Testing Beat Perception Without Sensory Cues To The Beat: The Beat-Drop Alignment Test (BDAT) Part 1

Abstract

Beat perception can serve as a window into internal time-keeping mechanisms, auditory–motor interactions, and aspects of cognition. One aspect of beat perception is the covert continuation of an internal pulse. Of the several popular tests of beat perception, none provide a satisfying test of this faculty of covert continuation. 

There is a close relationship between internal timing mechanisms and memory. The internal timing mechanism is an instinct of the human body. It helps people perceive the passage of time in a certain way, and the memory function related to the perception of the passage of time will also be strengthened.

When we have a clearer perception of the passage of time for an event, our memory of the event will be deeper. Our brain not only records the specific content of an event but also evaluates the time of the event at that moment and develops a timeline to help us better remember the event.

Our brains are more efficient at learning and remembering thanks to their internal timing mechanisms. We can better control our time allocation when studying, avoid the problem of time dispersion and inability to concentrate, and make it easier and faster to find relevant memory fragments when we need to recall something.

In addition, the internal timing mechanism can also help us develop better time management skills, allowing us to better adjust the rhythm of life, make full use of time, and improve work efficiency.

In summary, there is a strong connection between internal timing mechanisms and memory, and maximizing this connection requires paying attention to and taking full advantage of this mechanism. For everyone, learning how to use our internal timing mechanisms can sharpen our memories and make us more efficient and successful in life and work. It can be seen that we need to improve memory, and Cistanche deserticola can significantly improve memory, because Cistanche deserticola can also regulate the balance of neurotransmitters, such as increasing the levels of acetylcholine and growth factors. These substances are very important for memory and learning. In addition, Meat can also improve blood flow and promote oxygen delivery, which can ensure that the brain receives sufficient nutrients and energy, thereby improving brain vitality and endurance.

Click know supplements to improve memory

The current study proposes a new beat-perception test focused on covert pulse continuation: The Beat-Drop Alignment Test (BDAT). In this test, participants must identify the beat in musical excerpts and then judge whether a single probe falls on or off the beat. The probe occurs during a short break in the rhythmic components of the music when no rhythmic events are present, forcing participants to judge beat alignment relative to an internal pulse maintained in the absence of local acoustic timing cues. Here, we present two large (N > 100) tests of the BDAT. 

In the first, we explore the effect of test item parameters (e.g., probe displacement) on performance. In the second, we correlate scores on an adaptive version of the BDAT with the computerized adaptive Beat Alignment Test (CA-BAT) scores and indices of musical experience. Musical experience indices outperform the CA-BAT score as a predictor of the BDAT score, suggesting that the BDAT measures a distinct aspect of beat perception that is more experience-dependent and may draw on cognitive resources such as working memory and musical imagery differently than the BAT. The BDAT may prove useful in future behavioral and neural research on beat perception, and all stimuli and codes are freely available for download.

Keywords

Music cognition. Beat perception. Sensorimotor abilities.

Introduction

Musical beat perception involves inferring an underlying periodic pulse from an extract of music and anticipating the timing of each beat as the music unfolds (Patel & Iversen, 2014). Beat perception is a core aspect of music cognition and is a natural capacity in the general population (PhillipsSilver et al., 2011). However, significant variance is present in beat perception abilities (Sowiński & Dalla Bella, 2013). 

In some cases, individuals can demonstrate extreme difficulties performing tasks that require accurate beat tracking and synchronization (Palmer et al., 2014; Phillips-Silver et al., 2011), or show poor rhythm perception without any impairment in motor synchronization (Bégel et al., 2017).

