (2.50 cpd) generally reconfirmed those of Experiment 1. When the orientation difference between the suppressor and adaptor was 0°, the percentage data were well above the 50% level, indicating that the visible adaptor produced longer afterimages than the invisible one. However, when the difference was 90°, the percentage data were close to the 50% level, indicating that the visible and invisible adaptors produced afterimages with similar duration. Moreover, Experiment 3 also showed that the afterimage duration depended on the spatial frequency as well as the visibility of the adaptor. The significant simple main effect of spatial frequency in the 90° condition indicated that the visible adaptor of the lowest spatial frequency tested (1.00 cpd) produced longer afterimages than those of higher frequencies (1.75 and 2.50 cpd), although the 95% confidence interval did not exceed the 50% level. This result is inconsistent with the attention-based account of a relatively weak afterimage in the control condition of Experiment 1
The result for low spatial frequency is consistent with the natural prediction described in Experiment 1, namely, that longer afterimages are observed in the control (no-suppression) condition than in the suppression condition. This prediction did not hold in Experiment 1, probably because an adaptor of a higher spatial frequency was used. This result is further discussed below. 
General discussion
This study investigated whether the duration of an afterimage induced by an invisible stimulus varies with the depth of interocular suppression. An adapting Gabor stimulus (adaptor) was kept invisible throughout adaptation using interocular suppression that was produced by a counterphase-flickering Gabor (suppressor). In Experiment 1, we found that the afterimage duration varied with the orientation difference between the suppressor and adaptor (Figure 2). The afterimage duration was the shortest when the suppressor had the same orientation as the adaptor, inducing the strongest interocular suppression (Baker & Graf, 2009; Stuit et al., 2009; Yang & Blake, 2012). The physical properties of the two stimuli, except for the orientation, were kept constant across different conditions, so that the results could not be attributed to those properties. The results of Experiment 2 were not consistent with the possibility that the suppressor modulates the afterimage duration by reducing the contrast sensitivity in an orientation-selective fashion. Taken together with previous findings suggesting that the afterimage duration reflects the magnitude of visual responses caused by the adaptor (Georgeson & Turner, 1985; Kelly & Martinez-Uriegas, 1993; McLelland et al., 2010), it seems reasonable to conclude that interocular suppression, the depth of which depends on the orientation difference, disrupted the visual adaptation to the adaptor and thus shortened the afterimage duration. This conclusion implies that the afterimage duration can reflect and therefore be a measure of the graded strength of unconscious visual responses. 
If we agree that stronger interocular suppression shortens the afterimage duration, a longer afterimage could be expected in the no-suppression (control) condition. However, this was not found to be the case in Experiment 1 (Figure 2). This result can be understood in terms of the spatial-frequency dependency of afterimage formation. In Experiment 3, when the adaptor had a lower spatial frequency (1.00 cpd), the afterimage duration was longer in the control than in the suppression condition. By contrast, when the adaptor had a higher spatial frequency (2.5 cpd) as in Experiment 1, the afterimage duration was not always longer in the control condition (Figure 6). Thus, even when the suppressor was strong enough to render the adaptor invisible, it remained possible that it did not affect the afterimage duration much. In other words, whether or not visual awareness increases the afterimage duration depends on the spatial frequency of the stimulus. This spatial-frequency dependency of afterimage formation can be accounted for based on previous findings. 
