Optimization-Based Analysis of Music Intervals and Tuning Systems in Oral Traditions Using Pitch Histograms: A Case Study of Iranian Vocal Music

Sepideh Shafiei, Shapour Hakam · 2025

Read Optimization-Based Analysis of Music Intervals and Tuning Systems in Oral Traditions Using Pitch Histograms: A Case Study of Iranian Vocal Music by Sepideh Shafiei, Shapour Hakam (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-11956.txt.

Introduction

Sepideh Shafiei and Shapour Hakam present a 2025 computational method for reading tuning systems out of microtonal singing in oral traditions. Their case is Iranian classical vocal music, performed by one master across a full set of pieces. Such music uses intervals that fall between the notes of a standard keyboard, so staff notation alone misses its pitch world.

Pitch values are drawn straight from the recordings and lined up across pieces with time warping and pitch histograms. Tuning is then fit as a whole so that shared intervals line up across the repertoire. The aim is a data driven picture that holds both the bend of live singing and the steady system beneath it. The authors offer the method as a tool for music research on other oral traditions as well.

Catalog record

CREATORSepideh Shafiei, Shapour Hakam
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE21 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-11956

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Optimization-Based Analysis of Music Intervals and Tuning Systems in Oral Traditions Using Pitch Histograms: A Case Study of Iranian Vocal Music Sepideh Shafiei, Shapour Hakam Abstract This paper presents a computational methodology for analyzing music intervals and tunings in microtonal oral traditions, utilizing pitch histograms, Dynamic Time Warping (DTW), and optimization techniques. By extracting pitch frequencies directly from vocal performances and aligning them with MIDI notes via DTW, we determine musical intervals using histograms. This approach offers an efficient, performance-based, and instrument-independent alternative to traditional tuning system analysis. Optimization techniques are then employed to align intervals throughout the oral tradition repertoire, capturing the specific tunings and modes involved. Our methodology demonstrates the potential of computational techniques in advancing musicological and ethnomusicological research, revealing new insights into the studied traditions. 1 Introduction Understanding the exact musical intervals performed in various traditions is crucial for several reasons. First, it offers deeper insights into the cultural and theoretical foundations of a musical system, highlighting the unique characteristics that distinguish different traditions. Precise interval analysis can uncover the microtonal nuances often overlooked in conventional studies—an essential aspect of oral traditions, where music is transmitted through performance rather than notation. Additionally, analyzing performed musical intervals provides insights into intonation as an aspect of the performer’s style, capturing the expressive choices that shape musical traditions. Accurate interval identification also plays a vital role in preserving and documenting musical heritage, ensuring that the subtleties of performance practices are not lost over time. This preservation is crucial for maintaining the authenticity and continuity of cultural expressions across generations. Furthermore, precise interval analysis has practical applications in music education, performance practice, and digital archiving. In music education, a detailed understanding of intervals enhances the teaching and learning process, helping students grasp the nuances of different musical systems and their theoretical underpinnings. In performance practice, it enables musicians to reproduce traditional music with greater fidelity, respecting the intricate details of intervallic structures. For digital archiving, precise computational documentation ensures that recordings and notations of traditional music remain accurate, comprehensive, and accessible for future research and exploration. By employing computational methods such as pitch histograms, Dynamic Time Warping (DTW), and optimization techniques, researchers can achieve a more precise and systematic analysis of intonation and tuning. These methods allow for the direct extraction of pitch frequencies from multiple vocal performances, enabling the detection of microtonal intervals without reliance on predefined tuning systems. While MIDI note alignment is not central to our approach, we incorporate it where available using DTW to refine interval analysis. To demonstrate this methodology, we apply it to a comprehensive repertoire of 145 pieces performed by a master of Iranian Classical Vocal Music. This dataset is particularly well-suited for analysis due to its extensive coverage of traditional modes, rich microtonal details, and high-quality performances that exemplify the stylistic and intonational practices of the tradition. Optimization techniques further refine this process by systematically modeling intervallic relationships across the repertoire, capturing both intonational fluidity and underlying tuning structures. 