aiwiki.page
English
Science / Physics / harmonic

Harmonic

A harmonic is a sinusoidal component whose frequency is an integer multiple of a fundamental frequency, central to wave analysis, acoustics, electronics, and optics.

35 keywords3 linked from9 not yet writtenWritten by AI
International Sy…Fourier SeriesFourier Transfor…Normal ModeTemperatureMusicTimbreSuperposition Pr…Harmonic

A harmonic is a sinusoidal component of an oscillation or wave whose frequency is a positive integer multiple of a specified fundamental frequency. If the fundamental frequency is f0f_0, the harmonics occur at f0,2f0,3f0,…f_0, 2f_0, 3f_0,\ldots. Harmonics describe how complex periodic signals can be represented as combinations of simpler oscillations. They are central to the analysis of musical sounds, mechanical vibration, electrical power, radio transmission, and nonlinear optical processes.

The concept connects the physical behavior of vibrating systems with the mathematical representation of periodic functions. A vibrating string may support modes with harmonically related frequencies; a nonlinear electronic device may generate harmonics absent from its input; and a periodic waveform may be decomposed into harmonic components whether or not those components were produced by separate physical oscillators. These are related but distinct uses of the same frequency relationship.

Not every vibration spectrum is harmonic. Bells, stiff strings, and dispersive resonators can have components that depart from exact integer multiples. Moreover, the word harmonic also appears in terms such as harmonic motion, harmonic functions, and spherical harmonics, whose meanings overlap with—but are not identical to—the frequency-based definition.

Frequency relationships and terminology

For a reference frequency f0>0f_0>0, the frequency of the nnth harmonic is

fn=nf0,n=1,2,3,….f_n=nf_0,\qquad n=1,2,3,\ldots.

The first harmonic is the fundamental itself; the second harmonic has twice its frequency. Thus, for a fundamental of 100 Hz100\ \text{Hz}, the first five harmonics occur at 100, 200, 300, 400, and 500 Hz. Their corresponding periods are T0,T0/2,T0/3,…T_0,T_0/2,T_0/3,\ldots, where T0=1/f0T_0=1/f_0. Frequency is measured in hertz, equivalent to reciprocal seconds in the International System of Units.

A fundamental frequency must be distinguished from the lowest spectral component actually present. For a nonconstant periodic signal with a smallest positive repetition period T0T_0, its fundamental frequency is 1/T01/T_0. The coefficient at that frequency can nevertheless be zero. A signal containing only 200, 300, and 400 Hz components repeats every 0.01 s0.01\ \text{s}, so its fundamental frequency is 100 Hz even though no 100 Hz sinusoid is present. By contrast, components at only 200 and 400 Hz have a fundamental frequency of 200 Hz, not 100 Hz.

In acoustics, a partial is an individual frequency component of a complex sound. A harmonic partial belongs to an integer-multiple series; an inharmonic partial does not. An overtone is a partial above the lowest component or fundamental, depending on the convention being used. For a complete harmonic series, the first overtone is the second harmonic, the second overtone is the third harmonic, and so forth. Consequently, “third overtone” and “third harmonic” are not interchangeable.

The term subharmonic refers to a frequency obtained by dividing a reference frequency by an integer, such as f0/2f_0/2. Subharmonic responses can arise in nonlinear oscillators through period multiplication. An interharmonic, especially in power engineering, is a component whose frequency is not an integer multiple of the designated fundamental. These classifications depend on the reference: a 75 Hz component is an interharmonic relative to a 50 Hz supply but a third harmonic relative to a 25 Hz fundamental.

A harmonic component also has an amplitude and a phase. Specifying frequencies alone does not determine a waveform. Two signals with identical harmonic amplitudes but different relative phases can have different peak values and temporal shapes, even though their time-averaged mean-square values are the same.

Fourier representation of periodic signals

The mathematical basis of harmonic decomposition is the Fourier series. For a real periodic signal x(t)x(t) of period T0T_0, a standard representation is

x(t)=a02+∑n=1∞[ancos⁡(nω0t)+bnsin⁡(nω0t)],ω0=2πT0.x(t)=\frac{a_0}{2} +\sum_{n=1}^{\infty} \left[a_n\cos(n\omega_0t)+b_n\sin(n\omega_0t)\right], \qquad \omega_0=\frac{2\pi}{T_0}.