The afterimage duration has been shown to depend on the neural adaptation of two types of cells: phase-sensitive and phase-insensitive (Brascamp et al., 2010; Georgeson & Turner, 1985; Kelly & Martinez-Uriegas, 1993; Suzuki & Grabowecky, 2003). Phase-sensitive cells determine the intensity of the afterimage itself. Adapting phase-sensitive cells increases the genuine contrast of the afterimage, which lengthens the afterimage duration. Phase-insensitive cells, on the other hand, determine contrast sensitivity for the afterimage. Higher contrast sensitivity lengthens the time in which the contrast of the afterimage falls below the threshold. Adapting phase-insensitive cells decreases contrast sensitivity, which shortens the afterimage duration. Brascamp et al. (2010) showed that interocular suppression affects the adaptation of the two types of cells differently, depending on the spatial frequency of the adaptor. Then, the afterimage duration is determined based on the balance between the adaptation of phase-sensitive and phase-insensitive cells. Brascamp et al. (2010) also demonstrated that, in interocular suppression, the adaptor of a low spatial frequency (0.66 cpd) induces a shorter afterimage, whereas that of a high spatial frequency (3.3 cpd) induces a longer afterimage. 
In this study, we argue that the afterimage can be used to investigate visual responses to a stimulus that is not within awareness due to interocular suppression. There have been previous attempts to study unconscious visual processing using afterimages (Bartels & Logothetis, 2010; Brascamp et al., 2010; Dong, Holm, & Bao, 2017; Lak, 2008). However, these studies have focused on the correlation between the afterimage duration and the length of time in which the adaptor was visible while suppressed by binocular rivalry and other phenomena (Tsuchiya & Koch, 2005; van Boxtel et al., 2010). The new aspect of the afterimage method that we emphasize here is that the afterimage duration can be used as a probe to investigate unconscious visual processing, even when the suppressed adaptor is completely out of awareness. This aspect could be an advantage for the psychophysical investigation of unconscious visual processing. 
The afterimage method can be used to extend the scope of the investigation into unconscious visual processing. The quantitative examination of unconscious visual processing has been conducted using neurophysiological and brain imaging techniques (Sterzer, Stein, Ludwig, Rothkirch, & Hesselmann, 2014; Tong, Meng, & Blake, 2006), as these methods do not require a subjective response from observers. By combining the afterimage method with these techniques, we can explore crucial questions, such as what neural responses in which brain areas are closely linked with unconscious visual processing and are reflected in the afterimage duration. In psychophysical or perceptual studies, the prevalent paradigm for investigating unconscious processing involves comparing observers’ responses when the stimulus is visible to those when it is invisible (Breitmeyer, 2015; Pournaghdali & Schwartz, 2020; Sterzer et al., 2014; but see also He & MacLeod, 2001). In this paradigm, it is challenging to psychophysically quantify the depth of suppression when the suppressed stimulus is completely invisible. The afterimage duration could potentially allow us to monitor how deeply the invisible stimulus was suppressed. Moreover, it could serve as a useful tool to elucidate the nature of suppression at various degrees of depth. Quantifying the depth of suppression might contribute to elucidating controversial topics regarding unconscious processing under CFS, such as whether CFS selectively impairs specific processing like that mediated by the ventral pathway and whether unconscious processing under CFS is limited to basic features, with high-level semantic processing being impossible (Almeida, Mahon, Zapater-Raberov, Dziuba, Cabaço, Marques, & Caramazza, 2014; Hesselmann & Knops, 2014; Hesselmann et al., 2018; Rothkirch & Hesselmann, 2018). 
In conclusion, this study extended the utility of the afterimage as an effective psychophysical probe into unconscious visual processing during interocular suppression. Investigations combining the afterimage with other psychophysical, as well as neurophysiological and brain imaging, measures of unconscious processing could provide new insights into unconscious visual processing. 
Acknowledgments
The authors grateful to Guido Hesselmann, PhD, and an anonymous reviewer for their helpful comments and suggestions on an earlier version of the manuscript. 
Supported by Japan Society for the Promotion of Science KAKENHI Grant Numbers 26285162, 18K18686, and 20H01781. 
Commercial relationships: none. 
Corresponding author: Eiji Kimura. 
Address: Department of Psychology, Graduate School of Humanities, Chiba University, Yayoi-cho, Inage-ku, Chiba-shi, Chiba 263-8522, Japan. 
References
Almeida, J., Mahon, B. Z., Zapater-Raberov, V., Dziuba, A., Cabaço, T., Marques, J. F., & Caramazza, A. (2014). Grasping with the eyes: The role of elongation in visual recognition of manipulable objects. Cognitive, Affective, & Behavioral Neuroscience, 14, 319–335, https://doi.org/10.3758/s13415-013-0208-0. [PubMed]
