Fractal Tempo Fluctuation and Pulse Prediction

Area of Science:

• Cognitive Psychology
• Music Cognition
• Auditory Perception

Background:

• Music performance often features tempo fluctuations.
• Understanding how listeners process these tempo changes is crucial for music cognition.
• Previous research has explored tempo perception, but the role of fractal patterns in adaptation is less understood.

Purpose of the Study:

• To investigate human ability to adapt to fluctuating musical tempi.
• To determine if listeners can predict tempo changes based on underlying fractal structures in music.

Main Methods:

• Analysis of interbeat interval time-series using spectral and rescaled range methods.
• Recording of piano performances with natural expression across diverse musical styles.
• Participant beat-tapping experiments with original and mechanical music performances at different metrical levels.

Main Results:

• Identified long-range (1/f type) serial correlations and fractal scaling in musical performances.
• Participants successfully adapted to significant tempo fluctuations in music.
• Beat prediction was more accurate at the quarter-note metrical level compared to the eighth-note level.

Conclusions:

• Listeners demonstrate a capacity to adapt to tempo variations in music.
• The findings suggest that listeners may utilize long-range correlations and fractal scaling to anticipate tempo shifts.
• This predictive ability is influenced by the metrical level at which the beat is perceived.