June 10, 2019 Conference paper Open Access

Polyphonic overtone singing: An acoustic and physiological (MRI) analysis and a first-person description of a unique mode of singing

Proceedings from FONETIK 2019 Stockholm, June 10–12, 2019

Polyphonic Overtone Singing: an acoustic and physio-logical (MRI) analysis and a first -person description of a unique mode of singing

Anna-Maria Hefele1, Robert Eklund2, Anita McAllister3

1 http://www.anna-maria-hefele.com

2 Department of Culture and Communication, Linköping University, Sweden

3 Division of Speech and Language Pathology, Karolinska Institutet, Sweden

ama@anna-maria-hefele.com, robert@roberteklund.info, anita.mcallister@ki.se

Abstract This paper describes a unique singing mode, tentatively labeled “polyphonic overtone singing”. In overtone singing the vocal harmonics of a stabile fundamental frequency are filtered by the singer in such a way that specific upper harmonics are amplified, and heard clear ly, as a second musical voice. In the “throat singing” of Tuva (Mongolia) moving overtones usually occur over a stable drone. In polyphonic overtone singing not only the pitch of the overtones are changed and moving, but also the fundamental which results in two-voice singing.

Introduction

The historical records of overtone sing-ing (OTS) are fairly scarce. It is used in different ethnic traditions and musical styles. A well -known example is the “Tuvan throat-singing” (Levin & Edger-ton, 1999), which mainly occurs in Mongolia and Tibetan temples (Smith, Stevens & Tomlinson, 1967). Singers in these regions employ a variety of throat-singing styles, referred to as sygyt, chöömej and kargyraa (Grawunder, 1999:22). Chöömej is also used as an umbrella term for all central Asian techniques. In the west, the tradition of OT is not as strong, and the development of the technique began during the last century.

The earliest existing recording of western OTS is from 1929, sung by an Amer-ican country singer, Arthur Miles. However, Miles did not leave any marked traces in musical history, although he was a local celebrity and even released an album (Tongeren, 2004:161). Also in America, La Monte Young began to experiment with vocal overtones and started to use it as a musical parameter, although it was not yet mature overtone singing (Tongeren, 2004:166).

In 1968 Karlheinz Stockhausen published Stimmung for six vocalists, which can be considered the first composition in contemporary music with exact, notated vocal overtones, amplified through specific vowels (Saus, 2009:1).

Further, Michael Vetter (Germany), who collaborated with Stockhausen from 1969 (Tongeren, 2004:177), and David Hykes “laid the solid foundations for overtone singing as an independent vocal technique /…/ With their original examples, overtone singing became an art in itself” (Tongeren, 2004:175).

The Greek singer Demetrio Stratos (1945–1979) also belongs to the first generation of experimental western overtone singers, and before his early death he left his album Cantare la Voce(1978) with “vocal sounding overtone noises” (Tongeren, 2004:175).

Trân Quang Hai (1944–), a Vietnamese musician and researcher living in Paris, got in contact with Mongol overtone singing in 1969 and he “collaborated in one of the first compositions and public performances of OTS in Europe after Stockhausen .

Apart from his activities as a performer, Trân’s main contribution to the field of overtone singing are scientific and educational” (Tongeren, 2004:171).

From 1983 and on OTS became popular also outside the field of experi-mental music and the knowledge about the technique grew. Singing groups and overtone choirs were formed for chanting, singing mantras or practicing overtone singing together, so it was not only performed by professional musicians (Tongeren, 2004:185). So called “World Music” gained popularity in the west, which spurred an interested in overtone singing. Also enhanced by the Tuva Ensemble tours (Tongeren, 2004:187). To-day western overtone singing is used in many different musical genres. From a phonetic perspective, the formant tuning and the adjustments of the articulators during OTS, are of interest both with regard to aspects related to production and pedagogy.

Method and material

The study includes acoustic analysis of OTS performed by the first author (AMH in the following). In all spectrograms the program Overtone Analyzer(https://www.sygyt.com/en) was used.

