Overtone Singing Physics: Formant Tuning in Human Voice
Executive Summary & Theoretical Thesis: Non-Linear Supraglottal Acoustics in Biphonic Phonation
Biphonic vocalization, known across ethnomusicological and acoustic literature through the Central Asian traditions of Tuvan sygyt and kargyraa, Mongolian khöömei, and related Siberian lineages, constitutes an extreme physical realization of acoustic filtering. The central thesis of modern non-linear vocal mechanics asserts that biphonic vocalization operates not through independent physiological sound generators oscillating simultaneously, but via an extreme configuration of supraglottal transfer dynamics. The practitioner manipulates the internal geometry of the craniomaxillofacial airway to partition the vocal tract into a dual-cavity coupled resonator system. This system demonstrates how intentional manipulation of vocal tract standing waves concentrates acoustic energy into isolated spectral regions, transforming a multi-harmonic glottal pulse train into a discrete acoustic line spectrum characterized by a drone and a whistle-like melody.
Through targeted dynamic positioning of the tongue dorsum, tongue tip, and epiglottic structures, the singer aligns the lower supraglottal resonance peaks—principally the first and second formants ($F_1$ and $F_2$)—such that they converge directly onto a single selected integer harmonic ($n \in [6, 16]$) of the fundamental frequency ($f_0$). This precise convergence produces an amplification factor exceeding +20 dB relative to adjacent partials, effectively stripping the ambient acoustic output of broadband speech characteristics and outputting a quasi-sinusoidal acoustic beacon. Investigating these vocal tract acoustic resonance cavities clarifies the boundary conditions under which biological waveguides achieve narrow-band filtering, high quality factors ($Q > 30$), and direct impedance transformation.
The Glottal Source Function as an Impulse Comb Filter
The fundamental excitation mechanism underlying overtone singing tuvan throat singing formant tuning physics remains the vibrating true vocal folds (plicae vocales). In standard modal phonation, the transglottal airflow creates a volume velocity waveform whose derivative exhibits an asymptotic spectral roll-off of approximately -12 dB per octave, as modeled in Fant’s classical acoustic framework. In biphonic singing styles such as sygyt, this glottal source function is radically altered via hyper-adduction of the vocal processes, significantly elevating the subglottic pressure ($P_s$) and forcing an abbreviated glottal open quotient ($Q_o \le 0.30$).
This abrupt, sharp glottal closure delivers an acoustic excitation profile approaching a periodic Dirac delta comb in the time domain:
$$u_g(t) = \sum_{k=-\infty}^{\infty} U_0 , \delta(t - k T_0)$$
Because the derivative of this pulse train yields exceptionally high-frequency spectral components with a shallow roll-off rate (often approaching -6 dB per octave or flatter across the mid-frequency band), the source delivers high energy to upper harmonic partials up to 4 kHz. The glottal pulse spectrum acts as an unfiltered acoustic comb:
$$U_g(\omega) = \omega_0 \sum_{n=-\infty}^{\infty} \delta(\omega - n \omega_0)$$
This distribution supplies raw spectral materials across the mid-frequency band. The practitioner does not produce the high-pitch melodic line at the level of the larynx; rather, the glottis provides a rich harmonic continuum that supraglottal acoustic structures selectively amplify or attenuate.
Formant Clustering and the Collapse of Broadband Radiation
Standard human vowel production distributes acoustic energy across multiple formants ($F_1$ through $F_4$), creating phonemic vowel identity via relative formant dispersion. In biphonic vocalization mechanics, this broadband distribution collapses. The physical objective of formant tuning in sygyt is the precise mathematical superposition of two formant poles:
$$s_1, s_2 \to \sigma \pm j \omega_n$$
where $\omega_n = 2\pi n f_0$.
By driving the acoustic poles of the vocal tract transfer function into identical frequency coordinates, the system replaces vowel perception with the perception of a single, quasi-sinusoidal tone riding atop a low-frequency drone. This phenomenon occurs through the deliberate manipulation of cross-sectional area variations along the vocal tract axis $x$, effectively driving the reflection coefficients at the junction of the oral and pharyngeal cavities to enforce constructive interference. Under these geometric conditions, transmission zeros (anti-resonances) suppress energy in the spectral bands adjacent to the selected harmonic. The oral radiation impedance at the lips, working in tandem with an expanded pharyngeal volume, converts the tract from an open broadband radiator into a high-Q bandpass acoustic filter.
