Timbre ​

Objective ​

Understand timbre—the spectral "fingerprint" of sound—as a physical phenomenon arising from harmonic content, resonant filtering, material properties, and temporal evolution. Explore how the complete sound-producing chain—from the energy source through the vibrating body and its material to the resonating enclosure and the surrounding space—shapes the sound we hear, and how the human voice unifies all of these principles into the most flexible acoustic instrument ever evolved.

Relevance to visual music theory ​

Timbre is the primary axis along which visual music artists map colour, texture, shape, and motion to sound. A bright trumpet and a dark cello at the same pitch and loudness demand entirely different visual treatments. Understanding the physics of timbre—spectral content, spectral envelope, spectral evolution over time, noise floor, and formant structure—provides the precise, measurable vocabulary required for meaningful audiovisual translation. Every visual parameter you assign to a sound source should be grounded in a specific timbral dimension described in this unit.

Prerequisites and Unit Boundaries ​

This unit treats timbre as a physical and acoustic phenomenon: the measurable spectral and temporal properties of sound sources and the spaces they occupy. It builds directly on the wave physics of Unit 3.1 (Acoustics) and the physiological frequency analysis of Unit 3.2 (Hearing). It does not address the psychoacoustic judgment of how simultaneous timbres interact to produce consonance or dissonance (→ Unit 3.4), nor how cultures organise the pitch relationships embedded in the harmonic series into tuning systems (→ Unit 3.5). What you will carry forward from this unit:

• The complete sound-producing chain: energy source → vibrating body → material → resonator → space.
• The harmonic series as the spectral basis of tonal sound.
• Inharmonicity as the spectral basis of bells, drums, and metallic percussion.
• Formants as fixed resonant filters independent of pitch.
• The ADSR envelope and the spectral transient as twin components of temporal identity.
• The voice as the universal reference instrument that maps timbre to language.
• Reverberation and echo as timbral signatures of space.
• The Fourier Transform and the spectrogram as the tools that make all of the above visible.

I. What Is Timbre? ​

Definition ​

Timbre is The quality of sound that distinguishes different sources producing the same pitch at the same loudness—the composite spectral and temporal signature by which the auditory system identifies a sound's origin. The American National Standards Institute (ANSI) defines it formally as "that attribute of auditory sensation in terms of which a listener can judge that two sounds similarly presented and having the same loudness and pitch are dissimilar."

The Hook: Why Does a Piano Sound Like a Piano? ​

Play middle C (261.63 Hz) on a piano, a flute, and a violin. The fundamental frequency is identical. The loudness can be matched. Yet the brain identifies each instrument instantly—even from a single note, even in a fraction of a second. This discrimination is timbre. It is not a single physical quantity. It is an emergent property arising from at least five interacting dimensions:

1. Harmonic (spectral) content: Which overtones are present and at what relative amplitudes.
2. Spectral envelope: The overall shape of the energy distribution across frequency—the "silhouette" of the spectrum.
3. Temporal envelope: How the amplitude evolves over time—the attack, decay, sustain, and release.
4. Spectral evolution: How the spectrum itself changes over time—especially during the attack transient.
5. Noise and inharmonic components: The non-periodic, stochastic elements that add texture, grain, and "life."

Key insight: We read timbre unconsciously and instantaneously. When you hear a sound, your auditory system decodes the entire production chain—what kind of energy excited what kind of body, made of what material, resonating in what enclosure, in what kind of space. Timbre is the auditory answer to the question: what made this sound, and where?

II. The Sound-Producing System ​

Every acoustic sound can be decomposed into a chain of physical stages. Each stage shapes the final timbre, and the listener's brain reverses the chain to infer the source.

1. The Energy Source (Excitation) ​

The initial input of energy determines the character of the attack transient and the spectral content of the raw vibration:

• Impact: A strike (piano hammer, drumstick, mallet). Produces a broadband impulse rich in high frequencies, followed by a rapid spectral narrowing as the resonant modes of the struck body emerge.
• Pluck: A sudden displacement and release (guitar, harp, harpsichord, pizzicato). Similar to impact but with a different initial waveform determined by the shape and position of the plucking point.
• Friction (Bowing): A continuous stick-slip excitation (violin, cello, erhu, hurdy-gurdy). The bow maintains energy input, producing a sustained, harmonically rich waveform with a characteristic sawtooth-like shape due to the Helmholtz motion of the bowed string.
• Air stream: A jet of air directed across an edge (flute, recorder, organ pipe) or through a vibrating reed (clarinet, oboe, saxophone) or buzzing lips (trumpet, trombone). The nature of the air–edge or air–reed interaction determines the initial harmonic recipe.
• Vocal fold oscillation: Subglottal air pressure from the lungs sets the vocal folds into vibration. The mucosal wave of the folds produces a quasi-periodic pulse train rich in harmonics—the raw "buzzing" source before any resonant filtering.