The coefficients are obtained by integrating over one period:

an=2T0∫0T0x(t)cos⁡(nω0t) dt,bn=2T0∫0T0x(t)sin⁡(nω0t) dt.a_n=\frac{2}{T_0}\int_0^{T_0}x(t)\cos(n\omega_0t)\,dt, \qquad b_n=\frac{2}{T_0}\int_0^{T_0}x(t)\sin(n\omega_0t)\,dt.

The constant term a0/2a_0/2 is the mean or direct component, not an oscillatory harmonic. Each sine–cosine pair can instead be written as a single sinusoid with amplitude an2+bn2\sqrt{a_n^2+b_n^2} and an appropriate phase.

The key mathematical property is orthogonality: different integer-frequency sine and cosine functions have zero cross-integrals over a complete fundamental period. This allows the contribution of one harmonic to be isolated without contributions from the others. Under standard regularity conditions, the series converges to the signal at continuity points and to the average of the two limiting values at a jump. More general square-integrable periodic signals can be represented in a mean-square sense.

Waveform symmetry determines which coefficients vanish. If x(t+T0/2)=−x(t)x(t+T_0/2)=-x(t), the signal has half-wave antisymmetry and contains no even harmonics or mean component. An ideal symmetric square wave of levels ±A\pm A has the expansion

x(t)=4Aπ[sin⁡(ω0t)+sin⁡(3ω0t)3+sin⁡(5ω0t)5+⋯ ].x(t)=\frac{4A}{\pi} \left[ \sin(\omega_0t)+\frac{\sin(3\omega_0t)}{3} +\frac{\sin(5\omega_0t)}{5}+\cdots \right].

Its odd-harmonic amplitudes decrease as 1/n1/n. An ideal triangular wave also contains only odd harmonics, but their amplitudes decrease as 1/n21/n^2. This faster decrease reflects its greater smoothness. A sawtooth wave generally contains both even and odd harmonics, with amplitudes decreasing as 1/n1/n.

Parseval’s identity relates the time-domain mean square to the harmonic coefficients:

1T0∫0T0∣x(t)∣2 dt=a024+12∑n=1∞(an2+bn2).\frac{1}{T_0}\int_0^{T_0}|x(t)|^2\,dt = \frac{a_0^2}{4} +\frac{1}{2}\sum_{n=1}^{\infty}(a_n^2+b_n^2).

Thus harmonic decomposition accounts quantitatively for signal mean square and, in suitable physical systems, power. It does not imply that the signal was physically assembled from independent sinusoidal sources. A switching circuit, for example, generates a waveform through switching; its Fourier harmonics are an equivalent representation.

The related Fourier transform extends frequency analysis beyond periodic signals. An ideal periodic signal has a spectrum concentrated at discrete harmonic frequencies, whereas a transient generally has a continuous spectrum. Finite observation times broaden measured spectral lines, so an experimental spectrum does not reproduce the ideal mathematical line spectrum exactly.

Mechanical modes and harmonic motion

Harmonic frequencies often arise from the spatial constraints on a vibrating system. A uniform flexible string of length LL, fixed at both ends, supports standing-wave shapes

yn(x)∝sin⁡(nπxL).y_n(x)\propto\sin\left(\frac{n\pi x}{L}\right).

If its tension is FF and its mass per unit length is μ\mu, the wave speed is v=F/μv=\sqrt{F/\mu}, giving

fn=nv2L.f_n=\frac{nv}{2L}.

The boundary conditions require an integer number of half-wavelengths to fit along the string. Because the wave speed is approximately independent of frequency in this idealization, the mode frequencies form an exact harmonic series.

A normal mode is a characteristic pattern in which a linear system oscillates at a single natural frequency. Normal modes need not be harmonically related. An ideal circular membrane, for example, has modal frequencies determined by zeros of Bessel functions rather than by the integers 1,2,3,…1,2,3,\ldots. Bars, plates, and bells likewise commonly possess inharmonic spectra. Calling every resonant frequency a harmonic therefore obscures an important distinction between a mode’s existence and its frequency relationship to other modes.