+------------------------------------------------------------------------------------+
| VOCAL TRACT TRANSFER TOPOLOGY |
| |
| Glottal Source Pharyngeal Cavity Anterior Cavity |
| [ Broadband Comb ] =======> [ F1 Tuning Pole ] ==========> [ F2 Tuning Pole ] |
| \ / |
| \ / |
| v v |
| [ Compound Formant Clustering ] |
| | |
| v |
| [ Monochromatic Output Line ] |
+------------------------------------------------------------------------------------+Paradigm Shift: Active Filtering Versus Biological Multi-Voicing
Early European travelers and 19th-century phoneticians misclassified Tuvan and Mongolian overtone phonation as polyphonic vocal cord vibration, hypothesizing the presence of independent anatomical vibrators within the ventricular folds or piriform sinuses. Laboratory imaging, beginning with mid-20th-century tomographic scans and culminating in modern dynamic magnetic resonance imaging (MRI), overturned this multi-voicing hypothesis. Biphonic vocalization adheres to linear and weakly coupled non-linear source-filter acoustics.
The linear vocal tract transfer function linking volume velocity at the lips $U_{lip}(s)$ to volume velocity at the glottis $U_{glottis}(s)$ is defined in the complex frequency domain as:
$$T(s) = \frac{U_{lip}(s)}{U_{glottis}(s)} = \prod_{k=1}^{\infty} \frac{s_k s_k^}{(s - s_k)(s - s_k^)}$$
where $s_k = -\alpha_k + j\omega_k$ represents the complex conjugate pole pairs corresponding to tract resonances. In biphonic vocalization, lingual constriction divides the vocal tract into a dual-chamber system. When the exit constriction of the posterior cavity and the entrance impedance of the anterior cavity establish boundary conditions where $\omega_1 \approx \omega_2 \approx n\omega_0$, the transfer magnitude $|T(j\omega)|$ collapses into an isolated, high-amplitude resonant peak. Transfer zeros appear via acoustic shunts in the piriform sinuses and vallecular recesses:
$$Z(s) = \prod_{m=1}^{M} (s - z_m)(s - z_m^*)$$
These zeros attenuate spectral regions directly above and below the clustered pole pair, suppressing neighboring partials by more than 30 dB relative to the reinforced target harmonic.
The singer does not possess anomalous laryngeal anatomy. Rather, through empirical neuromuscular conditioning, the practitioner alters the supraglottal boundary configurations to isolate individual high overtones from the glottal pulse spectrum.
Historical Lineage & Experimental Precedents: From Central Asian Pastoralism to Modern Cineradiography
Nomadic Mimetic Traditions: The Khöömei, Sygyt, and Kargyraa Classifications
The genesis of biphonic vocalization is historically situated within the pastoral nomadism of the Altai-Sayan mountain range, spanning modern Tuva, western Mongolia, Khakassia, and the Altai Republic. Ethnomusicological analysis reveals an intimate connection between vocal development and animist acoustic mimesis. Nomadic herders engaged in active acoustic interrogation of natural environments, utilizing geographic features—such as mountain faces, frozen rivers, and steppes—as external resonators. The terminology itself reflects these environmental roots: sygyt (to whistle), kargyraa (to gurgle or wheeze), and khöömei (referring to the throat, pharynx, or guttural resonance).
Traditional Pastoral Practice Acoustic Physical Phenomenon
---------------------------------------------------------------------------------
Mountain Face Echo Mimicry ==========> Longitudinal Wave Phase Inversion
Wind/River Aeolian Resonances ===========> High-Q Cavity Filtering
Epic Guttural Narrative Phonation ======> Subharmonic Period-Doubling Bifurcation
These styles represent deliberate manipulations of acoustic impedance. Nomadic singers discovered that by configuring the craniomaxillofacial profile to imitate the wind shearing against cliffs or water echoing through caverns, they could project high-amplitude acoustic signals across long distances without exhausting metabolic energy. This oral acoustic lineage treated human anatomy as an adjustable acoustic waveguide long before formal wave mechanics mapped these internal configurations.
Early 20th-Century Spectrographic Inquiries and Helmholtz Resonators
The initial scientific classification of biphonic singing emerged during the mid-20th century. Early European observers documented the phenomenon using physical Helmholtz resonators—spherical brass enclosures tuned to discrete frequencies—placed adjacent to the singer’s mouth to detect isolated partials. These early mechanical experiments, however, struggled to determine whether the high-frequency whistle was generated aerodynamically via a vortex shedding mechanism (similar to an edge tone or whistling) or via acoustic resonance filtering.
The transition from subjective auditory observation to objective physical analysis began with the work of A. N. Aksenov (1964), who captured high-fidelity field recordings of Tuvan masters using portable Nagra reel-to-reel magnetic tape systems, later analyzed via early Kay Electric sound spectrographs. Aksenov’s spectrographic plates confirmed that the high-frequency melodic line did not drift independently of the drone; it tracked integer multiples of the fundamental frequency ($n f_0$). This confirmed the presence of harmonic filtering and disproved autonomous turbulence whistle theories. Theoretical advances by Gunnar Fant (1960) in Acoustic Theory of Speech Production and L. A. Chistovich’s speech analysis frameworks subsequently provided the mathematical basis to model vocal tract area functions using stepped-impedance acoustic transmission lines.
