Cross-modal grouping interactions between auditory and visual sequences have previously been demonstrated (O'Leary and Rhodes 1984, Perception & Psychophysics 33 565-569; Rahne et al 2008, Brain Research 1220 118-131). Three experiments were conducted here to determine whether cross-modal interaction precedes unimodal grouping and whether the temporal ventriloquism effect could be found between grouped auditory and visual sequences. We used a repeating four-tone auditory sequence (high-middle-middle-low) where the middle tones could group with either the high tone or the low tone, paired with a sequence of light flashes with a single flash one side and three the other. Experiment 1 showed that the temporal position of the isolated flash in the visual sequence had no effect on which of the auditory tones were perceived as isolated. Experiments 2 and 3 demonstrated that the temporal ventriloquism effect occurs between grouped auditory and visual sequences, as participants reported that the isolated light and tone from grouped visual and auditory sequences seemed to synchronise when they were between 120 and 240 ms apart. These results suggest that unisensory grouping can occur prior to cross-modal interaction.

The ability to make accurate audiovisual synchrony judgments is affected by the "complexity" of the stimuli: We are much better at making judgments when matching single beeps or flashes as opposed to video recordings of speech or music. In the present study, we investigated whether the predictability of sequences affects whether participants report that auditory and visual sequences appear to be temporally coincident. When we reduced their ability to predict both the next pitch in the sequence and the temporal pattern, we found that participants were increasingly likely to report that the audiovisual sequences were synchronous. However, when we manipulated pitch and temporal predictability independently, the same effect did not occur. By altering the temporal density (items per second) of the sequences, we further determined that the predictability effect occurred only in temporally dense sequences: If the sequences were slow, participants' responses did not change as a function of predictability. We propose that reduced predictability affects synchrony judgments by reducing the effective pitch and temporal acuity in perception of the sequences.