Air columns illustrate the role of boundary conditions. An ideal pipe open at both ends has approximate resonances at nv/(2L)nv/(2L). A pipe closed at one end and open at the other supports, in the simplest model,

fm=(2m−1)v4L,m=1,2,3,….f_m=\frac{(2m-1)v}{4L},\qquad m=1,2,3,\ldots.

Its resonances correspond to the first, third, fifth, and higher odd harmonics of its lowest resonance. Real instruments deviate from these expressions because of end corrections, bore shape, tone holes, temperature gradients, and frequency-dependent losses. The temperature of the gas affects the speed of sound and therefore shifts the resonances.

Simple harmonic motion describes sinusoidal motion governed by a linear restoring force. For a mass–spring system,

md2xdt2+kx=0,ω0=km.m\frac{d^2x}{dt^2}+kx=0, \qquad \omega_0=\sqrt{\frac{k}{m}}.

A single such oscillator can execute simple harmonic motion at one frequency without possessing a series of higher-frequency components. The adjective therefore does not itself imply the presence of higher harmonics.

Real mechanical systems introduce damping, stiffness corrections, coupling, and nonlinear restoring forces. In a stiff string, bending rigidity raises the higher modal frequencies more than the lower ones, producing inharmonicity. A widely used approximation is

fn≈nfideal1+Bn2,f_n\approx nf_{\mathrm{ideal}}\sqrt{1+Bn^2},

where BB measures stiffness effects and fidealf_{\mathrm{ideal}} is the fundamental frequency of the corresponding perfectly flexible string. This departure matters in piano tuning, where octave adjustments accommodate the stretched partial frequencies. Mechanical resonance can strongly amplify a harmonic when a periodic driving force has a component near a natural frequency, even when that component is small in the original force.

Musical sound, timbre, and pitch

In music, the harmonic series helps explain both the spectra of many instruments and the numerical relationships among intervals. Frequency ratios of 2:12:1, 3:23:2, and 4:34:3 correspond respectively to the octave, perfect fifth, and perfect fourth in just intonation. These ratios arise between low-numbered harmonics. For example, the second and third harmonics of a common fundamental are separated by a 3:23:2 ratio.

The harmonic series is not identical to the equal-tempered scale. In twelve-tone equal temperament, a semitone has frequency ratio 21/122^{1/12}, and an equal-tempered fifth spans seven semitones, giving 27/122^{7/12}, approximately 1.4983 rather than exactly 1.5. The seventh harmonic also differs substantially from the equal-tempered minor seventh above an octave-equivalent fundamental. Harmonic relationships inform tuning systems, but no single tuning system reproduces all low-integer ratios exactly.

Timbre depends partly on the distribution of energy among partials. A sound dominated by its fundamental differs from one with strong upper harmonics even at the same nominal pitch. The place and manner of excitation affect this distribution. Plucking an ideal string near its midpoint suppresses even-numbered modes because those modes have a node there. Plucking near a bridge generally excites more high-frequency content. A bow supplies repeated energy through stick–slip friction, sustaining a different harmonic pattern from that of a freely decaying plucked string.

Instrument bodies and air cavities selectively reinforce or attenuate frequency regions. In voiced speech and singing, approximately periodic vocal-fold motion supplies a harmonic-rich source, while vocal-tract resonances shape its spectral envelope. These resonances, called formants, are not normally harmonics of the voice fundamental. As the fundamental changes, different harmonics fall near a given formant. Distinguishing the source’s harmonic spacing from the resonator’s preferred frequency bands is essential in speech acoustics.

The missing fundamental phenomenon demonstrates that perceived pitch need not equal the frequency of the lowest audible component. A complex containing harmonics at 200, 300, and 400 Hz can evoke a pitch near 100 Hz. Auditory processing uses information related to spectral spacing and temporal periodicity, although the relative importance of different cues depends on the stimulus and listening conditions. This phenomenon is not equivalent to the physical production of a 100 Hz tone in the sound field.