These spectrographic evaluations demonstrated that the high whistle retained the periodic structure of the glottal pulse train, showing that its frequency was strictly coupled to the lower fundamental drone.
MRI and High-Speed Video Fluoroscopy of Subphonetic Lingual Geometries
The deployment of medical diagnostic technologies during the late 20th and early 21st centuries transformed the understanding of internal articulation in overtone singing. Groundbreaking research by Levin and Edgerton (1999), paired with dynamic fluoroscopic imaging and vocal tract area function extraction by Adachi and Yamada (1999), replaced indirect acoustical modeling with direct observational data. High-speed transnasal video-endoscopy illuminated the larynx, demonstrating distinct mechanical operations between styles.
In kargyraa, the false vocal cords (ventricular folds) adduct and vibrate in an exact 2:1 subharmonic frequency-locked cycle relative to the true vocal folds, generating a fundamental frequency an octave below the singer’s modal baseline ($f_0 / 2$). Conversely, in sygyt, the ventricular folds remain open or act as a static hyper-adducted collar, while the tongue undergoes radical shape changes.
Dynamic magnetic resonance imaging conducted by Story, Titze, and Hoffman (1996), and later applied specifically to Tuvan phonation, proved that sygyt requires a dual lingual constriction: the tongue dorsum rises toward the hard palate while the tongue tip curls backward into a retroflex posture or rests against the lower incisors, forming an anterior oral cavity with a volume below $5\text{ cm}^3$. This physical division establishes the geometry necessary for formant clustering.
Mathematical Formalism & Physical Mechanics: The Dual-Cavity Waveguide and Acoustic Impedance Matching
Webster Horn Equation and Area Function Discretization
Planar wave propagation through an arbitrary, non-uniform human vocal tract geometry is governed by Webster’s Horn Equation, which assumes low-frequency acoustic wave propagation along the longitudinal axis $x$:
$$\frac{1}{A(x)} \frac{\partial}{\partial x}\left(A(x)\frac{\partial p(x,t)}{\partial x}\right) - \frac{1}{c^2}\frac{\partial^2 p(x,t)}{\partial t^2} = 0$$
where $p(x,t)$ is the acoustic sound pressure, $c$ is the speed of sound in humid air at 37°C ($c \approx 353\text{ m/s}$), and $A(x)$ is the cross-sectional area function perpendicular to the sound propagation axis.
Glottis Lips
| |
v v
+-+----+-------------+------------------------------------+-----------+-+
| | | | | |
| A_1 | A_2 | ... | A_N | |
| | | | | |
+-+----+-------------+------------------------------------+-----------+-+
|<--- l_1 -------->|<-------------- l_2 --------------->|<- l_N --->|To model the acoustic response computationally, the continuous area function $A(x)$ is discretized into a chain of $N$ cylindrical acoustic tubes, each of length $l_k$ and cross-sectional area $A_k$. Within the $k$-th acoustic transmission line element, the pressure $p_k$ and volume velocity $U_k$ are related to the $(k+1)$-th element via the two-port transfer matrix:
$$\begin{bmatrix} p_k \ U_k \end{bmatrix} = \begin{bmatrix} \cosh(\gamma_k l_k) & Z_{0,k} \sinh(\gamma_k l_k) \ Z_{0,k}^{-1} \sinh(\gamma_k l_k) & \cosh(\gamma_k l_k) \end{bmatrix} \begin{bmatrix} p_{k+1} \ U_{k+1} \end{bmatrix}$$
where $Z_{0,k} = \frac{\rho c}{A_k}$ is the characteristic acoustic impedance of the cylinder, $\rho$ is the density of air ($\rho \approx 1.14\text{ kg/m}^3$), and $\gamma_k = \alpha_k + j\frac{\omega}{c}$ represents the complex propagation constant, which incorporates thermal dissipation and viscous boundary layer losses:
$$\alpha_k = \frac{1}{R_k c}\sqrt{\frac{\omega \eta}{2\rho}} + \frac{\gamma - 1}{c}\sqrt{\frac{\omega \kappa}{2\rho C_p}}$$
In these boundary loss terms, $\eta$ represents the dynamic viscosity of air, $\kappa$ the thermal conductivity, and $C_p$ the specific heat capacity at constant pressure.
In sygyt, the singer modifies $A(x)$ until a deep minimum ($A_c \to 0.05\text{ cm}^2$) emerges along the dental-alveolar ridge, creating a sharp impedance mismatch that acoustically isolates the anterior oral cavity from the posterior pharyngeal cavity.
