Overview
Sound production for speech begins with the time-varying glottal airflow produced by vocal fold vibration, but sound can be generated almost anywhere along the vocal tract through turbulence, pressure release, or tissue oscillation. Understanding both the generation and propagation of sound is fundamental to voice science, as these physical processes determine the acoustic characteristics we ultimately perceive as voice quality, pitch, and loudness.
This chapter examines the physical processes underlying sound generation and propagation. We begin by exploring various sound sources in the vocal tract, focusing particularly on the glottal flow pulse and its temporal characteristics. We then investigate how these sounds can be represented as frequency spectra, introducing the crucial concept of spectral slope and its relationship to voice quality. The chapter continues with detailed coverage of wave propagation in continuous media, including the fundamental space-time relationships that govern acoustic waves. Finally, we address sound reflection at media interfaces and wave interference patterns, concepts essential for understanding vocal tract resonance in the following chapter.
What You’ll Learn
Sound Generation Mechanisms
- Physical definition of sound: Local pressure disturbances in continuous media
- Primary sound sources: Time-varying glottal airflow and vocal fold vibration
- Secondary sound sources: Turbulence, transient sounds, and liquid movement
- Source-spectrum relationships: How temporal characteristics determine frequency content
Glottal Source Characteristics
- Flow pulse parameters: Open quotient, skewing quotient, and waveform shape
- Temporal features: Duration, amplitude, and phase relationships
- Quality correlates: How source parameters relate to perceived voice quality
- Clinical applications: Acoustic assessment of glottal valving
Frequency Domain Representation
- Spectrum fundamentals: Amplitude and phase spectra of complex sounds
- Harmonic relationships: Periodicity in time and harmonicity in frequency
- Spectral slope: Measurement and perceptual significance
- Fourier synthesis: Constructing complex waveforms from sinusoids
Wave Propagation Physics
- Propagation velocity: Dependence on medium properties
- Space-time duality: Wavelength, period, and their relationships
- Sinusoidal waves: Mathematical description and physical interpretation
- Wave impedance: Pressure-velocity relationships in propagating waves
Reflection and Interference
- Media interfaces: Reflection at boundaries between different materials
- Reflection coefficients: Quantifying reflected and transmitted energy
- Standing waves: Interference patterns from forward and backward waves
- Clinical relevance: Vocal tract resonance and impedance matching
Foundational Concepts
Understanding sound generation and propagation requires integrating concepts from mechanics, fluid dynamics, and wave physics. The glottal source acts as a periodic flow modulator, creating pressure disturbances that propagate through the vocal tract as acoustic waves. These waves reflect at tract boundaries, creating complex interference patterns that form the basis of vocal resonance.
The frequency spectrum provides an alternative representation to the time-domain waveform, revealing the harmonic structure underlying periodic sounds. Spectral characteristics like slope and harmonic spacing relate directly to perceptual qualities of voice. The relationship between temporal features (like open quotient) and spectral features (like harmonic amplitudes) exemplifies the powerful time-frequency transformations central to acoustic analysis.
Wave propagation involves the transmission of local pressure disturbances through a medium without net transport of the medium itself. The speed of this propagation depends on the medium’s elastic and inertial properties, establishing fundamental relationships between frequency, wavelength, and propagation velocity. These relationships govern all acoustic phenomena in the vocal tract.
Practical Applications
The concepts in this chapter have direct applications in clinical voice assessment, voice training, and acoustic analysis. Understanding glottal source characteristics enables clinicians to interpret inverse-filtered flow signals and relate acoustic measurements to underlying physiological conditions. Spectral analysis provides tools for quantifying voice quality objectively, while knowledge of wave propagation and reflection is essential for understanding how vocal tract shape influences the acoustic output.
For voice professionals, understanding the source-filter theory of voice production begins with mastering the material in this chapter. The glottal source provides the raw acoustic energy, while the vocal tract (covered in Chapter 6) acts as an acoustic filter. The interplay between source characteristics and tract resonances determines the final radiated sound.
This chapter provides the acoustic foundation necessary for understanding vocal tract resonance, formant production, and the complete source-filter theory of voice production developed in subsequent chapters.
Check Your Understanding
Five short questions on how voice sound is generated and how it travels. Each one explains itself as soon as you answer, so treat them as a way to test your reasoning rather than a test to pass.
Correct Not quite Best answer
Harmonic spacing equals F₀, and pitch did not change. Harmonic positions are set by the repetition rate of the pulses, not by their shape; what the shape changes is how strong each harmonic is.
Correct Not quite Best answer
Sharp changes in the flow waveform produce high-frequency content. A short closing phase, complete closure, and a large skewing quotient all flatten the roll-off from the roughly −12 dB/octave of modal voice toward −6 to −10 dB/octave, and H1−H2 falls (H2 may even exceed H1). Breathy voice with gradual, incomplete closure does the reverse, with slopes of −15 to −20 dB/octave and a dominant fundamental.