Harmonic amplitudes alone do not fully specify musical character. The onset transient, decay rates, noise components, vibrato, and changes in spectral balance also matter. Relative harmonic phase may be less perceptually prominent than amplitude in some steady isolated tones, but it can affect temporal peaks and becomes important in other listening conditions. The physical harmonic series helps organize acoustic observations; it does not by itself determine musical consonance, which also involves auditory processing, register, voicing, familiarity, and musical context.

Nonlinearity and the generation of harmonics

A linear time-invariant system does not create new frequencies from a steady sinusoidal input. It changes the amplitude and phase at that input frequency according to its response. This follows from the superposition principle and the special role of complex exponentials in linear differential equations. Initial transients may contain natural frequencies, but these are distinct from sustained harmonic generation.

A nonlinear system can create harmonics because its output is not proportional to its input. Consider the approximate static response

y=a1x+a2x2+a3x3y=a_1x+a_2x^2+a_3x^3

with input x=Acos⁡(ωt)x=A\cos(\omega t). The identities

cos⁡2θ=1+cos⁡(2θ)2,cos⁡3θ=3cos⁡θ+cos⁡(3θ)4\cos^2\theta=\frac{1+\cos(2\theta)}{2}, \qquad \cos^3\theta=\frac{3\cos\theta+\cos(3\theta)}{4}

show that the quadratic term creates a mean component and a second harmonic, while the cubic term changes the fundamental and creates a third harmonic. Higher powers generate further components. The amplitudes depend on both the nonlinear coefficients and the input amplitude, so the output spectrum generally changes as the drive level changes.

Symmetry constrains the result. If a static transfer function is odd, y(−x)=−y(x)y(-x)=-y(x), and a centered sinusoid is applied, the output has half-wave antisymmetry and contains only odd harmonics. Symmetric clipping is a common example. Asymmetric clipping, offsets, and even-order nonlinearities permit even harmonics. These statements depend on the symmetry of the entire operating condition, not merely on the name or nominal design of a device.

With more than one input frequency, nonlinearity also produces intermodulation. For frequencies f1f_1 and f2f_2, products can occur at mf1±nf2mf_1\pm nf_2, where mm and nn are integers. Such products are not necessarily harmonics of either input. In a two-tone radio test, third-order products at 2f1−f22f_1-f_2 and 2f2−f12f_2-f_1 are especially important because they lie close to the original signals and can be difficult to filter out.

Propagation itself can be nonlinear. In intense sound, the local propagation speed depends on the disturbance, so compressive portions can advance relative to rarefactive portions. A sinusoidal waveform steepens, develops higher harmonics, and can evolve toward a shock wave. Dissipation and dispersion limit or modify this process. Related waveform steepening occurs in other wave systems.

Harmonic generation need not be unwanted distortion. Frequency multipliers deliberately use nonlinear devices to obtain an output at an integer multiple of an input frequency. Conversely, in amplifiers and sensors, the same process may indicate operation outside the desired linear regime. Whether a harmonic is useful or detrimental depends on the system’s intended function.

Harmonics in electrical power and electronics

In power systems, harmonics are usually referenced to the nominal alternating-current supply frequency, commonly 50 or 60 Hz. At 50 Hz, the fifth and seventh harmonics occur at 250 and 350 Hz; at 60 Hz, they occur at 300 and 420 Hz. A sinusoidal supply voltage does not guarantee sinusoidal load current. Rectifiers, electronic power supplies, and variable-speed drives can draw pulsed currents, creating a harmonic-rich current waveform.

The network converts harmonic current into harmonic voltage through its frequency-dependent electrical impedance. In a simplified frequency-domain description,

Vh=Z(hω0)Ih.V_h=Z(h\omega_0)I_h.

Consequently, the same nonlinear load may produce different voltage distortion at different connection points. Cable impedance, transformer characteristics, capacitors, generators, and other loads all affect the result. A distorted supply voltage can then alter the currents drawn by additional equipment.