Coupled Resonator Pole Convergence: Merging $F_1$ and $F_2$
Under typical phonetic conditions, the first formant ($F_1$) is primarily governed by the pharyngeal volume and the lingual constriction aperture, while the second formant ($F_2$) depends on the length and volume of the anterior oral cavity. The vocal tract can be modeled as two coupled resonators:
Posterior Cavity (V_1) Anterior Cavity (V_2)
+-----------------------+ +-----------------------+
| | Constrict| | Lip Orifice
| Pharyngeal Resonator |====(A_c)==| Oral Resonator |====(A_lip)==> P_rad
| | | |
+-----------------------+ +-----------------------+Approximating both chambers as coupled Helmholtz resonators yields:
$$\omega_{1,2}^2 = \frac{1}{2}\left( \omega_{a}^2 + \omega_{b}^2 \pm \sqrt{(\omega_{a}^2 - \omega_{b}^2)^2 + 4 \kappa_{12}^2} \right)$$
where $\omega_a$ and $\omega_b$ are the uncoupled eigenfrequencies of the pharyngeal and oral cavities, and $\kappa_{12}$ is the acoustic coupling factor dependent on the constriction area $A_c$.
In biphonic vocalization mechanics, the practitioner adjusts the tongue position to decrease $V_2$ and enlarge $V_1$. By expanding $V_1$, $F_1$ shifts downward; simultaneously, retracting and elevating the tongue raises $F_1$ into a range between 800 and 1200 Hz. Concurrently, by minimizing the anterior length $L_2$ and volume $V_2$, $F_2$ is elevated out of its nominal speech envelope and shifted downward or upward toward $F_1$.
When the cross-sectional area of the constriction approaches a critical limit, the boundary conditions permit the two resonant poles to merge. This convergence eliminates formant-dispersion, compounding the acoustic amplifications of $F_1$ and $F_2$:
$$|T_{\text{compound}}(\omega)| \approx |T_{F1}(\omega)| \cdot |T_{F2}(\omega)|$$
This compounding produces an acoustic peak with a high quality factor:
$$Q = \frac{\omega_{\text{peak}}}{\Delta\omega_{-3\text{dB}}} > 35$$
The vocal tract standing waves align, phase-locking constructive reflections directly back toward the laryngeal outlet, creating a stable resonant envelope around the target harmonic.
Ventricular Fold Phonation and Subharmonic Period-Doubling Bifurcations
While sygyt achieves its spectral purity via linear and supraglottal filtering of an adducted glottal source, kargyraa introduces non-linear dynamics into the excitation mechanism itself. Laryngeal biomechanics during kargyraa involve the recruitment of the ventricular folds (plicae vestibulares), situated superior to the true vocal cords. Under heightened subglottic pressure ($P_s > 3.0\text{ kPa}$) and medial laryngeal constriction, Bernoulli forces and tissue viscoelasticity destabilize the ventricular folds.
The interaction between the true vocal folds and the ventricular mucosal tissue induces a period-doubling-bifurcation. The physical mechanism is modeled as two aerodynamic oscillators coupled in series:
$$m_1 \ddot{x}_1 + r_1 \dot{x}1 + k_1 x_1 = F{\text{aero},1}(x_1, x_2, P_s)$$
$$m_2 \ddot{x}_2 + r_2 \dot{x}2 + k_2 x_2 = F{\text{aero},2}(x_1, x_2, P_s)$$
where $m_1, r_1, k_1$ represent the mass, damping, and stiffness of the true vocal folds, and $m_2, r_2, k_2$ correspond to the mechanical properties of the ventricular folds.
Because the ventricular folds possess greater mass and lower tissue tension, their natural oscillation frequency sits below that of the true folds. Aerodynamic driving pulses from the glottis induce limit-cycle oscillations in the ventricular folds at exactly half the fundamental frequency of the true vocal fold oscillations:
$$f_{\text{ventricular}} = \frac{1}{2} f_0$$
This period-doubling inserts subharmonic spectral lines between every integer harmonic of the true glottal comb:
$$f_m = m \left(\frac{f_0}{2}\right), \quad m \in {1, 2, 3, 4, \dots}$$
The acoustic spectrum gains partials at odd multiples of $f_0 / 2$. This mechanism accounts for the deep timbre of kargyraa, whose acoustic energy drops into a bass register without requiring an abnormally long biological vocal tract.
Empirical Evidence & Observational Data: Spectrographic Profiling and Aerodynamic Metrics
Narrowband FFT Analysis of Sygyt Harmonic Trajectories
High-resolution Fast Fourier Transform (FFT) spectrographic investigations clarify the quantitative distribution of energy in sygyt. In standard operatic phonation, the “singer’s formant” aggregates $F_3$, $F_4$, and $F_5$ across a broad band between 2.5 kHz and 3.2 kHz, spanning a 700 Hz window with a moderate quality factor ($Q \approx 4\text{ to }6$). This distribution balances acoustic energy to help an unamplified voice carry over an orchestra.