Correct Not quite Best answer
A steep slope with a dominant fundamental is the signature of breathy phonation—the starting point here, not the destination. Gradual flow changes remove high frequencies; abrupt ones add them.
Correct Not quite Best answer
Closing speed within a cycle and the number of cycles per second are independent. The folds can close more abruptly at exactly the same F₀; only the pulse shape—and hence the harmonic amplitudes—changes.
Correct Not quite Best answer
The chapter is explicit that F₀ is unchanged—the vibration rate is set by the tissue's mass and tension, not by the gas it moves. What rises is the set of vocal tract resonances, not the pitch.
Correct Not quite Best answer
The direction is backwards. From λ = c/f, a faster medium gives a longer wavelength for the same frequency. It is the increase in c, not a shortening of λ, that raises the resonance frequencies for a tract of fixed length.
Correct Not quite Best answer
Propagation velocity depends only on the medium: c rises when density falls. For a tube of fixed length the resonances scale with c (F ≈ (2n−1)c/4L), so all formants shift up—by a factor of roughly 2.5 for heliox. Vowels and timbre are transformed, but the glottal source, and therefore the pitch, is untouched. The chapter notes this should only be done with supervision and medical-grade gas.
Correct Not quite Best answer
Breathing a gas does not raise the tract's pressure, and the folds' stiffness is a tissue property. The relevant change is in the gas's density and therefore the speed at which pressure disturbances travel through it.
Air particles cannot penetrate the wall, so they crowd at the interface and pressure doubles. The reflected wave has the same amplitude and polarity as the incident wave (r = +1) and travels back the way it came.
Choose an answer for this row, then check again.
Particles meet no resistance and spread out, depleting the pressure to zero at the interface. The reflection is complete but inverted (r = −1): a compression returns as a rarefaction. A vacuum reflects as effectively as a brick wall, just with opposite phase.
Choose an answer for this row, then check again.
With z₂ = z₁ the reflection coefficient is zero: nothing is reflected and all the energy is transmitted. Impedance matching is the condition for maximum energy transfer.
Choose an answer for this row, then check again.
r = (2z₁ − z₁)/(2z₁ + z₁) = +1/3. One-third of the pressure reflects with the same polarity, and the transmitted pressure is pᵢ(1 + r) = 4/3 pᵢ—the interface "transforms up" the pressure while particle velocity drops to conserve energy.
Choose an answer for this row, then check again.
Correct Not quite Best answer
The open quotient is adjusted mainly by the gap between the vocal processes; it increases with abduction. Normal values run about 0.4–0.7, and above 0.7 the steady flow u₀ grows relative to the acoustic flow u_ac, raising the chance of glottal turbulence and producing a breathy quality, especially at low pitch. Q₀ and u₀/u_ac are therefore useful acoustic indices of glottal valving.
Correct Not quite Best answer
Pressed voice goes with a low open quotient—below about 0.4—because the folds spend most of the cycle closed. A Q₀ of 0.8 means the opposite: the glottis is open for 80% of each cycle.
Correct Not quite Best answer
The typical range the chapter gives is 0.4 to 0.7. A value of 0.8 lies above it, and combined with a high u₀/u_ac ratio it signals incomplete valving rather than normal phonation.
Correct Not quite Best answer
Brassiness is tied to a large skewing quotient (rapid closure), which boosts the third harmonic—a different parameter from the open quotient. A high Q₀ with excess steady flow points to gradual, incomplete closure and reduced high-frequency energy, not a brassy edge.
Correct Not quite Best answer
Propagation velocity is a property of the medium (c = √(K/ρ)), not of the source or its effort. Loud and soft sounds travel at the same speed. Nothing the singer does changes how fast the wave moves.
Correct Not quite Best answer
At the wavelengths of voice fundamentals (1–3 m) and low formants, the mouth opening is far smaller than a wavelength and radiates nearly omnidirectionally. Directional beaming only begins for the highest frequencies, so "aiming" is not what carries the voice.
Correct Not quite Best answer
The singer cannot alter how the medium propagates the wave, but the source and resonator are very much under control: the chapter lists impedance matching at the lips, strong excitation of vocal tract resonances, and glottal flow shape as the real ingredients of "projection."
Correct Not quite Best answer
Once a disturbance is created the medium takes over and propagates it at its own velocity. What a vocalist can do is generate more acoustic power at the glottis, shape the flow pulse for useful high-frequency energy, and tune the vocal tract so resonances are strongly excited and radiate efficiently from the lips. "Projection" is a matter of source and filter, not of throwing sound.
Want more? The practice section has recall flashcards for this chapter's key terms, plus these questions one at a time.