Understanding the ability to perceive a musical beat can inform models of human time-keeping mechanisms and temporal adaptation (Palmer et al., 2014). These mechanisms have been linked to certain behavioral and cognitive traits (e.g., impulsivity; Allman & Meck, 2012) including attention levels and learning (Taatgen et al., 2007). Moreover, deficits in time perception are associated with several neurological and psychiatric conditions (e.g., Parkinson’s disease, schizophrenia, attention deficit hyperactivity disorder, and autism; Allman & Meck, 2012; Breska & Ivry, 2018; Grahn & Brett, 2009; Puyjarinet et al., 2017; Turgeon et al., 2012), and atypical rhythmic timing abilities are associated with some childhood language disorders (e.g., dyslexia, stuttering, developmental language disorder; Ladányi et al., 2020)

For these reasons, the evaluation of beat perception ability is of interest not only in education (Ladányi et al., 2020; Ozernov-Palchik et al., 2018) but also in clinical settings (Cochin De Cock et al., 2018), and beat-perception tests can serve as a useful tool in both contexts. The need for a reliable measure of beat perception is evident from the number of beat perception tests developed over the past decade or so, including the Beat Alignment Test (BAT; Iversen & Patel, 2008), the Harvard Beat Assessment Test (H-BAT; Fujii & Schlaug, 2013), and the Battery for the Assessment of Auditory Sensorimotor Timing Abilities (BAASTA; Dalla Bella et al., 2017), which combines the BAT with several other perceptual and production-based measures of timing abilities. 

Notably, research with the BAASTA found that performance on the BAT was not correlated with tests of interval timing but was correlated with performance on paced tapping to a rhythmic stimulus, suggesting that the BAT is sensitive to beat-based timing mechanisms (Dalla Bella et al., 2017; cf. Fiveash et al., 2022). Recently, two updated versions of the BAT have been developed: the Adaptive Beat Alignment Test (A-BAT; Ross et al., 2018) and the Computerized Adaptive Beat Alignment Test (CA-BAT; Harrison & Müllensiefen, 2018). We note in passing that the rhythm and meter subtests of the Montreal Battery for Evaluation for Amusia (MBEA; Peretz et al., 2003) are not sensitive tests of beat perception abilities, as recently documented by Peretz and colleagues (Tranchant et al., 2021).

The human faculty of rhythm has recently been shown to be multidimensional: individuals’ rhythmic abilities vary along the independent axes of rhythm production, sequence memory-based rhythm perception, and beat-based rhythm perception (Fiveash et al., 2022; Tierney & Kraus, 2015). But even these subskills may be composed of independent proficiencies. The process of identifying and maintaining a beat percept is not well understood, but it likely involves a complex interplay of processing auditory features, measuring durations, and internally producing similar durations, allowing the listener to continue to perceive a beat through complex rhythms in which not all beats are marked by sound events (Cannon & Patel, 2021). 

Existing beat perception tests emphasize some of these aspects of beat perception at the expense of others. The BAT, for example, asks participants to determine whether a metronomic sequence of tones is on or off the beat of a musical excerpt. A participant could succeed by adopting a strategy of comparing the timing of the tones to the local acoustic features associated with the music’s beat; thus, recognizing the acoustic events associated with beats may be more important for passing this test than the ability to internally continue the pulse with accurate timing. 

Perceptual tasks using metronomic clicks or tones (e.g., the isochrony detection tasks included in the BAASTA; Dalla Bella et al., 2017) emphasize the measurement and production of durations but do not test the ability to continue a pulse drawn spontaneously from the complex acoustic features of music. This distinction may be quite important: non-isochrony detection could be performed by simple comparison of consecutive time intervals, and some participants may be able to synchronize with simple metronome clicks but cannot extract a beat from music, as in the case of beat-deafness reported in Phillips-Silver et al. (2011).

This paper reports the development of the new Beat Drop Alignment Test (BDAT), which aims to bring the process of covert music-induced pulse continuation into focus. The BDAT employs naturalistic musical stimuli and asks participants to judge whether a single probe event falls on the musical beat or not. Crucially, the probe occurs within a bar of music from which all rhythmic cues have been removed after the first beat, eliminating the ability to do the task based on judging the alignment of the probe with locally prominent acoustic events. Hence, to perform well on the BDAT participants are required to (1) extract a mental representation of the musical beat, (2) continue this mental representation with accurate timing over a short period when there are no sensory cues to the beat, and (3) to compare their mental representation of the beat to a single probe sound.

In line with findings from the beat processing literature, we hypothesize that the difficulty of items on this test will mainly depend on three factors: the degree of probe displacement from the target beat (Harrison & Müllensiefen, 2018), the probe displacement direction about the target beat (early/late; Van Der Steen et al., 2014), and the metrical strength of the target beat (Patel et al., 2005). In addition, we hypothesize that individuals with extensive training in music or dance will show superior performance on the BDAT task compared with the general public, and that performance on the BDAT will be associated with the level of musical experience (i.e., general musical sophistication; GMS) as measured by the Gold-MSI (Müllensiefen et al., 2014).

