Drum loop rhythmic analysis

Sound by its very nature is temporal, and in its most generic sense, the word rhythm is used to refer to all of the temporal aspects of a musical work, whether represented in a score, measured form a performance, or existing only in the perception of the listener [35]. The drum loop music is taken as the stimuli in this study because of its obvious and simple rhythm characteristic. Drum loops are prerecorded percussive riffs that are designed to create a continuous beat or pattern when played repeatedly. Loops are usually compiled in commercially available databases containing several hundreds, or even thousands, of these riffs. These collections are widely used in computer music composition and production as a means to generate high-quality music tracks in a quick and easy manner. This study utilizes the techniques in the field of audio signal processing and music analysis to extract some features from the drum loops and discusses their effects on the modulation of autonomic nervous system.

Entrainment describes a process whereby two rhythmic processes interact with each other in such a way that they adjust towards and eventually ‘lock in’ to a common phase and/or periodicity [7]. For example, we tap foot and shake body to the beat of a song.

Similarly, there are many naturally occurring rhythms within the human body such as the heartbeat, blood circulation, respiration and many others. Therefore, the relationship between the musical rhythmic characteristics and heart rhythm is what this study wants to explore.

So the next problem is how to quantitatively define the musical characteristic, rhythm.

Intuitively, the first feature of rhythm is its speed. The speed of heart rhythm is called heart rate and the speed of musical rhythm is called tempo. Tempo is usually indicated in beats per

Fig. 3.17(a): The sound wave of one single note

Fig. 3.17(b): Onset, attack and transient

minute (BPM) in modern music. The beat means the exact time we nod our head or tap our feet to the rhythm. It is one temporal aspects of a musical work existing in the perception of the listener and a fundamental unit of the temporal structure of music. Once the beats in one piece of music were detected, the tempo could be decided as the unit, beats per minute.

Is the tempo enough to describe the musical rhythm fully? Through the automatic beat tracking algorithm, each onset in a piece of music which probably makes us tap to follow will be identified. And it can be found that the intervals between each identified beat are almost the same. If the other components of a piece of music were removed except the beats, the remains are only the repeated and equal spaced sound pulses. These pulses can’t make us feel rhythmic. So it is not enough to represent the musical rhythm by the only one feature, tempo.

Observing the characteristics of heart rhythm, it can be found the variability of heart rate exists in a stable and near constant heart rate. Inspired by the similarity, the second feature, complexity, is proposed to be the second quantitative measure to describe the musical rhythm.

As the tempo is to musical rhythm, so is the average heart rate to the heart rhythm. As the complexity is to musical rhythm, so is the heart rate variability to the heart rhythm. That’s why the feature, complexity, is chosen. The heart rhythm is just like a piece of music. If it is just a monotone pulse, it will be not good to listen, in other words, not a healthy heart rhythm.

3.2.1 Tempo

To find the exact time when we nod our heads or tap our feet is called “beat tracking.”

Automatic beat tracking is an essential task for many applications such as musical analysis, automatic rhythm alignment of multiple musical instruments, cut and paste operations in audio editing, beat driven special effects.

Music is expressed by the successive notes. These notes record the relating temporal information. Identifying and characterizing these notes is an important aspect of the following steps of music analysis. Here some nouns must be explained first. In the Fig. 3.14 (a), the sound wave of one single note is shown. The definitions of onset, attack and transient are shown in Fig. 3.14(b) and were explained in [36]. An onset can be defined as the instant when the attack transient begins, thus marking the beginning of the note. So the first step of music analysis is to detect the onset. In the general case of a polyphonic signal, where multiple sound objects may be present at a given time, the onset detection is not easy. The procedure employed in the majority of onset detection algorithms is illustrated in Fig. 3.15: from the original audio signal, which can be pre-processed to improve the performance of subsequent stages, a detection function is derived at a lower sampling rate, to which a peak-picking algorithm is applied to locate the onsets.

Once the rhythmic events (the onsets) have been determined, the beat tracking algorithm will be applied. The beat tracking algorithm adopted in this work is developed by Simon Dixon [37]. First, the time intervals between pairs of events are determined. These data are

Fig. 3.18: Flowchart of a standard onset detection algorithm

clustered to generate a ranked list of tempo hypotheses. The top ranked clusters represent a set of hypotheses as to the basic tempo of the music. The processing mentioned above is called tempo induction. The tempo induction algorithm computes the approximate inter-beat interval, but not calculates the beat times. In order to calculate beat times, a multiple hypothesis search is employed, with an evaluation function selecting the hypothesis that fits the data best. In this work, the interactive beat tracking and visualization system developed by Simon Dixon is used to determine the drum loop tempo.

3.2.2 Complexity

As mentioned above, the sound with a fixed tempo doesn’t make people feel rhythmic at all. Therefore, the musical complexity is proposed to be another important feature to describe the musical rhythm. The notion of complexity has generally been studied in the context of information theory and is closely connected with concepts such as randomness, information, regularity, and coding. Some measures of complexity that corresponds to a high degree with a human’s subjective notion of complexity have been discussed [38-39]. Because these measures are made to fit the human perception on temporal pattern complexity, the questionnaires are used to collect the subjects’ opinions about musical complexity directly in this study. When the relationship between the musical rhythmic characteristics (tempo and complexity in this study) and human heart rhythm is understood, the beat tracking algorithm and the complexity measure can be served as the automatic musical rhythmic characteristics extractor and the corresponding effect on human heart rhythm can also be conducted automatically.