2. The Vibrating Body (Oscillator) ​

The physical element that sustains oscillation and determines the fundamental frequency: Strings: The frequency of a vibrating string is governed by: $$f = \frac\sqrt{\frac{\mu}}$$ where $L$ is the vibrating length, $T$ is the tension, and $\mu$ is the linear mass density (mass per unit length). Changing any of these three parameters changes the pitch. A guitarist shortens $L$ by pressing a fret; a violinist increases $T$ by tuning the peg; a piano uses strings of increasing $\mu$ (thicker, wound strings) for lower notes. Air columns: The resonant frequency of an air column depends on its length and its boundary conditions:

• Open pipe (open at both ends, e.g., flute, organ flue pipe): supports all harmonics. $f_n = \frac{nv}{2L}$, $n = 1, 2, 3, 4...$
• Closed pipe (closed at one end, e.g., clarinet, some organ pipes): supports only odd harmonics. $f_n = \frac{nv}{4L}$, $n = 1, 3, 5, 7...$ This boundary-condition difference is the primary reason a clarinet and a flute of similar length sound fundamentally different (see §V). Membranes (2D): A circular drumhead vibrates in modes described by Bessel functions, not simple sine waves. Its overtone frequencies are not integer multiples of the fundamental (see §VI). Plates, bars, and bells (2D/3D): These stiff, often geometrically complex bodies produce strongly inharmonic overtone series determined by their shape, thickness profile, and material properties (see §VI). Vocal folds: Two mucosal folds in the larynx that vibrate when subglottal pressure exceeds the threshold of phonation. Their vibration frequency (the fundamental frequency of the voice, $f_0$) is controlled by the tension, mass, and length of the folds, adjusted by the laryngeal muscles. Typical $f_0$ ranges: bass voice ~80–300 Hz; soprano ~250–1000 Hz.

3. Material Properties ​

The physical substance of the vibrating body determines which overtones are emphasised and how quickly they decay:

Practical rule: A steel string sustains longer and sounds brighter than a nylon string at the same pitch because steel has lower internal damping and higher stiffness. A wooden xylophone bar sounds warmer than a metal glockenspiel bar because wood absorbs high-frequency energy more readily through internal friction.

4. The Resonator (Filter) ​

The resonator selectively amplifies certain frequencies of the raw vibration while attenuating others. It acts as a spectral filter applied to the vibrating body's output. The resonator's own resonant frequencies (determined by its geometry, volume, and material) are the frequencies that will be amplified. This is the critical distinction: the vibrating body determines pitch; the resonator determines timbre. A violin string vibrating in free air is nearly inaudible. Coupled to the violin body, the same string produces a rich, powerful, projecting tone—because the body's resonant modes amplify specific frequency bands and radiate them efficiently into the air. Types of resonators in instruments:

III. Resonance and Resonators (Detailed) ​

1. Helmholtz Resonance ​

A Helmholtz resonator is an air cavity with a narrow neck. The air in the neck oscillates as a mass; the air in the cavity compresses and expands as a spring. The system has a single, sharply tuned resonant frequency: $$fH = \frac\sqrt{\frac{V \cdot L}}}$$ where $v$ = speed of sound, $A$ = neck cross-sectional area, $V$ = cavity volume, and $L_{\text{eff}}$ = effective neck length (including end corrections for radiation impedance).

Demonstration: Blow across the top of a bottle. The pitch depends on the volume of air inside. Add water (reducing $V$): the pitch rises. This is a tunable Helmholtz resonator. The American folk instrument jug (as in jug band music) exploits exactly this principle—the player buzzes their lips into the mouth of a ceramic jug, and the jug's Helmholtz resonance amplifies and colors the sound. Helmholtz resonance in instruments:

• The main air resonance of a guitar body (typically ~90–110 Hz) is a Helmholtz mode: air oscillates in and out of the soundhole, with the cavity acting as the spring. This mode is responsible for the guitar's low-frequency warmth.
• The bass reflex port on a loudspeaker cabinet is a precisely tuned Helmholtz resonator designed to extend bass response by coupling the speaker's back-radiation to the room.

2. Structural Resonances ​

Beyond the air cavity, the physical walls of an instrument body vibrate with their own resonant modes:

• A violin top plate (spruce) has dozens of distinct vibrational modes, each amplifying a specific frequency band. The lowest modes (~280 Hz and ~460 Hz on a fine violin) are the most important for the instrument's "voice." The legendary quality of Stradivari and Guarneri violins is partly attributed to the specific modal structure of their plates—a result of wood selection, arching geometry, thickness graduation, and 300 years of aging.
• A piano soundboard (spruce) has a dense forest of resonant modes that collectively amplify the entire audible range. The soundboard is the piano's true "speaker": without it, the strings would produce almost no audible sound.
• A drum shell (wood or metal) has its own resonances that colour the drumhead's vibration. A maple snare drum shell and a steel snare drum shell at the same diameter and head tension produce noticeably different timbres—the wood shell absorbs more high-frequency energy, producing a warmer tone.

3. The Brass Bell: Impedance Matching as Timbral Design ​

The flared bell of a brass instrument (trumpet, trombone, horn, tuba) serves a specific acoustic function: gradual impedance matching between the high-impedance standing wave inside the narrow tube and the low-impedance free air outside. Without the bell, most of the wave's energy would reflect back into the tube at the open end (due to the impedance mismatch described in Unit 3.1, §IV). The bell's exponential flare reduces this reflection, allowing sound to radiate efficiently—but it does so in a frequency-dependent manner. Low frequencies, whose wavelengths are much larger than the bell, still reflect back (sustaining the standing wave inside the instrument). High frequencies, whose wavelengths are comparable to the bell diameter, radiate outward. The cutoff frequency of the bell determines which harmonics escape and which remain trapped. A wider bell (tuba) has a lower cutoff, radiating more bass. A narrower bell (trumpet) has a higher cutoff, producing a brighter, more penetrating sound. The bell is thus a high-pass filter whose cutoff frequency is a primary determinant of brass timbre.