Apthorp, D., Griffiths, S., Alais, D., & Cass, J. (2017). Adaptation-induced blindness is orientation-tuned and monocular. i-Perception, 8(2), 2041669517698149, https://doi.org/10.1177/2041669517698149. [PubMed]
Baayen, R. H., Davidson, D. J., & Bates, D. M. (2008). Mixed-effects modeling with crossed random effects for subjects and items. Journal of Memory and Language, 59, 390–412, https://doi.org/10.1016/j.jml.2007.12.005.
Baker, D. H., & Graf, E. W. (2009). On the relation between dichoptic masking and binocular rivalry. Vision Research, 49, 451–459, https://doi.org/10.1016/j.visres.2008.12.002. [PubMed]
Baker, D. H., & Meese, T. S. (2012). Interocular transfer of spatial adaptation is weak at low spatial frequencies. Vision Research, 63, 81–87, https://doi.org/10.1016/j.visres.2012.05.002. [PubMed]
Baker, D. H., Vilidaite, G., Lygo, F. A., Smith, A. K., Flack, T. R., Gouws, A. D., & Andrews, T. J. (2021). Power contours: Optimising sample size and precision in experimental psychology and human neuroscience. Psychological Methods, 26, 295–314, https://doi.org/10.1037/met0000337. [PubMed]
Bartels, A., & Logothetis, N. K. (2010). Binocular rivalry: A time dependence of eye and stimulus contributions. Journal of Vision, 10(12):3, 1–14, https://doi.org/10.1167/10.12.3.
Bjørklund, R. A., & Magnussen, S. (1981). A study of interocular transfer of spatial adaptation. Perception, 10, 511–518, https://doi.org/10.1068/p100511. [PubMed]
Blake, R., & Logothetis, N. K. (2002). Visual competition. Nature Reviews Neuroscience, 3, 13–21, https://doi.org/10.1038/nrn701. [PubMed]
Brainard, D. H. (1997). The Psychophysics toolbox. Spatial Vision, 10, 433–436, https://doi.org/10.1163/156856897X00357. [PubMed]
Brascamp, J. W., van Boxtel, J. J. A., Knapen, T. H. J., & Blake, R. (2010). A dissociation of attention and awareness in phase-sensitive but not phase-insensitive visual channels. Journal of Cognitive Neuroscience, 22, 2326–2344, https://doi.org/10.1162/jocn.2009.21397. [PubMed]
Breitmeyer, B. G. (2015). Psychophysical “blinding” methods reveal a functional hierarchy of unconscious visual processing. Consciousness and Cognition, 35, 234–250, https://doi.org/10.1016/j.concog.2015.01.012. [PubMed]
Brindley, G. S. (1962). Two new properties of foveal after-images and a photochemical hypothesis to explain them. The Journal of Physiology, 164, 168–179, https://doi.org/10.1113/jphysiol.1962.sp007011. [PubMed]
Cass, J., Johnson, A., Bex, P. J., & Alais, D. (2012). Orientation-specificity of adaptation: Isotropic adaptation is purely monocular. PLoS One, 7(11), e47425, https://doi.org/10.1371/journal.pone.0047425. [PubMed]
Dong, B., Holm, L., & Bao, M. (2017). Cortical mechanisms for afterimage formation: Evidence from interocular grouping. Scientific Reports, 7, 41101, https://doi.org/10.1038/srep41101. [PubMed]
Georgeson, M. A., & Turner, R. S. E. (1985). Afterimages of sinusoidal, square-wave and compound gratings. Vision Research, 25, 1709–1720, https://doi.org/10.1016/0042-6989(85)90143-9. [PubMed]
Gilroy, L. A., & Blake, R. (2005). The interaction between binocular rivalry and negative afterimages. Current Biology, 15, 1740–1744, https://doi.org/10.1016/j.cub.2005.08.045. [PubMed]
Green, P., & MacLeod, C. J. (2016). SIMR: An R package for power analysis of generalized linear mixed models by simulation. Methods in Ecology and Evolution, 7, 493–498, https://doi.org/10.1111/2041-210X.12504.
Han, S., & Alais, D. (2018). Strength of continuous flash suppression is optimal when target and masker modulation rates are matched. Journal of Vision, 18(3):3, 1–14, https://doi.org/10.1167/18.3.3.
Han, S., Blake, R., & Alais, D. (2018). Slow and steady, not fast and furious: Slow temporal modulation strengthens continuous flash suppression. Consciousness and Cognition, 58, 10–19, https://doi.org/10.1016/j.concog.2017.12.007. [PubMed]
He, S., & MacLeod, D. I. A. (2001). Orientation-selective adaptation and tilt after-effect from invisible patterns. Nature, 411, 473–476, https://doi.org/10.1038/35078072. [PubMed]
Hesselmann, G., Darcy, N., Rothkirch, M., & Sterzer, P. (2018). Investigating masked priming along the “vision-for-perception” and “vision-for-action” dimensions of unconscious processing. Journal of Experimental Psychology: General, 147, 1641–1659, https://doi.org/10.1037/xge0000420. [PubMed]
Hesselmann, G., & Knops, A. (2014). No conclusive evidence for numerical priming under interocular suppression. Psychological Science, 25, 2116–2119, https://doi.org/10.1177/0956797614548876. [PubMed]