Additional distinctive features of the BDAT

The BDAT has other appealing features from the standpoint of testing beat perception. First, it uses naturalistic music tracks. By using varied, complex instrumentation, the BDAT aims to maintain the listener’s interest and avoid the fatigue that can come from psychophysical timing tasks which use simple and timbrally uniform sounds on each trial. Second, it is quickly administered: The BDAT does not exceed 15 minutes and can be as short as 7 minutes. Because it was constructed as an adaptive test, testing length (number of items) and duration (in minutes) are flexible to ensure the highest testing efficiency in the shortest time. Third, the BDAT has unambiguously defined beat times. 

Since the BDAT stimuli were created specifically for the test (using a synthesizer), beat locations were based on the MIDI grid used to align the electronic instruments, and probe sounds were placed relative to these beat locations. Computer-generated stimuli have been previously used in BAASTA (Dalla Bella et al., 2017), but not in the BAT, A-BAT, or CA-BAT. Fourth, unlike the CA-BAT, the BDAT uses a one-alternative forced-choice (1-AFC) task paradigm: participants listen to each stimulus once rather than making a judgment comparing two stimuli as in the CA-BAT. This makes trials comparatively short and reduces the working memory load on participants. Finally, the BDAT allows for investigation of the impact of implied meter on beat perception in the absence of local acoustic cues of beat strength.

Study design

The aim of Study 1 (calibration) was to construct the main BDAT paradigm and obtain data for estimating an explanatory item response theory (IRT) model. The explanatory IRT contributes to the test’s validity by assessing its assumptions and hypotheses. In addition, the explanatory IRT model provides the basis for estimating the difficulties of the test items, which are later used for estimating participant abilities in the adaptive test. 

Previous research (e.g., Nguyen, 2017) suggests that beat processing abilities differ in individuals with extensive musical training versus the general public. Therefore, the secondary aim of Study 1 was to investigate whether these two groups differ in their performance on BDAT.

Study 2 aimed to explore relationships among three measures: beat perception as measured by BDAT performance, beat perception as measured by CA-BAT performance, and self-reported musical sophistication based on the Gold-MSI questionnaire. This information is important to understanding the nature of the ability that is quantified by the BDAT.

Study 1

Methods

Participants

A total of 1361 participants were recruited for an online experiment. Eleven participants with incomplete responses were disregarded at the data analysis stage, leaving 125 participants: 72 identified as female and 53 as male, ages ranging from 18 to 66 years (M = 30.06, SD = 8.79). Participants were recruited through social media and email invitations. The sample consisted of 40 (32%) self-declared musicians, 19 (15.2%) dancers, 64 (51.2%) individuals who did not identify as musicians or as dancers, and two (1.6%) individuals who preferred not to disclose this information. 

Participants were considered musicians or dancers if they were currently studying for a music or dance degree at a higher education institution if they had graduated from such an institution, or if they had more than 10 years of experience actively engaging in music or dance activities in a professional or semiprofessional setting. All participants provided informed consent to participate in the study.

Materials

BDAT Thirty novel musical tracks were composed and synthesized by the first author in the style of contemporary electronic dance music (EDM) using Ableton Live 10. Clips were created to be naturalistic musical stimuli without any vocals, and therefore included diverse instrumentation and sound effects, and employed a variety of tonalities. The sound was the same in both the left and right channels. All musical events were aligned to a MIDI grid to ensure timing precision. All clips consisted of 6 bars of music and were composed using the time signature of 4/4 in the tempo of 125 bpm (i.e., 480 ms between beats), which resulted in clips of approximately 11.5- s duration. Common time (4/4) was chosen as a norm of contemporary dance music, and the tempo (which is in the range of EDM heard in clubs) was decided based on the results of a small pilot study that identified 125 bpm as the most natural sounding tempo for the chosen musical style.2

All clips were structured in the following way: musical material was introduced in Bars 1–3, Bar 4 contained the beat-drop (from Beats 2–4) and a probe sound during this beat-drop that was either aligned or misaligned with the underlying beat and in Bars 5–6 the musical material returned. (The probe sound was a 16th-note duration woodblock pitched at F2, duration = 120 ms; Fig. 1). A beat-drop is a sudden absence of most sounds for a short period, most commonly used as a compositional device to create musical tension in electronic dance music. During the beat drop, there were no rhythmic cues, though rhythmically ambiguous sound (e.g., a drone) was present to fill the four-beat gap and to help build expectation for the return of the rhythm.