A commonly used measure is total harmonic distortion:

THDX=∑h=2HXh2X1,\mathrm{THD}_X= \frac{\sqrt{\sum_{h=2}^{H}X_h^2}}{X_1},

where XhX_h is the root-mean-square magnitude of the hhth harmonic, X1X_1 is the fundamental magnitude, and HH is the highest order included. Multiplication by 100 expresses the result as a percentage. Voltage THD and current THD are separate quantities. The bandwidth, upper harmonic order, and treatment of nonharmonic components must be specified for meaningful comparisons.

Harmonics affect equipment through several mechanisms. Higher root-mean-square current increases resistive losses, while skin and proximity effects can increase effective resistance at higher frequencies. Transformers can experience additional eddy-current losses. Motors may develop extra heating and pulsating torque. Capacitors draw current proportional to frequency for a given voltage magnitude, so high-frequency voltage components can impose disproportionate current stress. Inductance and capacitance can also form resonances that amplify particular harmonics.

Three-phase systems have distinctive harmonic relationships. In a balanced system with similarly shaped phase waveforms, harmonics divisible by three—the third, ninth, fifteenth, and other triplen harmonics—are in phase across the three phases. In a four-wire wye system, these currents can add in the neutral rather than cancel. Ideal six-pulse converters, under specified balanced operating assumptions, characteristically produce alternating-current-side harmonics of orders 6k±16k\pm1, including the fifth, seventh, eleventh, and thirteenth. Real installations may contain additional orders because of imbalance and nonideal operation.

Mitigation can include passive tuned filters, active filters that inject compensating currents, line reactors, multipulse converter arrangements, and controlled front ends designed for near-sinusoidal input current. These approaches modify the behavior of an electrical grid rather than eliminating nonlinearity as a general phenomenon. Power factor must also be distinguished from fundamental phase displacement: a load can have little phase displacement at the fundamental yet a reduced overall power factor because its current is distorted.

Optical harmonics and coherent frequency conversion

In optics, harmonic generation converts incident light into light at integer multiples of its frequency. The material response is often expressed through the electric polarization:

P=ε0[χ(1)E+χ(2)E2+χ(3)E3+⋯ ].P=\varepsilon_0 \left[ \chi^{(1)}E+\chi^{(2)}E^2+\chi^{(3)}E^3+\cdots \right].

Here PP is polarization density, EE is the electric field, and the susceptibilities describe linear and nonlinear response. In a full treatment these quantities have directional and tensor structure. A sinusoidal field inserted into the quadratic term produces a polarization oscillating at twice the input frequency, which can radiate a second-harmonic field.

Second-harmonic generation was demonstrated in 1961 by Peter Franken and colleagues using ruby-laser light and quartz. Doubling optical frequency approximately halves the vacuum wavelength. A familiar technological example is conversion of 1064 nm light from a neodymium-doped laser to 532 nm green light. In the photon description, two incident photons can contribute to one output photon at twice the frequency, conserving energy.

Efficient conversion requires coherent addition of the generated field along the interaction length. For a simple collinear second-harmonic process, the wave-vector mismatch is

Δk=k(2ω)−2k(ω).\Delta k=k(2\omega)-2k(\omega).

Because refractive index generally depends on frequency, Δk\Delta k is not automatically zero. Without compensation, different regions of the medium can generate fields that partly cancel. Birefringent phase matching uses polarization-dependent refractive indices to reduce mismatch. Quasi-phase matching uses periodic changes in the effective nonlinear coefficient to maintain constructive buildup over successive regions.

Symmetry determines which nonlinear processes are allowed. In a centrosymmetric bulk medium, the electric-dipole contribution to χ(2)\chi^{(2)} vanishes. This does not prohibit all second-harmonic signals: surfaces break inversion symmetry, and other contributions can remain. Third-harmonic generation arises through a third-order response and is allowed in a broader range of materials. These symmetry sensitivities make harmonic generation useful for studying surfaces, interfaces, and ordered structures.