Iseki, R. (2023). Anovakun (version 4.8.8). Retrieved from http://riseki.php.xdomain.jp/index.php?ANOVA%E5%90%9B.
Kelly, D. H., & Martinez-Uriegas, E. (1993). Measurements of chromatic and achromatic afterimages. Journal of the Optical Society of America A, 10, 29–37, https://doi.org/10.1364/JOSAA.10.000029.
Kleiner, M., Brainard, D. H., & Pelli, D. G. (2007). What's new in Psychtoolbox-3? Perception, 36(ECVP Abstract Supplement), 14.
Lack, L. C. (1978). Selective attention and the control of binocular rivalry. The Hague, The Netherlands: Mouton.
Lak, A. (2008). Attention during adaptation weakens negative afterimages of perceptually colour-spread surfaces. Canadian Journal of Experimental Psychology, 62, 101–109, https://doi.org/10.1037/1196-1961.62.2.101. [PubMed]
Lenth, R. (2023). emmeans: Estimated marginal means, aka least-squares means (version 1.8.5). Retrieved from https://CRAN.R-project.org/package=emmeans.
Ling, S., & Blake, R. (2009). Suppression during binocular rivalry broadens orientation tuning. Psychological Science, 20, 1348–1355, https://doi.org/10.1111/j.1467-9280.2009.02446.x. [PubMed]
Loomis, J. M. (1972). The photopigment bleaching hypothesis of complementary after-images: A psychophysical test. Vision Research, 12, 1587–1594, https://doi.org/10.1016/0042-6989(72)90031-4. [PubMed]
Loomis, J. M. (1978). Complementary afterimages and the unequal adapting effects of steady and flickering light. Journal of the Optical Society of America, 68, 411, https://doi.org/10.1364/JOSA.68.000411. [PubMed]
McLelland, D., Baker, P. M., Ahmed, B., & Bair, W. (2010). Neuronal responses during and after the presentation of static visual stimuli in macaque primary visual cortex. Journal of Neuroscience, 30, 12619–12631, https://doi.org/10.1523/jneurosci.0815-10.2010. [PubMed]
Morey, R., & Rouder, J. (2022). BayesFactor: Computation of Bayes factors for common designs (version 0.9.12-4.4). Retrieved from https://CRAN.R-project.org/package=BayesFactor.
Pelli, D. G. (1997). The VideoToolbox software for visual psychophysics: Transforming numbers into movies. Spatial Vision, 10, 437–442, https://doi.org/10.1163/156856897X00366. [PubMed]
Pournaghdali, A., & Schwartz, B. L. (2020). Continuous flash suppression: Known and unknowns. Psychonomic Bulletin & Review, 27, 1071–1103, https://doi.org/10.3758/s13423-020-01771-2. [PubMed]
Rothkirch, M., & Hesselmann, G. (2018). No evidence for dorsal-stream-based priming under continuous flash suppression. Consciousness and Cognition, 64, 84–94, https://doi.org/10.1016/j.concog.2018.05.011. [PubMed]
Shimojo, S., Kamitani, Y., & Nishida, S. (2001). Afterimage of perceptually filled-in surface. Science, 293, 1677–1680, https://doi.org/10.1126/science.1060161. [PubMed]
Sterzer, P., Stein, T., Ludwig, K., Rothkirch, M., & Hesselmann, G. (2014). Neural processing of visual information under interocular suppression: A critical review. Frontiers in Psychology, 5, 00453, https://doi.org/10.3389/fpsyg.2014.00453.
Stuit, S. M., Cass, J., Paffen, C. L. E., & Alais, D. (2009). Orientation-tuned suppression in binocular rivalry reveals general and specific components of rivalry suppression. Journal of Vision, 9(11):17, 1–15, https://doi.org/10.1167/9.11.17. [PubMed]
Suzuki, S., & Grabowecky, M. (2003). Attention during adaptation weakens negative afterimages. Journal of Experimental Psychology: Human Perception and Performance, 29, 793–807, https://doi.org/10.1037/0096-1523.29.4.793. [PubMed]
Tong, F., Meng, M., & Blake, R. (2006). Neural bases of binocular rivalry. Trends in Cognitive Sciences, 10, 502–511, https://doi.org/10.1016/j.tics.2006.09.003. [PubMed]
Tsuchiya, N., & Koch, C. (2005). Continuous flash suppression reduces negative afterimages. Nature Neuroscience, 8, 1096–1101, https://doi.org/10.1038/nn1500. [PubMed]
van Boxtel, J. J. A. (2017). Different signal enhancement pathways of attention and consciousness underlie perception in humans. The Journal of Neuroscience, 37, 5912–5922, https://doi.org/10.1523/JNEUROSCI.1908-16.2017. [PubMed]
van Boxtel, J. J. A., Tsuchiya, N., & Koch, C. (2010). Opposing effects of attention and consciousness on afterimages. Proceedings of the National Academy of Sciences, USA, 107, 8883–8888, https://doi.org/10.1073/pnas.0913292107.
Wagenmakers, E.-J. (2007). A practical solution to the pervasive problems of p values. Psychonomic Bulletin & Review, 14, 779–804, https://doi.org/10.3758/BF03194105. [PubMed]
Yang, E., & Blake, R. (2012). Deconstructing continuous flash suppression. Journal of Vision, 12(3):8, 1–14, https://doi.org/10.1167/12.3.8.
Figure 1.
 