IV. The Harmonic Series: The Spectral Foundation of Tonal Sound ​

1. Definitions ​

• Fundamental frequency ($f_0$): The lowest frequency of a periodic vibration. It determines the perceived pitch.
• Harmonics: Frequencies that are exact integer multiples of the fundamental: $f_0, 2f_0, 3f_0, 4f_0, ...$. The fundamental is the 1st harmonic.
• Overtones: All frequencies above the fundamental. The 1st overtone = 2nd harmonic; the 2nd overtone = 3rd harmonic, etc.
• Partials: The most general term for any frequency component of a sound, whether harmonic or not. All harmonics are partials, but not all partials are harmonics.

2. The Natural Overtone Series ​

The harmonic series is one of the most important structures in all of acoustics and music. For a fundamental of 100 Hz:

The harmonic series encodes music. The intervals between successive harmonics are the very intervals that form the foundation of musical systems worldwide: octave (2:1), perfect fifth (3:2), perfect fourth (4:3), major third (5:4), minor third (6:5). The harmonic series is not an abstract mathematical construct—it is a physical fact of vibrating bodies, and it is the reason these intervals "exist" in nature before any culture codifies them. Unit 3.5 will explore how human societies have organised these natural ratios into tuning systems. The Compressing Series: The intervals between successive harmonics become progressively smaller as you ascend: octave, fifth, fourth, major third, minor third, and ever-smaller intervals. The series naturally compresses toward a dense cluster at high frequencies. This compression is the physical reason why the upper partials of a complex tone blend into a continuous spectral texture rather than standing out as individually audible pitches.

3. Standing Waves: How Harmonics Arise ​

Harmonics are not abstract mathematical entities. They are standing waves—physical vibrational patterns that arise when a wave reflects between two boundaries and interferes with itself. On a string fixed at both ends: The string can vibrate in modes where its length accommodates exactly $n$ half-wavelengths: $$\lambda_n = \frac, \quad f_n = \frac = n \cdot f_0$$ The 1st mode ($n = 1$) is the fundamental; the 2nd mode ($n = 2$) has a node at the midpoint (frequency = $2f_0$, one octave higher); the 3rd mode ($n = 3$) has two nodes (frequency = $3f_0$, an octave plus a fifth higher), and so on. In normal playing, a string vibrates in all modes simultaneously. The relative amplitudes of the modes—determined by how and where the string is excited—create the instrument's characteristic timbre. Natural harmonics (flageolets) on strings: By lightly touching a vibrating string at a nodal point (without pressing it to the fretboard), a player suppresses the fundamental and all modes that require an antinode at that point, isolating a single harmonic:

• Touch at the midpoint ($L/2$): isolates the 2nd harmonic (one octave above the open string).
• Touch at $L/3$: isolates the 3rd harmonic (an octave plus a fifth above).
• Touch at $L/4$: isolates the 4th harmonic (two octaves above). The resulting tone is ethereal and flute-like—a near-pure sine wave stripped of the fundamental's complexity. This technique, used on guitar, violin, cello, harp, and many other string instruments, provides a direct physical demonstration that the harmonic series is real and audible.

V. Timbral Signatures by Instrument Family ​

The harmonic recipe of an instrument is determined by the interplay of its oscillator type, geometry, boundary conditions, and excitation mechanism. Here we survey the major families.

1. Strings (Chordophones) ​

General character: Strings fixed at both ends support all harmonics (both odd and even). The relative amplitude of each harmonic depends on:

• Excitation point: Plucking at the midpoint ($L/2$) suppresses even harmonics; plucking near the bridge emphasises high harmonics (brighter, more "twangy").
• Excitation method: A pluck produces a sharp initial spectrum that decays progressively from the top. A bow produces a sustained sawtooth-like waveform (Helmholtz motion) rich in all harmonics.
• String material and construction: Steel strings are brighter (lower damping, higher stiffness) than nylon or gut. Wound strings (a core wrapped with wire) increase mass without excessive stiffness, allowing low-pitched strings on manageable instrument sizes. Piano inharmonicity (quantified): Real piano strings have finite stiffness, which adds a bending-resistance restoring force alongside the tension restoring force. This causes the overtones to be progressively sharper than exact integer multiples: $$f_n = n f_0 \sqrt$$ where $B$ is the inharmonicity coefficient, typically $10^$ to $10^$ for piano strings. The higher the partial, the sharper it becomes. On a concert grand, the 16th partial may be ~20 cents sharp. This physical reality is the reason piano tuners "stretch" octaves (Unit 3.5): they tune upper octaves slightly sharp and lower octaves slightly flat so that the inharmonic partials of different notes align better across the keyboard.

2. Air Columns (Aerophones) ​

The harmonic content of wind instruments is governed by the boundary conditions of the air column (open vs. closed ends) and the excitation mechanism (air jet, single reed, double reed, lip buzz). Open cylindrical pipe (flute, organ flue pipe):

• Supports all harmonics: $f_n = nf_0$, $n = 1, 2, 3, 4...$
• The flute's characteristic purity is not primarily a consequence of its pipe geometry (which allows all harmonics) but of its excitation mechanism: the air jet at the embouchure is a gentle, quasi-linear driver that preferentially excites the fundamental. The jet's weak non-linearity means upper harmonics have relatively low amplitude. The resulting spectrum is dominated by the fundamental with a gently declining harmonic series—close to a sine wave with subtle overtone coloring.
• Breath noise at the embouchure edge adds a continuous broadband component to the spectrum. This noise is not a defect; it is integral to the flute's "airy" identity. Closed cylindrical pipe (clarinet):
• Supports only odd harmonics: $f_n = nf_0$, $n = 1, 3, 5, 7...$
• The single reed is a strongly non-linear valve: it opens and closes abruptly against the mouthpiece, generating a rich harmonic spectrum. The closed-pipe geometry then filters this spectrum, suppressing even harmonics.
• The result is the clarinet's characteristic hollow, dark, woody quality in the low (chalumeau) register—physically identical to a square wave's odd-harmonic signature.