Energy Profile: Operatic Singer's Formant (Broadband Distribution)
Amp
^ .-.
| / \ (Broadband peak over F3, F4, F5)
| / \
| .--' '--.
+--------+----+-----+----+-------------------------------------> Freq (kHz)
2.0 2.5 3.0 3.5
Energy Profile: Tuvan Sygyt Biphonic Phonation (Discrete Line Spectrum)
Amp
^ | (Discrete overtone isolation: partial n = 9)
| | Harmonic Bandwidth < 45 Hz
| | Peak-to-Adjacent Ratio > 28 dB
| |
±-------±-------±-----±------------------------------------> Freq (kHz)
f0 n*f0 (n+1)*f0
Tuvan sygyt demonstrates an alternative acoustic topology. Narrowband spectrographic metrics show that up to 65% of total radiated acoustic power can be confined to a single integer harmonic. In an FFT profile with a 2048-point Hanning window, the isolated harmonic displays an acoustic bandwidth narrower than 45 Hz at the -3 dB margin. Neighboring partials ($n-1$ and $n+1$) are attenuated by 20 to 35 dB relative to the reinforced harmonic peak.
Harmonic Number (n) Center Frequency (Hz) Relative Amplitude (dB) Bandwidth (-3 dB, Hz)
-------------------------------------------------------------------------------------------------
Fundamental (f_0) 352.0 -12.4 18.2
Partial n = 6 2112.0 -24.8 26.4
Partial n = 7 2464.0 -21.5 28.1
Partial n = 8 (Target) 2816.0 0.0 (Reference Peak) 32.0
Partial n = 9 3168.0 -28.7 29.5
Partial n = 10 3520.0 -34.1 34.2
The table above demonstrates this selective filtration during a steady-state sygyt vocalization on fundamental pitch $F_4$ ($f_0 = 352\text{ Hz}$). The eighth harmonic (2816 Hz) functions as the radiated pitch center, while the fundamental drone serves as a lower acoustic anchor.
Impedance Transduction at the Velopharyngeal Port
A crucial requirement for achieving the high-Q filtering characteristic of sygyt is the absolute decoupling of the nasal cavity from the acoustic path. The human nasopharynx, lined with vascularized mucosal tissue, introduces heavy acoustic damping through viscous dissipation and thermal conduction.
In aerodynamic experiments measuring nasal airflow ($V_{\text{nasal}}$), overtone practitioners maintain complete velopharyngeal closure ($V_{\text{nasal}} = 0\text{ mL/s}$). The musculus levator veli palatini and musculus constrictor pharyngis superior contract firmly, pressing the soft palate against the posterior pharyngeal wall.
Quantitative MRI and aerodynamic studies (Story et al., 1996; Titze, 2008) reveal distinct cross-sectional area profiles for high-Q vocal tract resonances. If a velopharyngeal leakage gap of merely $A_{\text{nasal}} \ge 0.03\text{ cm}^2$ opens, the transfer function introduces an acoustic zero at:
$$f_{\text{zero}} \approx \frac{c}{2\pi} \sqrt{\frac{A_{\text{nasal}}}{V_{\text{nasal-cavity}} \cdot l_{\text{port}}}}$$
This zero attenuates the compound formant amplitude by 14 to 18 dB, causing the isolated overtone to collapse back into a standard nasalized vowel. High transglottal driving pressures ($P_s \in [2.5, 4.2]\text{ kPa}$) demand an input acoustic resistance ($R_g \ge 120\text{ acoustic ohms}$) that can only be sustained when the velopharyngeal port provides an acoustic seal.
These aerodynamic constraints demonstrate that sygyt requires a complete acoustic seal along the secondary branch to prevent energy dissipation.
Nasal Shunt Dynamic:
Velum Open (A_nasal > 0) ===> Formant Zero Injected ===> High-Q Peak Collapses
Velum Closed (A_nasal = 0) ===> Pole Reinforcement ===> High-Q Overtone Maintained
Comparative Energetics: Normal Phonation vs. Formant Focused Output
The metabolic and aerodynamic expenditure required for overtone phonation differs significantly from speech and classical singing. In standard conversational modal voice, subglottic pressure hovers between 0.5 and 1.0 kPa, with a transglottal volume velocity of approximately 150 to 250 mL/s. Radiated acoustic efficiency rarely exceeds 0.01% to 0.1%.
