From the perspective of someone with an interest in perceptual bistability, perhaps the most exciting aspect of this study is the parallels the authors draw between MIB and other bistable phenomena. This comment serves to provide an additional angle on these parallels.

The authors emphasize the correspondence of their observations during MIB with predictions from 'Levelt's proposition 2' that has been formulated for a different case of perceptual bistability (Levelt, 1968). They also emphasize that their observations during Troxler fading are quite different from those during MIB, noting that comparable deviations from Levelt's proposition 2 have been found at certain stimulus values for other bistable phenomena. In this comment I aim to highlight just how remarkable the correspondence is between the present observations and those observed for other bistable phenomena, by summarizing the patterns that have been found for other bistable phenomena and then placing the present observations in the context of those patterns.

I will follow the authors' lead by conceptualizing MIB as a competition between the target and the mask. One can then take the effects of varying target strength and mask strength that are reported in this paper, and evaluate them in the context of existing work on other bistable phenomena, which has characterized the effects of varying the strengths of the perceptual alternatives. That work has revealed two rules that summarize the observations for other bistable phenomena:

1. Changing the strength of one competing percept while keeping the strength of the other percept fixed primarily affects the mean dominance duration of the stronger of the two alternatives (Brascamp et al., 2006; Klink et al., 2008; Moreno-Bote et al., 2010). For instance, increasing the strength of an already strong percept will primarily act to further increase that percept's own dominance durations, whereas increasing the strength of a weak percept will primarily reduce the other percept's dominance durations.
2. A closely related observation is that the maximum rate of perceptual alternations between the competing alternatives occurs around the point where both alternatives are equally strong (Moreno-Bote et al., 2010).

These two rules apply to the full range of 'percept strengths' that can be tested and they somewhat deviate from Levelt's proposition 2, which is accurate for a subsection of this range. The rules are presented graphically in the left part of the linked figure (http://home.kpn.nl/brasca...), using 'A' and 'B' to denote the two percepts, and with the top two panels together reflecting rule 1 and the bottom panel reflecting rule 2.

To evaluate the results of the present study in the context of these two rules one must accept the following:

1. 'Target invisible' epochs in the present study can be conceptualized as periods of mask dominance. This conceptualization helps draw parallels with other bistable phenomena where each percept corresponds to a part of the visual input. In this case the 'visible' percept would correspond to the target and the 'invisible' percept would correspond to the mask.
2. A reduction of target luminance means a reduction in the strength of the 'visible' percept, whereas a change from a normal MIB mask, via a slow mask and a static mask, to no mask at all in the case of Troxler fading, constitutes a gradual reduction in the strength of the 'invisible' percept.

With these assumptions in hand it turns out that the two rules formulated above capture the present findings to a remarkable extent, not just the findings for MIB but also those for Troxler fading. This is illustrated in the right part of the linked figure (http://home.kpn.nl/brasca...).

What is the upshot of this analysis in terms of the conclusions of the present study? Whereas the present study emphasizes the differences between MIB and Troxler fading, this analysis emphasizes that both can be viewed as part of a continuum that is remarkably similar to the continuum observed when varying percept strength for other bistable phenomena. I find this similarity exciting and I hope that this angle provides further support for the authors' view that the present findings bolster MIB's place in the same category as other bistable phenomena. With regard to the authors' view that MIB is categorically different from Troxler fading in that only the former depends on competition between two perceptual interpretations: this view is not necessarily at odds with the idea that MIB and Troxler fading are both on the same continuum, because the absence of a mask may place Troxler fading at an extreme part of this continuum where one percept ('visible') is so much stronger than the alternative ('invisible') that it hardly experiences any competition from this alternative.

Levelt, W. J. M. (1968). On binocular rivalry. The Hague/Paris: Mouton.
Brascamp, J. W., Ee, R. V., Noest, A. J., Jacobs, R., & Berg, A. V. D. (2006). The time course of binocular rivalry reveals a fundamental role of noise. Journal of Vision, 6(11): 8, 1244–1256.
Klink, P. C., van Ee, R., & van Wezel, R. J. (2008). General validity of Levelt’s propositions reveals common computational mechanisms for visual rivalry. PloS One, 3, e3473.
Moreno-Bote, R., Shpiro, A., Rinzel, J., & Rubin, N. (2010). Alternation rate in perceptual bistability is maximal at and symmetric around equi-dominance. Journal of Vision, 10(11): 1, 1–18.