Illustration of a trial sequence: a randomly generated matrix (M1) was presented for a randomized duration between 1000 and 1500 ms.

<p>Then, this matrix was replaced by a second matrix (M2), which was identical to M1 except that all pixels in a annulus shaped region reversed luminance polarity. After a variable duration, either an auditory or a visual reference stimulus was presented for 500 ms, while M2 remained on the screen. The sequence was repeated, now with M2 as the first and M1 as the second matrix. Participants could adjust the interval between the matrix transition and the reference stimulus onset using four keys on an external response box in steps of -100, -10, 10, and 100 ms, respectively. The sequence was repeated until the participant indicated with a fifth key that perceptual synchrony between the onset of the reference stimulus and–depending on the task in a given trial–the on- or offset of the fading percept was established.