Parameter Conversational Speech Operatic Bel Canto Tuvan Sygyt
-------------------------------------------------------------------------------------------------
Subglottic Pressure (P_s) 0.5 - 1.0 kPa 1.5 - 2.5 kPa 2.5 - 4.2 kPa
Transglottal Flow (U_mean) 150 - 250 mL/s 100 - 200 mL/s 40 - 80 mL/s
Closed Quotient (Q_c) 0.40 - 0.50 0.55 - 0.65 0.70 - 0.85
Peak Formant Bandwidth 120 - 250 Hz 80 - 150 Hz 25 - 45 Hz
Harmonic Reinforcement (F_n/F_0) -18 to -30 dB -6 to -12 dB +18 to +26 dB
In sygyt, the singer elevates $P_s$ to between 2.5 and 4.2 kPa while restricting mean transglottal airflow to less than 80 mL/s, and in some practitioners below 40 mL/s. This behavior corresponds to high vocal fold adduction, with the glottal closed quotient reaching $Q_c \in [0.70, 0.85]$.
Rather than releasing energy through high volume flow, the glottis operates against high downstream acoustic impedance. Energy transfer efficiency is concentrated into the narrow bandwidth of the aligned formants. The human airway functions not as a passive dissipative pipe, but as a tuned acoustic transmission line that maximizes output power within a narrow spectral window.
Comparative Acoustic Typologies: Western Phonation vs. Central Asian Overtone Systems
Acoustic Impedance Matching vs. Radiative Dissipation
The primary divergence between Western classical singing (such as the bel canto school) and Central Asian biphonic traditions lies in the physical management of acoustic-impedance at the airway boundaries. The Western operatic tradition models the vocal tract as an impedance matching horn designed to transfer energy smoothly from the glottis to a free radiation field across a wide frequency range. The operatic vocal tract minimizes internal reflections that could destabilize vocal fold oscillation, maintaining a broad acoustic output that fills large performance spaces.
Western Bel Canto Phonation:
[Glottis] === High Inertance ===> [Tract as Radiative Horn] === Wide Energy Transfer ===> [Auditorium]
Tuvan Sygyt Phonation:
[Glottis] === High Resistance ===> [Dual-Chamber Resonator] === Reflected Phase Trap ===> [Monochromatic Beam]
Central Asian overtone systems use internal acoustic reflections constructively. In sygyt, the tract creates an internal acoustic trap. The severe lingual constriction establishes a localized standing wave between the tongue and pharyngeal structures, trapping energy within the posterior and anterior resonators.
Radiation from the lips is minimized for all frequencies except the narrow compound formant band. The system functions as a narrow band-pass filter, reflecting non-targeted harmonics back toward the glottis while transmitting the selected overtone as an acoustic line.
Linear Filter Approximation vs. Non-linear Source-Tract Interaction
Standard acoustic descriptions of speech rely on linear source-filter theory (Fant, 1960), which assumes that the glottal source and supraglottal resonator function independently without meaningful physical feedback:
$$P_{\text{rad}}(s) = U_g(s) \cdot T(s) \cdot R(s)$$
While this linear approximation describes normal vowels reasonably well, it breaks down under the acoustic conditions of biphonic singing.
As demonstrated by Titze (2008), when vocal tract formants are tuned close to individual harmonics of the glottal comb, the supraglottal vocal tract input impedance ($Z_{\text{in}}$) feeds back directly onto the glottal dynamics. When the input impedance of the vocal tract is reactive (inertive) at the frequency of vocal fold oscillation, acoustic pressure reflections assist the tissue in closing more rapidly during each cycle.
In sygyt, this source-tract interaction is pronounced. The compound formant peak presents a mechanical load back to the glottis. Reflected acoustic pressure cycles modulate the glottal flow pulses non-linearly, sharpening the instant of glottal closure and generating higher spectral energy across upper partials. This circular feedback loop demonstrates non-linear vocal mechanics: the filter actively modifies its own excitation source.
Morphological Variances Across Western Voicing and Eastern Phonation
Sygyt (Supraglottal Formant Focus)
- Primary Excitation Source: True vocal folds (plicae vocales) vibrating in a hyper-adducted, high-impedance modal register with an extended closed quotient ($Q_c > 0.70$).
- Acoustic Chamber Division: Tract bifurcated by an alveolar-palatal lingual constriction into a wide pharyngeal cavity and a small oral cavity ($V < 5\text{ cm}^3$).
- Harmonic Selection: Extraction and dynamic tracking of high-order single partials ($n \in [6, 16]$); all adjacent partials attenuated by >25 dB.
- Ventricular Fold Dynamic: Ventricular folds remain abducted or hyper-adducted statically to narrow the laryngeal outlet without self-oscillating.
- Spectral Radiation: Quasi-sinusoidal acoustic line spectrum displaying an anomalous high quality factor ($Q > 30$).
Kargyraa (Glottal/Ventricular Phonation)
- Primary Excitation Source: Dual oscillatory sources: true vocal fold vibration coupled aerodynamically to ventricular fold (plicae vestibulares) vibration.