Register-dependent caveat: The odd-harmonic suppression is most pronounced in the lowest register. When the player overblows, the clarinet jumps not to the 2nd harmonic (an octave, as a flute would) but to the 3rd harmonic (a twelfth)—because the 2nd harmonic is suppressed. In the upper registers (clarion, altissimo), tone-hole effects and reed non-linearity allow even harmonics to emerge, brightening the tone. Common error in introductory texts: Many sources mistakenly attribute odd-harmonic emphasis to the flute and even-harmonic emphasis to the clarinet. This is physically backwards. The closed-pipe geometry of the clarinet is what suppresses even harmonics. Conical bore instruments (oboe, saxophone, brass):

• A conical bore, even when closed at one end by a reed or mouthpiece, supports all harmonics (both odd and even). The oboe's rich, nasal, penetrating quality comes from its strongly non-linear double reed exciting a full harmonic series, which is then shaped by the narrow conical bore into a characteristically bright spectral envelope.
• Brass instruments (trumpet, trombone, horn) use a cup or funnel mouthpiece where the player's lips buzz, generating a rich harmonic source. The long conical/cylindrical bore and the flared bell (§III.3) filter and radiate this spectrum. A trumpet has a brighter spectrum (more high harmonics radiated by its relatively small bell) than a French horn (whose very large bell radiates lower frequencies more efficiently, producing a darker, rounder tone).

3. Struck and Plucked Bars (Idiophones) ​

Xylophone and marimba (wooden bars):

• A bar free at both ends vibrates in bending modes whose frequencies follow: $$f_n \propto (2n+1)^2$$
• The overtones are not integer multiples of the fundamental. The first overtone of an untreated bar is at approximately 2.76 × $f_0$—between a musical octave (2×) and a twelfth (3×).
• Instrument makers undercut (thin) the bars at specific points to shift the first few overtones into more harmonic alignment. On a well-tuned marimba, the first overtone is tuned to exactly 4:1 (two octaves); the second overtone to approximately 10:1.
• The wood's high internal damping rapidly attenuates high-frequency overtones, producing a warm, mellow, "round" tone with a relatively fast decay. Glockenspiel and vibraphone (metal bars):
• Same physics as wooden bars, but metal (steel, aluminium) has much lower internal damping. High-frequency overtones ring much longer, producing a bright, shimmering, "metallic" sustain.
• The vibraphone adds a motor-driven rotating disc in the resonator tube beneath each bar, periodically opening and closing the tube. This modulates the amplitude of the fundamental at ~3–6 Hz, creating the vibraphone's characteristic vibrato (technically an amplitude tremolo, not a frequency vibrato).

4. Membranes (Membranophones): The Drum Spectrum ​

A circular membrane (drumhead) fixed at its perimeter vibrates in two-dimensional modes described by Bessel functions. The modal frequencies are: $$f = \frac\sqrt{\frac{\sigma}}$$ where $\alpha_$ are the zeros of the Bessel functions, $R$ is the membrane radius, $T$ is the tension, and $\sigma$ is the surface mass density. The resulting overtone ratios for a circular membrane are:

1. The heavy copper bowl beneath the drumhead acts as a loaded boundary, altering the modal frequencies.
2. Careful tensioning and head selection shift the first several overtone ratios closer to a harmonic series (approximately 1 : 1.5 : 2 : 2.5...), allowing the ear to fuse them into a recognisable pitch.

Visual mapping note: The inharmonic spectrum of drums and metal percussion is the acoustic analogue of non-repeating, asymmetric visual patterns (fractals, Voronoi tessellations, organic textures). Harmonic sounds map naturally to periodic, symmetric, crystalline visual structures; inharmonic sounds map to aperiodic, organic, amorphous ones. This distinction is a foundational rule for visual music grammar.

5. Bells and Gongs: Complex 3D Vibrators ​

Bells and gongs are among the most acoustically complex instruments. Their three-dimensional geometry and non-uniform thickness produce overtone series with strongly non-integer ratios: Church bell (typical):

VI. The Temporal Envelope: ADSR and Beyond ​

The ADSR Model ​

The temporal evolution of a sound's amplitude is conventionally modelled as four phases:

Important clarification: ADSR is a synthesiser convention, not a physical law. Real acoustic instruments have far more complex envelopes. A piano note has no true "sustain"—it decays continuously from the moment of impact. A bowed string has a quasi-steady state with complex micro-fluctuations in amplitude and spectrum. A struck bell has multiple partials decaying at different rates (the high partials die in milliseconds; the fundamental can ring for minutes). ADSR is a useful first approximation, but treating it as a complete description of temporal evolution would be an oversimplification.

The Spectral Transient: The Hidden Dimension ​

The ADSR model describes amplitude evolution, but during the attack phase, the spectrum itself evolves rapidly. This spectral transient is often more important for instrument identification than the amplitude envelope. Examples:

• A piano hammer strike initially excites a dense, broadband burst of energy (including very high frequencies from the hammer impact and string noise). Within 20–50 ms, these high-frequency components decay, leaving the steady-state harmonic series. The initial burst is the "percussive" identity of the piano.
• A bowed violin attack begins with a chaotic "scratch" noise (the bow catching the string) that transitions into the periodic Helmholtz sawtooth motion within 30–100 ms. The scratch is essential for identification—without it, the violin sounds like an organ.
• A trumpet attack begins with a burst of broadband lip noise before the standing wave in the tube stabilises. This transient "spit" is a defining feature of brass articulation.