Schematic diagram of the stimulus sequence in Experiments 1a and 1b. A suppressor was presented to the observer's dominant eye (DE, left column), and an adaptor was presented to the non-dominant eye (NDE, middle column). A negative afterimage of the adaptor was seen on the blank display after the offset of the dichoptic stimuli. A typical percept is illustrated in the right column of the figure. A random noise mask was presented after the observer's response only in Experiment 1b.
Figure 1.
 
Schematic diagram of the stimulus sequence in Experiments 1a and 1b. A suppressor was presented to the observer's dominant eye (DE, left column), and an adaptor was presented to the non-dominant eye (NDE, middle column). A negative afterimage of the adaptor was seen on the blank display after the offset of the dichoptic stimuli. A typical percept is illustrated in the right column of the figure. A random noise mask was presented after the observer's response only in Experiment 1b.
Figure 2.
 
Results of Experiments 1a and 1b. Two experiments were conducted with a basically identical procedure but different sets of equipment. The afterimage duration was plotted as a function of the orientation difference between the suppressor and adaptor (open symbols). The filled symbols represent the results in the control (no-suppression) condition. The black symbols and lines show the mean afterimage duration averaged across Experiments 1a and 1b. The error bars indicate ±1 SEM across different observers. The asterisks denote the conditions in which the afterimage duration was significantly longer than that in the 0° condition.
Figure 2.
 