- Acoustic Chamber Division: Broad pharyngeal expansion accompanied by epiglottic retroflexion, establishing a deepened uniform acoustic waveguide.
- Harmonic Selection: Period-doubling bifurcation producing an exact subharmonic frequency division ($f_0 / 2$), with odd intermediate partials throughout.
- Ventricular Fold Dynamic: Ventricular folds undergo limit-cycle mechanical oscillations, striking each other once for every two cycles of the true vocal folds.
- Spectral Radiation: Dense, multi-harmonic comb with rich low-frequency energy extending downward into the sub-fundamental bass envelope.
These morphological profiles demonstrate the physiological versatility of the human vocal tract. By adjusting tissue compliance, constriction areas, and aerodynamic flow rates, practitioners access diverse mechanical states, transitioning between subharmonic frequency division and narrow-band filtering.
Metaphysical Implications & Unified Synthesis: Cymatics, Sacred Ratio, and Morphogenetic Acoustics
Harmonic Series as Integer Foundations of Cymatic Geometry
The extraction of pure harmonics in overtone singing provides an acoustic demonstration of cymatics-modal-geometry. When longitudinal acoustic waves propagate through a medium and encounter defined boundaries, they establish standing wave patterns characterized by nodes (points of zero acoustic displacement) and antinodes (points of maximum displacement).
In two-dimensional and three-dimensional spaces, these nodal distributions dictate the geometric organization of matter, as demonstrated in classical Chladni plate dynamics:
$$\nabla^2 \psi - \frac{1}{c^2}\frac{\partial^2 \psi}{\partial t^2} = 0 \quad \text{subject to} \quad \left.\frac{\partial \psi}{\partial \vec{n}}\right|_{\partial \Omega} = 0$$
Linear Harmonic Spectrum 2D Cymatic Modal Distribution
---------------------------------------------------------------------------------
f_0 Fundamental (Low Drone) ===> Radial Symmetry / Macro-Boundary Ring
n = 6 to 8 (Mid-register Sygyt) ===> Hexagonal / Octagonal Geometric Arrays
n = 12 to 16 (High Sygyt Whistle) ===> Dense Tessellated High-Order Nodes
The vocal tract serves as a biological instrument that isolates individual integers of the harmonic series:
$$f_n = n \cdot f_0, \quad n \in \mathbb{N}$$
By shifting the compound formant from one integer harmonic to another ($n = 6, 7, 8, 9, \dots$), the overtone singer modifies the nodal geometry of the standing waves inside the airway. The transitions heard in Tuvan sygyt are not merely aesthetic; they reflect fundamental structural modes of acoustic wave mechanics.
Standing Waves as Biological Waveguides and Form-Generative Fields
The human vocal tract in sygyt functions as an organic waveguide. By elevating the tongue blade and curling the tongue tip backward, the singer converts soft tissue into an acoustic boundary that enforces precise boundary conditions:
Tongue Blade Elevated (Z -> Inf) Anterior Lip Aperture (A_lip -> 0)
| |
v v
[ Acoustic Velocity Node ] [ Acoustic Pressure Antinode ]
| |
+======== Dual-Cavity Standing Wave ====++These spatial distributions generate stable standing waves within the anterior oral chamber, with a quarter-wavelength matched to the cavity length:
$$\lambda_n = \frac{4 L_{\text{anterior}}}{2m - 1}$$
These standing waves generate significant acoustic radiation pressure gradients inside the micro-cavity, demonstrating how bio-acoustic systems can shape dynamic energy fields through internal geometry.
Harmonic Purity and Coherence in Archeoacoustic Architecture
The extraction of high-order integer partials links human vocal production with physical space. In megalithic acoustic enclosures and sacred stone chambers, the resonant behavior of the space is defined by its architectural eigenmodes, analyzed within archaeoacoustic-megalithic-resonators. Standard speech and broad vocalization scatter energy randomly across the room’s reflective surfaces, exciting competing reflections that create phase cancellation and diffuse reverberation.
In a rectangular stone enclosure with rigid boundaries, the resonant eigenfrequencies (modal room resonances) are governed by the Rayleigh formula:
$$f_{p,q,r} = \frac{c}{2} \sqrt{\left(\frac{p}{L_x}\right)^2 + \left(\frac{q}{L_y}\right)^2 + \left(\frac{r}{L_z}\right)^2}$$
where $p, q, r$ represent modal integers and $L_x, L_y, L_z$ designate chamber dimensions. When an overtone singer aligns a compound vocal formant ($n f_0$) with an architectural eigenmode ($f_{p,q,r}$), the human-tract resonator and the architectural cavity lock into an acoustic coupled system:
$$\text{Coupled System: } \quad Z_{\text{total}}(\omega) = Z_{\text{tract}}(\omega) + Z_{\text{chamber}}(\omega)$$
This coupling reduces radiative losses, amplifying ambient sound pressure and establishing longitudinal-waves that remain stable throughout the architectural volume.