Perceptual experiment: If you remove the first ~50 ms of a recorded piano note and play the remainder, listeners can no longer reliably identify it as a piano. It sounds like a harmonium or an organ. The attack transient is the acoustic "fingerprint" of the excitation mechanism.

VII. The Fourier Transform and Spectral Analysis ​

Decomposing Complex Sound ​

The Fourier Transform is the mathematical operation that decomposes any complex waveform into its constituent sinusoidal components, each defined by a frequency, an amplitude, and a phase: $$X(f) = \int_^ x(t) , e^ , dt$$ In practice, digital audio uses the Fast Fourier Transform (FFT), a computationally efficient algorithm that operates on discrete, windowed segments of a signal.

• Spectrum: A snapshot of the frequency content at a single moment—amplitude (and optionally phase) as a function of frequency. A spectrum reveals the harmonic recipe of a sound: which overtones are present and how strong they are.
• Spectrogram: A time-frequency representation showing how the spectrum evolves over time. Frequency is mapped to the vertical axis, time to the horizontal, and amplitude to colour or brightness. The spectrogram is the most powerful single tool for visualising timbre.

Spectral Descriptors ​

Beyond visual inspection, a spectrum can be characterised by numerical descriptors—measurable, single-number summaries of spectral shape. These are essential for automated audiovisual mapping:

The Role of Phase ​

A spectrum has two components: magnitude (amplitude of each frequency) and phase (the timing offset of each sinusoidal component). For steady-state tones, the ear is largely phase-deaf—altering the relative phase of harmonics in a sustained sound has little effect on perceived timbre (a fact known since Ohm's acoustic law and confirmed by Helmholtz). However, during transients (attack, decay), phase relationships are critical to the waveform's shape, which in turn affects the spectral transient and thus the perceived attack character. This is relevant for additive synthesis and for any visual representation that attempts to display waveform shape rather than spectral content.

VIII. Noise as a Timbral Component ​

Definition ​

In acoustics, noise is any sound produced by a collection of uncoordinated, non-periodic sources—a signal with no discernible fundamental frequency or harmonic structure. Noise is characterised not by specific frequencies but by the statistical distribution of energy across the spectrum.

Types of Noise ​

Visual mapping bridge: This correspondence between acoustic noise colours and visual noise textures is not a metaphor—it is a direct mathematical equivalence. White noise in audio (flat power spectrum) and white noise in an image (random, uncorrelated pixel values) share the identical statistical structure. Pink noise (1/f spectrum) in audio and Perlin noise in computer graphics both exhibit self-similar, fractal structure with equal energy per octave. Brown noise and deep visual blur both have spectra dominated by low frequencies. Any visual music system that maps audio noise to visual texture can use this correspondence directly.

Noise in Acoustic Instruments ​

Noise is not an impurity to be eliminated. It is an essential component of nearly every acoustic timbre:

• Flute: Breath noise at the embouchure edge (the turbulent air jet that fails to excite the resonance) is integral to the flute's "airy" identity. Remove it digitally, and the flute sounds like a cheap synthesiser.
• Violin: The "scratch" of the bow at attack; the continuous low-level rosin noise during sustain.
• Piano: The hammer impact noise ("thump"), the damper lift noise, and the sympathetic rattle of undamped strings.
• Snare drum: The buzz of the snare wires against the bottom head is a broadband noise component that defines the instrument. Without the wires, it is a tom.
• Voice: The aspiration noise of the consonants /h/, /s/, /f/, /sh/ is entirely aperiodic. Speech is a continuous alternation between periodic (voiced) and noisy (unvoiced) segments.

IX. The Human Voice: The Universal Reference Instrument ​

The human voice occupies a unique position in this unit. It is not merely one more instrument to catalogue; it is the instrument that every human already knows intimately. Our lifelong experience with voices makes the voice the natural reference point for explaining all timbral concepts. More importantly, the voice is the physical bridge between abstract acoustic phenomena and the concrete, communicative world of speech and language.

1. The Source-Filter Model ​

The voice operates on a source-filter model (Gunnar Fant, 1960):

• Source (the larynx): The vocal folds vibrate as air from the lungs passes through the glottis, producing a quasi-periodic pulse train rich in harmonics. The spectrum of this raw buzz falls off at approximately 12 dB per octave—a declining harmonic series with strong low harmonics and progressively weaker high harmonics.
• Filter (the vocal tract): The pharynx, oral cavity, and nasal cavity form a tube of approximately 17 cm in adult males (shorter in women and children). This tube has its own resonant frequencies (formants), which selectively amplify certain harmonics of the source signal while attenuating others. The crucial insight: the source determines pitch; the filter determines timbre (vowel quality, voice character). When a singer changes pitch, the harmonic spacing changes but the formant positions remain approximately fixed. When a speaker changes from /a/ to /i/ without changing pitch, the harmonics stay the same but the formant positions shift dramatically.

Analogy: Think of the harmonics as a comb with teeth at $f_0, 2f_0, 3f_0, ...$. The formants are a fixed envelope draped over the comb. When the singer raises pitch, the comb teeth spread apart (wider spacing), but the envelope stays in place. Different teeth now fall under the envelope peaks, changing which harmonics are amplified—but the overall timbral "shape" remains recognisable as the same vowel.