The MRI recordings were made at Freiburg University Hospital, usinga 3T Siemens Prisma Fit scanner. MRI recordings were completed in collaboration with professor Bernhard Richter, Dr Michael Burdumy and the singer at an earlier date. The MRI-illustrations are provided by the singer AMH. The singerAMH is a classically trained soprano with a wide frequency range of around ~100–2000 Hz. She started with overtone singing in 2005 and now cooperates with several musicians and composers to increase knowledge about the musical possibilities of the technique. AMH is also teaching overtone singing in masterclasses around the world.

The techniques

In the western style of overtone singing a distinction can be made between the so-called one- cavity-technique, where harmonics get amplified through specific vowel shapes and the so-called two -cavity-technique, where the tongue is raised and divides the oral cavity into two resonance chambers. This results in very clear, whistle-like harmonics. The differentiation between these two tech-niques was first performed by the ethno-musicologists Tran Quang Hai and Hugo Zemp, based on physiological research (Jentsch 2007:55). The one-cavity technique enhances the harmonics through changing the vowel shapes, mostly over a stabile fundamental. This is usually the first technical stage to learn in OTS.

Results

Below different possibilities with OTS are explored. As seen in the spectrogram in Figure 1, overtones can be clearly visualized over a stabile fundamental. Note that a drone would have the same func-tion. Below the spectrogram the simulta-neous MRI registrations are shown. Also note the different and peculiar tongue shapes used to enhance the partials. This is especially apparent in the initial frames. The technique used for this anal-ysis is the two-cavity technique. Another possibility within the two-cavity technique is to move overtone and fundamental in parallel, or to leave the harmonics stabile and move the fundamental through the subharmonic series as shown in Figure 2.By alternating the fundamental frequency different overtones can be enhanced as shown in Figure 3. Here two alternating fundamentals and their harmonics are combined using the two-cavity technique.

Proceedings from FONETIK 2019 Stockholm, June 10–12, 201992Figure 1. Harmonics 4-10 amplified in the two-cavity technique from fundamental C4. Figure 2. A stable harmonic and a moving fundamental.Figure 3. Overtones over two alternating fundamentals.In Figure 4 AMH illustrates how overtones and fundamentals can be moved in opposite directions. This results in creating a musical counterpoint.Thus, simple classical songs or folk melodies can be easily arranged for OTS. In Figure 5 the spectrogram of the first phrase of “Sehnsucht nach dem Frühlinge” by W.A. Mozart is shown. The fundamental is not moved randomly but rather chosen within the overtone- and subharmonic series and within the singer’s voice range. Within these options harmonics are chosen that deliver a harmony or create a counterpoint move-ment. This provides a harmonic context to the melody in OT. Figure 4. Fundamental and overtones moving in opposite directions, resulting in a musical counterpoint.To result in polyphonic overtone singing, different combinations of movements for overtones and fundamentals are used. In the figure three combinations of movements for the overtones and the fundamentals are displayed.Proceedings from FONETIK 2019 Stockholm, June 10–12, 201993Figure 5. Three combinations of movements for overtones and fundamentals are shown. In this example it is the first phrase of “Sehnsucht nach dem Frühlinge” by W.A. Mozart.The physiology of overtone singing .In the one-cavity-technique the vowel transition between /i/ and /u/ is used, as in the slowly spoken words “oui” and “you”. Here only the movement of the second formant is enhancing harmonics, see Figure 6. In the second panel of the figure the same movement was executed but while whispering, to illustrate the ef-fects of the vocal tract.In the two-cavity technique the ele-vation and shape of the tongue creates two separate resonatory chambers. The tongue shape is similar to a / l/ or / ɹ /. The third formant is lowered and brought close to the second by a raised tongue. The 2nd and 3rd formant are coordinated and moved through the spectrum almost at the same frequency (Saus, 2016:450), amplifying the same harmonic. Also the impact of other formants is lowered.In both techniques, the first formantis kept low to improve control and pro-vide a larger range of options in the har-monic scale. In overtone singing melo-dies are created by formant tuning, pri-marily involving the second and third formant (F2, F3). The shape and eleva-tion of the tongue, jaw position, lip rounding and spreading plays key roles in this process. Figure 6. To the left overtones and f0 is shown while moving from /i/ to /u/ in one-cavity technique. To the right the same movement from /i/ to /u/ while whispering, illustrating filtering of the vocal tract.Discussion In the present study OTS produced by one singer using two techniques was an-alysed. Previously the enhancement of harmonics has been attributed to a nar-rowing of pharynx and the velar constriction, adjusting the mouth opening and a tuning of F1 and F2 (Klingholz 1993). In this study we showed that the formant tuning mainly involves F2 and F3 and that several other alterations are involved, such as tongue elevation and shape, lips and jaw position. Also the alterations differ depending on whether a one- or a two-cavity technique was used. Thus, OTS requires precision and timing of the motor control of the articulators and a very accurate pitch perception.