The probe in the fourth bar was placed to either coincide with the beat of the music (ON condition), or it was shifted away from the beat (OFF condition). In the ON condition, the probe was placed on the 3rd (strong) beat or the 4th (weak) beat (Palmer & Krumhansl, 1990). In the OFF condition, variations were created by manipulating the following parameters: metrical strength (strong/weak; probe was manipulated away from the 3rd or 4th beats), probe displacement direction (early/ late; the probe would be placed either before or after the beat), probe displacement (degree of displacement from the actual beat location; see Fig. 1 and supplementary sound examples).

Probe displacement settings

Probe displacement was measured in units of beat period and ranged from 15% to 45% of the beat period in 7 steps. As noted by Harrison and Müllensiefen (2018), the relationship between probe displacement (in % of units of a beat) and perceptual difficulty is not linear (see Harrison & Müllensiefen, 2018, for details), therefore probe displacement points were established by finding equal distances on the perceptual accuracy scale that accounts for the relationship between physical probe displacement and perceived difficulty, and is expressed by the following formula:

Expressed as a percentage, these calculations resulted in the following displacement points: 15%, 18%, 20%, 23%, 26%, 31%, 45%. These levels were chosen as they were above the displacement hearing threshold (Harrison & Müllensiefen,2018) and below a displacement of 50% (since an event occurring exactly halfway between beats, a frequent location for rhythmic events, might be mistaken for “on the beat”). Also, a 50% displacement following the 3rd beat coincides with a 50% displacement preceding the 4th beat, and such conceptual overlap would interfere with later analyses.

Variations of the experimental stimuli were created using automated item generation coded in Python. 2 ON and 28 OFF variations were produced for each of the 30 tracks, which yielded a total of 900 items (audio clips).

As noted earlier, a 1-AFC paradigm was employed for the test: participants listened to 30 clips of music and answered whether the probe was on or off the beat. Every correct response scored 1 point, and every incorrect response scored 0. Participants heard each of the 30 tracks once. Each track was presented in either ON or OFF condition with a 50% chance. The perturbation for the OFF tracks was chosen randomly from the 28 options: seven options with the probe coming before Beat 3 (early/strong beat), seven options with the probe coming after Beat 3 (late/strong beat), seven options with the probe coming before Beat 4 (early/weak beat), and seven options with the probe coming after Beat 4 (late/weak beat).

The BDAT can be downloaded and installed (https:// github.com/klausfrieler/BDT/). All BDAT materials can be accessed online (https://osf.io/jpc29/).

Gold-MSI To evaluate the level of musical sophistication of the participants, the Goldsmiths Musical Sophistication Index was used (Gold-MSI; Müllensiefen et al., 2014). The GoldMSI assesses a broad range of self-reported musical skills and behaviors on five dimensions: active engagement, perceptual abilities, musical training, singing abilities, and emotions. Study 1 only made use of two subscales: Musical Training (Gold-MSI_MT), composed of seven items assessing the extent of musical training and practice, and Perceptual Abilities (Gold-MSI_PA) composed of nine items that represent the self-assessment of cognitive and perceptual musical abilities. These scales can be downloaded.

Procedure

The experiment was conducted online using the Qualtrics survey platform. Participants provided their informed consent and familiarized themselves with the task by listening to a sample track presented first in the ON condition and then in the OFF condition. They were asked to use headphones and were asked not to tap or otherwise move to the beat of the music, as movement to the music has been demonstrated to affect beat perception (Manning & Schutz, 2013; Morillon et al., 2014). The BDAT was followed by the Gold-MSI questionnaire. Participants were also asked to rate test difficulty on the 7-point Likert scale, with 1 being easy and 7 being extremely difficult.

For more information:1950477648nn@gmail.com