2 Choice of Repertoire To demonstrate our methodology and examine its results, we have selected a comprehensive repertoire of 145 pieces from Iranian classical music, performed by one of the twentieth century’s masters, Karimi. These performances have been publicly recorded and transcribed by the prominent ethnomusicologist Masoudieh [ 1 ] . The transcriptions were later digitized in Finale, with pitch bends incorporated into the MIDI playback. This data has been compiled into a database of Iranian classical music as part of the first author’s dissertation [ 2 ] . Iranian vocalists are among the few musicians who continue to preserve and transmit music through oral tradition, relying on master-to-student transmission rather than musical notation. In the foreword to Marufi’s radif, Barkeshli emphasizes that the human voice provides a more authentic reference for Iranian musical intervals than musical instruments. Unlike instruments, which may introduce structural tuning constraints, the human voice remains unbounded, allowing for greater microtonal flexibility [ 3 ] . According to Barkeshli, measuring the Persian scale through vocal performances offers a more genuine representation of traditional intonation. Furthermore, in Iranian classical music, the human voice serves as the primary pitch reference, with instruments—especially unfretted ones like the kamancheh and ud—seeking to closely follow vocal intonation patterns. It is important to note that while this study employs Karimi’s repertoire as a case study, the proposed computational methodology is repertoire-independent and can be easily adapted to analyze other musical traditions with minimal modification. The chosen repertoire serves as a representative example due to its comprehensive coverage of traditional modes, rich microtonal details, and the high-level artistry of its performance. Moreover, the fluidity of intonation in these performances provides an ideal dataset for studying how microtonal intervals are realized dynamically in practice, rather than being confined to fixed theoretical models. To illustrate the methodology, we have selected the first piece of the chosen repertoire, referred to as Example 1, throughout the paper. Figure 1 presents its transcription. In this notation, the p-shaped sign on the note E represents koron , a microtonal modification in Iranian Classical Music that lowers the pitch by approximately a quarter-tone, though its exact tuning varies depending on performance context. Figure 1: Transcription of Example 1 3 Pitch Recognition There are various algorithms for pitch recognition, all of which rely on fundamental frequency estimation to determine pitch. The fundamental frequency (F0), the lowest frequency component of a sound wave, corresponds to its most dominant perceived pitch. Monophonic voice pitch estimation algorithms have been extensively studied in Music Information Retrieval (MIR) literature. Gómez et al. evaluated two leading algorithms, CREPE and pYIN, for monophonic vocal pitch recognition [ 4 ] . Their evaluation on the iKala dataset reported 91% Raw Pitch Accuracy for pYIN and 90.5% for CREPE. Another notable method, SPICE [ 5 ] , employs a self-supervised learning technique for pitch estimation. For this study, we selected pYIN due to its strong performance and extensive validation in MIR research. To minimize errors, we performed manual octave-error corrections when necessary. The output from pYIN is stored in .csv format, allowing for easy substitution with alternative pitch recognition algorithms for experimentation or future improvements. To generate the pitch file, we used Sonic Annotator with the following parameters: a step size of 256, a block size of 2048, low amplitude suppression of 0.1, onset sensitivity of 0.7, prune threshold of 0.1, and a threshold distribution of 2. Pitch values extracted in Hertz (Hz) were converted to the cents system, where each octave equals 1200 cents. The interval between two notes with frequencies f_{1} and f_{2} in cents is given by the following formula. 1200\times\log_{2}(\frac{f_{2}}{f_{1}}) (1) 4 Interval Estimation 4.1 Pitch Histogram Pitch histograms are a valuable tool in Music Information Retrieval (MIR) for visualizing and analyzing musical intervals. They provide a statistical representation of pitch frequencies within a piece, revealing patterns and structures that characterize specific musical traditions. This technique is particularly valuable for handling large datasets and is applicable across various musical genres, including microtonal music in oral traditions. Histograms have been applied extensively to non-Western and microtonal musical traditions, providing insights into unique interval patterns and tuning systems. Serra et al. used histograms to analyze the maqam system in Arab music,

Read the complete text