2. Formants: The Timbral DNA of Speech ​

Formants are the resonant frequency peaks of the vocal tract. They are named $F_1, F_2, F_3, ...$ in ascending order. The first two formants ($F_1$ and $F_2$) are the primary determinants of vowel identity:

Key physical fact: $F_1$ correlates inversely with tongue height (high tongue → low $F_1$, and vice versa). $F_2$ correlates with tongue frontness/backness (front tongue → high $F_2$; back tongue → low $F_2$). The entire vowel space of a language can be plotted as a two-dimensional $F_1 \times F_2$ chart—a visual map of articulatory gesture to acoustic output. Higher formants ($F_3$, $F_4$, $F_5$): contribute to voice quality, speaker identity, and the perception of "brightness" or "darkness" of a particular voice. They are less important for vowel identity but critical for distinguishing individual speakers (forensic voice analysis).

3. The Singer's Formant: Acoustic Projection Without Amplification ​

Trained classical (operatic) singers develop a remarkable spectral peak in the range 2,500–3,200 Hz, variously called the "singer's formant," the "ring," or the "carrying power" of the voice. This peak arises from the clustering of the 3rd, 4th, and 5th vocal tract resonances ($F_3$, $F_4$, $F_5$), achieved through:

• Precise lowering of the larynx.
• Widening of the pharyngeal space (lowered epiglottis, expanded hypopharynx).
• Narrowing of the epilaryngeal tube (the short tube just above the vocal folds). Why it works: The orchestral accompaniment has its maximum spectral energy between 400–600 Hz, with a natural dip in the 2–4 kHz region. The singer's formant places maximum vocal energy precisely in this dip, allowing the voice to cut through a 70-piece orchestra without electronic amplification. It is a solution discovered empirically by centuries of vocal pedagogy and explained physically by resonance acoustics. The ear canal resonance (Unit 3.2, §I.2) peaks at ~3,400 Hz, overlapping precisely with the singer's formant band. The singer is, in effect, broadcasting on the exact frequency to which the listener's ear is most sensitive.

4. The Low Formant: Warmth and "Body" ​

In contrast to the high singer's formant, the frequency range 350–700 Hz (approximately G3–F5) encompasses what singers and voice teachers describe as the "low formant" or "chest resonance." This region contributes:

• Warmth and richness to the voice.
• Perception of "massiveness" and fullness.
• The sensation of chest vibration (vibratory feedback from the thorax and sternum). The balance between the low formant (warmth, body) and the singer's formant (brilliance, projection) defines the aesthetic character of different voice types: a deep bass emphasises the low formant; a lyric soprano emphasises the singer's formant.

5. The Resonance Theory of Singing (Bel Canto and Morozov) ​

The resonance theory of singing, rooted in the Italian Bel Canto tradition (from the late 16th century), holds that the vocal tract must function as a maximally efficient resonator. The goal is not to push more air through the folds (force-based singing) but to shape the resonances so that the acoustic output is maximised for a given vocal fold effort. Historical aphorisms of the Bel Canto masters:

• "Sing without disturbing the flame of a candle" — efficient resonance amplifies the voice; brute-force air pressure does not.
• "Sound should be in the nose, but there should be no nose in the sound" — the sensation of resonance in the nasal/pharyngeal cavities without actual nasalisation.
• "A voice devoid of resonance is stillborn and cannot carry." — Giacomo Lauri-Volpi. V.P. Morozov's Research (1950s–present): Russian scientist Vladimir Petrovich Morozov formalised the resonance theory and identified seven critical functions of the singing resonators:

1. Energetic: Resonators amplify vocal sound by increasing the acoustic efficiency (КПД) of the vocal apparatus—more loudness per unit of vocal fold effort.
2. Generative: Resonators are not passive filters. They are an integral, active part of the sound-generation system itself, participating in the oscillation process through acoustic feedback to the vocal folds.
3. Phonetic: Resonators shape vowels and consonants, enabling intelligible diction.
4. Aesthetic: Resonators determine the core aesthetic qualities of the singing voice: brilliance (zvonkost'), softness (myagkost'), projection or "flight" (polyotnost'), voice type, and vibrato.
5. Protective: Well-organised resonators protect the larynx and vocal folds from overload through seven direct and indirect mechanisms. The primary protective mechanism: by amplifying sound through resonance and redistributing its spectral energy into the range of maximal auditory sensitivity (2–4 kHz), resonators free the singer from the need to strain the larynx to achieve the necessary loudness. The direct protective mechanism involves strong reactive (back-pressure) effects from the resonators that facilitate the oscillation of the vocal folds, reducing the collision force required.
6. Indicator: The physical vibration of resonators—felt in the face, chest, skull, and sternum—serves as a proprioceptive feedback mechanism, allowing the singer to "tune" their resonator system in real time by biofeedback (the "feel" of a well-placed note).
7. Activating (Malyutin Effect): Vibration of the resonators reflexively increases the muscular tone of the larynx, vocal folds, and the entire vocal apparatus. Resonance activates the whole system, enhancing coordination and responsiveness.