Under these conditions, the performer uses precise vocal tuning to probe the resonant characteristics of the physical environment, using harmonic sound to activate the natural frequencies of the architectural space.
Frequently Asked Questions: Advanced Technical Inquiries into Biphonic Mechanics
Physiological Limits of the Overtone Extraction Spectrum
What physical and anatomical constraints dictate the absolute upper frequency boundary for isolated overtones in sygyt?
The upper limit for harmonic extraction in sygyt sits near 4000 to 4500 Hz, typically corresponding to the 16th through 18th harmonics of a standard male fundamental pitch ($f_0 \approx 250\text{ to }300\text{ Hz}$). This limit is enforced by three primary factors:
First, the resonant frequency of the anterior oral cavity depends inversely on its physical volume:
$$F \approx \frac{c}{2\pi} \sqrt{\frac{A_{\text{constriction}}}{V_{\text{anterior}} \cdot l_{\text{effective}}}}$$
To drive the anterior cavity resonance above 4500 Hz, $V_{\text{anterior}}$ must shrink below $1.5\text{ cm}^3$. Human oral anatomy cannot reduce this volume further without having the tongue press completely against the hard palate and teeth, which closes the cavity entirely and stops acoustic transmission.
Second, as cross-sectional areas shrink and frequencies climb, acoustic boundary layer losses scale with the square root of frequency:
$$R_{\text{loss}} \propto \sqrt{\omega}$$
Viscous drag against oral mucosa and thermal exchange with tissue dissipate wave energy, lowering the cavity’s quality factor ($Q$).
Third, the glottal comb source exhibits a spectral roll-off of at least -6 to -12 dB per octave; above 4 kHz, the raw harmonic energy delivered by the glottis diminishes, making it difficult to amplify overtones above ambient noise levels.
The Role of the Ventricular Folds in Subharmonic Modulation
Do the ventricular folds oscillate during the high-pitched sygyt style, or is their activation restricted to kargyraa?
In sygyt, the ventricular folds do not oscillate. Endoscopic video-stroboscopy confirms that mechanical oscillation of the ventricular folds occurs in kargyraa, where they serve as a secondary acoustic oscillator that induces a period-doubling bifurcation ($f_0 / 2$).
During sygyt, the ventricular folds adopt an entirely different mechanical role: they hyper-adduct or constrict statically without mucosal oscillation. This static supraglottic constriction forms a narrow laryngeal exit nozzle (aryepiglottic funnel), which increases the acoustic input impedance ($Z_{\text{in}}$) of the vocal tract directly above the true vocal cords.
Style Ventricular Fold Action Mechanical Consequence
-------------------------------------------------------------------------------------------
Sygyt Static Adduction / Constriction Acoustic impedance matching nozzle (No oscillation)
Kargyraa Dynamic Aerodynamic Oscillation Period-doubling bifurcation (f0 / 2 subharmonic)
This nozzle stabilizes aerodynamic pressure across the glottis and facilitates source-tract back-coupling, strengthening high partials in the glottal volume velocity pulse without introducing subharmonic distortion.
Thermal and Viscous Boundary Losses in Micro-Vocal Cavities
Why is unmodulated turbulent transglottal airflow so detrimental to the production of high-Q overtone lines?
Acoustic transductance in sygyt relies on high coherence within the vocal tract. If the vocal folds fail to adduct completely during the closed phase—permitting unmodulated, turbulent transglottal air leakage—two acoustical failures occur immediately:
First, the unmodulated airflow introduces broad-spectrum aerodynamic noise across the mid-frequency band via turbulent vortex shedding at the glottal aperture:
$$p_{\text{turbulent}} \propto \rho U_0^2$$
This acoustic noise fills the spectral valleys between harmonics, reducing the spectral contrast between the target overtone and surrounding frequencies, which degrades the perceived purity of the biphonic line.
Second, incomplete glottal closure introduces high resistive losses at the lower boundary of the vocal tract. The open glottis couples the tract to the subglottic trachea and pulmonary network, which introduces substantial acoustic damping. This damping lowers the quality factor of the supraglottal cavities:
$$Q = \frac{\omega M}{R_{\text{internal}} + R_{\text{glottis}} + R_{\text{radiation}}}$$
When $R_{\text{glottis}}$ drops due to an open glottal aperture, total dissipation surges, and the compound formant broadens. To isolate an overtone with a bandwidth under 45 Hz, the singer must maintain complete glottal closure ($Q_c \ge 0.70$), sealing the acoustic waveguide at its source and maximizing energy concentration in the compound formant.