Perceived continuity and pitch perception

Three experiments investigated the importance of perceived stimulus continuity for the perception of the fundamental frequency $\text{(F0)}$ of an unresolved complex tone. The $\text{F0}$ of the complex was 250 Hz and the harmonics were bandpass filtered between 5500 and 7500 Hz. In the first experiment, $\text{F0}$ discrimination was measured for single-burst tones with durations of 20, 40, and 80 ms, and for stimuli containing two 20- or 40-ms tone bursts separated by an 8- or 16-ms gap. For the single-burst conditions, there was a large decrease in threshold as the duration was increased from 20 to 40 ms. However, performance in the gapped conditions was much worse than that for the single-burst condition with the same cumulative duration (e.g., two 20-ms bursts separated by 8 ms produced higher thresholds than one 40-ms burst). Adding a bandpass noise (with the same spectral envelope as the tone) in the gap between the two tone bursts improved performance to the level of the single-burst condition. When the noise was added, the two discrete tone bursts were perceived as one single tone burst interrupted by the noise, and this seemed to facilitate discrimination. In a second experiment, the effects on pitch of an envelope delay (phase shift) of 0.75 periods between two tone bursts separated by an 8-ms gap were investigated. If the gap was silent, the pitch of the pair was unaffected by the phase shift. However, if the gap contained the bandpass noise, the phase shift between the bursts did produce a significant downward shift in the pitch of the pair. Finally, the third experiment showed that presenting a noise before a single 20-ms burst may improve discrimination performance in some listeners, but not sufficiently to account for the results of the first experiment purely in terms of an improvement in the discriminability of the second tone burst in the pair. The experiments suggest that a level decrease between two tone bursts may disrupt or reset a long integration mechanism, decreasing performance. When there is no level decrease between the bursts, the auditory system may assume that the two bursts belong to the same single tone and analyze them together in order to derive $\text{F0.}$