6. The Voice as Universal Timbral Reference ​

The voice's supreme importance for this course lies in its role as the Rosetta Stone of timbre:

• Formants demonstrate that resonator shape determines timbre independently of pitch. Every student already knows this intuitively—they can sing any vowel at any pitch.
• The harmonic series is directly audible in overtone singing (Tuvan, Mongolian throat singing), where singers isolate individual harmonics of a single sustained drone by precisely tuning their vocal tract formants to align with specific harmonics.
• The source-filter model is the same architecture used by all acoustic instruments: a vibrating source (string, reed, lip buzz) filtered by a resonator (body, bore, bell). The voice simply makes this architecture maximally transparent because we control every parameter consciously.
• Noise in speech (the /s/, /f/, /sh/ consonants) is the same broadband, aperiodic energy that gives cymbals, snare drums, and the flute their non-tonal character. Speech is, physically, a continuous alternation between tonal (harmonic) and noisy (stochastic) sound—exactly the alternation that occurs in any complex musical texture.

X. What Sound Tells Us About the Space: Reverberation and Echo ​

The timbre of a sound is not determined solely by its source. The acoustic environment in which the sound propagates adds its own spectral and temporal signature. We hear the space along with the source.

1. Direct Sound, Early Reflections, Late Reverberation ​

When a sound source emits a pulse in an enclosed space, the listener receives three temporally distinct components:

1. Direct sound (0 ms): The first-arriving wavefront, traveling the shortest path from source to listener. This carries the "purest" representation of the source's timbre.
2. Early reflections (1–80 ms): Discrete reflections from nearby surfaces (floor, ceiling, walls). These arrive within the integration time of the auditory system and are fused with the direct sound (the Precedence Effect, Unit 3.2, §V.4). They provide the brain with information about room size and shape.
3. Late reverberation (>80 ms): A dense, diffuse cloud of thousands of overlapping reflections. The reverberation tail decays exponentially. Its rate of decay, measured as RT60 (the time for sound to decay by 60 dB), is the primary metric of a room's acoustic character: | Space | Typical RT60 | Acoustic character | |---|---|---| | Recording studio (dead room) | 0.2–0.4 s | Dry, intimate, "close" | | Living room | 0.3–0.6 s | Natural, conversational | | Concert hall | 1.5–2.5 s | Warm, enveloping, "alive" | | Cathedral | 4–8+ s | Massive, ethereal, blurred | | Anechoic chamber | ~0 s | No reflections; unsettling, "dead" |

2. Reverberation as a Spectral Filter ​

Reverberation is not spectrally neutral. High frequencies are absorbed more readily by air, surfaces, and furnishings than low frequencies (Unit 3.1, §III.1). As a result, the reverberation tail of a room acts as a progressive low-pass filter: the longer the sound decays, the more the high-frequency content is stripped away, leaving a warm, bass-heavy residue. This is why a cathedral organ sounds massive and enveloping: the long reverberation time sustains the low harmonics for seconds while the high harmonics die away quickly, creating a spectral tilt toward bass that fuses successive notes into a continuous harmonic wash.

3. Echo vs. Reverberation ​

An echo is a discrete, audible repetition of a sound arriving more than ~50 ms after the direct sound (the threshold at which the Precedence Effect fails and the brain perceives a separate event). Echoes require a large, reflective surface at sufficient distance (at least ~17 m for a 50 ms delay in air). Reverberation is the dense, continuous decay of thousands of overlapping reflections arriving too close together to be perceived as individual echoes. The distinction is temporal density: sparse, separated reflections = echo; dense, overlapping reflections = reverberation.

Visual mapping note: Reverberation is the acoustic equivalent of motion blur, glow, or bloom in visual design. Just as motion blur smears a moving object across time, reverberation smears a sound's energy across time. Short reverberation = sharp, crisp visual edges. Long reverberation = soft, glowing, diffuse visual halos. A dry recording maps to hard-edged, high-contrast visuals; a reverberant recording maps to soft-edged, atmospheric visuals.

XI. The Missing Fundamental: Pitch Without a Frequency ​

One of the most remarkable demonstrations that timbre and pitch are psychoacoustically linked: If you play a series of harmonics—say 200, 300, 400, 500 Hz—but physically remove the fundamental (100 Hz), the brain still perceives the pitch as 100 Hz. The auditory system infers the fundamental from the spacing pattern of the upper harmonics (each pair is separated by 100 Hz). This is why:

• Small speakers (phone speakers, laptop speakers) that cannot physically reproduce frequencies below 200 Hz can still convey the impression of bass. The brain reconstructs the missing fundamental from the higher harmonics.
• Telephone lines (bandwidth 300–3,400 Hz) can transmit recognisable voice pitch even though the fundamental frequency of most male voices (~100–150 Hz) falls below the transmission band. The missing fundamental is a powerful demonstration that pitch is a percept, not a frequency. It is computed by the auditory system from the harmonic template, not simply read off the lowest frequency present.