Results of Experiments 1a and 1b. Two experiments were conducted with a basically identical procedure but different sets of equipment. The afterimage duration was plotted as a function of the orientation difference between the suppressor and adaptor (open symbols). The filled symbols represent the results in the control (no-suppression) condition. The black symbols and lines show the mean afterimage duration averaged across Experiments 1a and 1b. The error bars indicate ±1 SEM across different observers. The asterisks denote the conditions in which the afterimage duration was significantly longer than that in the 0° condition.
Figure 3.
 
Schematic diagram of the stimulus sequence in Experiment 2. Two dynamic stimuli were presented to the dominant eye (DE, left column), and the target was presented to the non-dominant eye (NDE, middle column). A typical percept is illustrated in the right column of the figure.
Figure 3.
 
Schematic diagram of the stimulus sequence in Experiment 2. Two dynamic stimuli were presented to the dominant eye (DE, left column), and the target was presented to the non-dominant eye (NDE, middle column). A typical percept is illustrated in the right column of the figure.
Figure 4.
 
Results of Experiment 2. The mean contrast threshold in the log unit is plotted for the suppression (open squares) and control (filled square) conditions. The thresholds for individual observers are also shown using gray circles. Error bars denote ±1 SEM across different observers.
Figure 4.
 
Results of Experiment 2. The mean contrast threshold in the log unit is plotted for the suppression (open squares) and control (filled square) conditions. The thresholds for individual observers are also shown using gray circles. Error bars denote ±1 SEM across different observers.
Figure 5.
 
Schematic diagram of the stimulus sequence in Experiment 3. A suppressor was presented to the observer's dominant eye (DE, left column), and two identical adaptors were presented to the non-dominant eye (NDE, middle column). Although one of the two adaptors was rendered invisible by the suppressor during the adaptation, two afterimages were perceived after the stimulus offset. A dynamic mask was presented following the observer's response. A typical percept is illustrated in the right column of the figure.
Figure 5.
 
Schematic diagram of the stimulus sequence in Experiment 3. A suppressor was presented to the observer's dominant eye (DE, left column), and two identical adaptors were presented to the non-dominant eye (NDE, middle column). Although one of the two adaptors was rendered invisible by the suppressor during the adaptation, two afterimages were perceived after the stimulus offset. A dynamic mask was presented following the observer's response. A typical percept is illustrated in the right column of the figure.
Figure 6.
 
Results of Experiment 3. The percentage of trials in which the visible adaptor produced a longer afterimage than the invisible one is shown. Different symbols represent different spatial frequencies of the adaptor, as shown in the legend. The error bars denote 95% confidence intervals. The dashed line represents the 50% level, which indicates that the two afterimages were comparable in duration.
Figure 6.
 
Results of Experiment 3. The percentage of trials in which the visible adaptor produced a longer afterimage than the invisible one is shown. Different symbols represent different spatial frequencies of the adaptor, as shown in the legend. The error bars denote 95% confidence intervals. The dashed line represents the 50% level, which indicates that the two afterimages were comparable in duration.
Table 1.
 
Bayes factor analysis of the afterimage duration using linear mixed-effects models. PID, participant ID.
Table 1.
 
Bayes factor analysis of the afterimage duration using linear mixed-effects models. PID, participant ID.
Table 2.
 
Bayes factor analysis of the contrast threshold using linear mixed-effects models. PID, participant ID.
Table 2.
 
Bayes factor analysis of the contrast threshold using linear mixed-effects models. PID, participant ID.
Table 3.
 
Bayes factor analysis of the results of Experiment 3 using linear mixed-effects models. PID, participant ID; SF, spatial frequency.
Table 3.
 
Bayes factor analysis of the results of Experiment 3 using linear mixed-effects models. PID, participant ID; SF, spatial frequency.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×