OT singing: A first-person account

When learning OTS , one learns to perceive harmonics as musical notes or pitches, while they are usually perceived as timbre or different vowels. This also leads to a more detailed perception of timbre and vowels, which is useful while learning other singing techniques or working with other singers. An improved perception and more precise control of your sound can lead toincreased blending and enhanced intona-tion. Learning overtone singing increases the ability to control the first 3 formants precisely and autonomously. This could be a useful ability also in Proceedings from FONETIK 2019 Stockholm, June 10–12, 201994other singing techniques in order to improve specific resonance strategies.

In polyphonic overtone singing not only the harmonics but also the funda-mental is moved within the musical con-text to enlarge the musical possibilities of the technique.When overtones and fundamentals are moved simultaneously a very specific coordination is required. The clos-est analogy for me (AMH) as a singer is that of “juggling”.

On terminology – what to call it

On the note of terminology the term “polyphonic overtone singing” has been questioned, e.g. by Maxfield and Titze (2015:470), who write: “Ms. Hefele dubs her technique “polyphonic over-tone singing.” In reality, like most over-tone singers, she selectively amplifies overtones to create a melody above a rel-atively stable fundamental frequency (F0). While her F0 does move, it may be questioned whether its movement rise to the level of an independent melody. As such, calling her technique “polyphony” – a musical style with its own rich history – might be a stretch. Stuart Hinds (2005) points out that he and other overtone singers are indeed capable of a style of overtone singing that could justifiably be called polyphonic, but this practice seems to be an extension of the style and not the most common practice”. Ethnomusicologist Carole Pegg notes that English terms for the practice include “biphonic singing,” “split-tone singing,” “Jew’s harp voice,” and “throat sing-ing.” French terms are no less varied, including “chant diphonique,” “voix dé-doublée,” and “voix guimbarde”.While it is (trivially) true that the set of upper harmonics are defined by the fundamental frequency, this does not completely rule out the use of the word “polyphony”. Although composers such as Palestrina (c1525–1594) were free, in theory, to put in any note in the upper voices, he most surely would never had considered going outside of the “true harmony”, making the end result similar to AMH’s singing. Palestrina was considered the “golden standard” of polyphony during his lifetime. This links to the second argumentagainst the term, that the singing tech-nique used by Hefele is not true polyph-ony, since “poly-” means “many”, and OTS can only produce two separate voices (Maxfield and Titze 2015:470).This argument has more merit since many languages make use of a “one, two, many” (singular, dual, plural) grammar. However, the suggested terms “biphonic singing” or “chant diphonique”, while technically accurate, might be somewhat misleading since “biphonic(al)” or the more commonly used term diplophonic phonation is well established in pathological voice production (e.g. Colton, Casper, Leonard, 2011) commonly referring to abnormal vocal fold vibrational patterns (Sveç, 2000) sometimes also including the ventricular folds (Maryn, De Bodt, Van Cauwenberge 2003). The term is also used to describe animal vocalizations (Wilden et al. 1998; Volodin & Vo-lodina, 2012), where it is a common ine.g. African Wild Dogs (Lycaon pictus) and dholes (Cuon alpinus). Also, and perhaps more importantly, biphonation implies that the sound source are the vo-cal folds vibrating at different frequen-cies or irregularly. This, of course, is very far from the precise control of the fundamental and the harmonics Hefele, and other OT singers use to create their two -part singing.It is evident that OTS, as performed by AMH, is distinct from Tuva throat-singing but what term best describes it is not entirely clear.Overtone singing as a pedagogical tool

Although interesting in and by itself, OTS can also be used in the teaching of general phonetics and acoustic analysis, as tested for years by the second author (RE). Very often it is difficult to make students understand both that there are Proceedings from FONETIK 2019 Stockholm, June 10–12, 201995

For the comple article, click on the link below:

https://zenodo.org/record/3246011