XII. Summary ​

Key Takeaways ​

1. Timbre is multidimensional: It arises from the interaction of harmonic content, spectral envelope, temporal envelope (ADSR), spectral transient (attack evolution), noise floor, and the acoustic signature of the space.
2. Every sound is a chain: Energy source → vibrating body → material properties → resonator → space. Each stage shapes the final timbre, and the listener's brain reverses the chain to identify the source and its environment.
3. The harmonic series (integer multiples of the fundamental) is the spectral basis of all tonal sound. Harmonics produce no beats with each other, fusing into a single coherent pitch. The intervals of the harmonic series are the physical origin of musical intervals.
4. Inharmonic spectra (drums, bells, gongs, bars) arise from 2D/3D vibration modes governed by Bessel functions and plate mechanics. They produce sounds without definite pitch—or with ambiguous pitch—because the partials do not fuse.
5. Resonators act as spectral filters on the raw vibration: the body of a string instrument, the bore and bell of a wind instrument, the vocal tract of a singer. The resonator determines timbre; the vibrating body determines pitch.
6. Material properties (stiffness, density, internal damping, elasticity) govern which overtones are emphasised and how quickly they decay. Metal rings; wood absorbs. Steel is bright; felt is dark.
7. The human voice is the universal reference instrument. The source-filter model (vocal folds → vocal tract) is the archetype of all acoustic sound production. Formants are the resonant peaks of the vocal tract; they determine vowel identity and are independent of pitch. The singer's formant (2.5–3.2 kHz) enables operatic projection. Morozov's seven functions of vocal resonators describe a complete system of acoustic efficiency, protection, and aesthetic control.
8. Noise is an essential timbral component, not an impurity. Breath noise, bow scratch, hammer thump, and consonant aspiration are integral to the identity of their respective sounds. Acoustic noise colours (white, pink, brown) correspond directly to visual noise textures (static, Perlin, deep blur).
9. Reverberation and echo are the timbral signatures of space. Reverberation acts as a progressive low-pass filter, smearing sound across time. The RT60 metric characterises the acoustic "personality" of an environment.
10. The Fourier Transform and the spectrogram make all of the above visible. Spectral descriptors (centroid, flux, flatness, rolloff) provide measurable, mappable quantities for audiovisual translation.

Applications ​

Sound design and synthesiser programming. Audio mixing and mastering (EQ as spectral sculpting). Acoustic instrument design and lutherie. Vocal pedagogy and forensic voice analysis. Architectural acoustics and room design. Audiovisual art and visual music.

XIV. Closing Thoughts ​

Timbre is where physics becomes identity. The same mathematical laws that govern a vibrating string also shape the cry of a newborn, the roar of a cathedral organ, and the whisper of wind through a bamboo flute. Every sound carries within it a complete physical story: what was struck, bowed, plucked, or blown; what it was made of; what shape contained it; and what space surrounded it. The human auditory system reads this story in milliseconds, without conscious effort, with a discrimination that no artificial system has yet matched. For the visual music artist, timbre is the richest axis of translation. Pitch gives you a note. Loudness gives you a magnitude. But timbre gives you everything else: the colour, the texture, the shape, the grain, the material, the space, and the identity of the source. By understanding the harmonic series, resonance, formant structure, spectral evolution, and noise, you acquire the complete vocabulary for translating the infinite sonic palette into visual form. In the next unit, we follow these timbral signatures into the domain of interaction: what happens when two or more sounds—each with their own harmonic fingerprint—collide in the same critical band, and how the brain judges these collisions as consonance, dissonance, or something more complex still.

XV. Suggested Further Reading and Exploration ​

• Hermann von Helmholtz, On the Sensations of Tone as a Physiological Basis for the Theory of Music (1863/1885). The foundational text linking acoustic physics to musical perception. Helmholtz's descriptions of resonance, combination tones, and vowel synthesis remain remarkably relevant.
• Neville H. Fletcher & Thomas D. Rossing, The Physics of Musical Instruments, 2nd ed. (Springer, 1998). The definitive reference on the acoustics of every instrument family: strings, wind, percussion, and voice.
• Gunnar Fant, Acoustic Theory of Speech Production (Mouton, 1960). The foundational text on the source-filter model of voice production.
• Johan Sundberg, The Science of the Singing Voice (Northern Illinois University Press, 1987). The most accessible scientific treatment of vocal acoustics, formants, and the singer's formant.
• V.P. Morozov, Искусство резонансного пения [The Art of Resonance Singing] (Moscow, 2002). The definitive treatment of the resonance theory of singing and the seven functions of vocal resonators.
• Jean-Claude Risset, "The Perception of Musical Tones" (in The Psychology of Music, ed. Diana Deutsch, 3rd ed., 2013). Risset's pioneering work on timbral illusions, spectral morphing, and the computer synthesis of timbre.
• Stephen McAdams, "The Perception of Timbre" (in The Psychology of Music, ed. Diana Deutsch, 3rd ed., 2013). The modern review of timbre as a perceptual, multidimensional space. McAdams' work on "timbre space" maps instruments in a perceptual coordinate system—directly applicable to visual mapping.
• Harry Partch, Genesis of a Music (1949/1974). Partch's deeply original exploration of timbre, intonation, and the physical basis of musical sound, including his 43-tone scale and custom-built instruments.
• Albert Bregman, Auditory Scene Analysis (1990). How the brain groups spectral components into auditory "objects"—essential for understanding how timbre enables the separation of simultaneous sound sources.
• PRAAT: Boersma, P. & Weenink, D. (2024). Praat: doing phonetics by computer [Computer program]. http://www.praat.org. Free, open-source.
• Experiment—Cymatics and Chladni patterns: If you have access to a speaker driver and a thin metal plate, sprinkle sand on the plate and sweep the excitation frequency. Photograph the resulting nodal patterns at each resonant mode. These patterns are the visual equivalent of the harmonic series—standing wave modes made visible. Compare the increasing complexity of higher modes to the increasing density of the harmonic series.
• Experiment—Overtone singing: Listen to recordings of Tuvan khoomei or sygyt throat singing. The singer sustains a single drone (the fundamental) while dramatically shifting the vocal tract resonances to isolate and amplify individual harmonics—making the harmonic series directly audible as a melody of overtones above a fixed bass note. Try it yourself: sustain a low drone and slowly shape your mouth from /u/ to /i/, listening for the high-pitched whistling overtone that emerges.