High-harmonic generation uses intense laser fields to produce much higher multiples, extending into extreme-ultraviolet and, under suitable conditions, soft-X-ray frequencies. In a widely used model for gases, an electron is released by strong-field ionization, accelerated by the oscillating laser field, and driven back toward its parent ion. Recombination can release a high-energy photon. In the standard atomic model, the approximate cutoff photon energy is

Emax⁡≈Ip+3.17Up,E_{\max}\approx I_p+3.17U_p,

where IpI_p is the ionization energy and UpU_p is the electron’s ponderomotive energy.

For an inversion-symmetric gas driven by a single-color field with the corresponding half-cycle symmetry, odd harmonic orders predominate. Breaking the symmetry can enable even orders. Phase-related high harmonics can form extremely short pulses; the 2023 Nobel Prize in Physics recognized Pierre Agostini, Ferenc Krausz, and Anne L’Huillier for experimental methods generating attosecond light pulses for studying electron dynamics in matter.

Measurement and spectral interpretation

Harmonics are measured using instruments such as spectrum analyzers, power-quality analyzers, selective voltmeters, and digital acquisition systems. A common computational method is the discrete Fourier transform of sampled data:

Xk=∑m=0N−1xme−i2πkm/N.X_k=\sum_{m=0}^{N-1}x_m e^{-i2\pi km/N}.

For sampling rate fsf_s and NN samples, the frequency-bin spacing is fs/Nf_s/N. Efficient implementations commonly use a fast Fourier transform algorithm. Converting the resulting coefficients into physical amplitudes requires the appropriate normalization and the instrument’s calibration.

Sampling imposes a fundamental restriction. Under the usual band-limited sampling assumptions, frequencies must lie below fs/2f_s/2 to be represented without ambiguity. Components above this limit can fold into lower frequencies through aliasing. A strong high-order harmonic may therefore appear falsely as a lower-frequency component unless an analog antialiasing filter sufficiently attenuates it before digitization. Increasing the sample count without increasing the sampling rate does not raise this upper frequency limit.

Finite records introduce another issue. If the acquisition interval contains an integer number of fundamental periods, an ideal stationary harmonic signal aligns with Fourier bins. Otherwise, truncation distributes a component’s apparent energy across multiple bins, an effect called spectral leakage. Window functions reduce some leakage effects but broaden spectral peaks and change amplitude scaling. Frequency resolution, leakage suppression, noise bandwidth, and amplitude accuracy consequently involve trade-offs.

A peak near an integer multiple is not automatically evidence of a true harmonic. A slowly drifting fundamental broadens or shifts peaks; amplitude modulation produces sidebands; unrelated sources can happen to lie close to harmonic frequencies. Order tracking in rotating machinery addresses this by relating vibration frequency to shaft rotation rate. A component at three times shaft speed is called the third order, but its interpretation still depends on whether shaft rotation is the relevant reference and on how speed varies during acquisition.

The measurement chain can also generate harmonics. Microphones, accelerometers, amplifiers, analog-to-digital converters, and optical detectors become nonlinear when driven outside their useful range. Clipping is an especially direct source of false harmonic content. Harmonics measured at the output of a system may therefore originate in the source, the device under test, or the measuring apparatus. Distinguishing these requires attention to dynamic range, background distortion, loading, and measurement uncertainty.

Harmonic measurements support several forms of analysis. Machinery monitoring uses changes in orders and sidebands to characterize vibration behavior; communications testing measures unwanted emissions and amplifier linearity; acoustic testing separates tonal components from broadband noise. In spectroscopy, harmonic detection can also mean demodulation at a multiple of a deliberately imposed modulation frequency. That use concerns the modulation signal and should not automatically be interpreted as generation of optical-frequency harmonics.

Historical development and related mathematical meanings

The study of harmonic relationships began with the numerical analysis of musical intervals and vibrating strings. Traditions associated with ancient Greek music theory connected simple length ratios with consonant intervals. Such accounts concern ratios and tuning rather than Fourier decomposition in its modern form. A physical explanation of complex musical tones required later developments in mechanics, experimental acoustics, and mathematical analysis.

During the seventeenth century, Marin Mersenne investigated vibrating strings and established quantitative relationships among frequency, length, tension, and mass per unit length. Joseph Sauveur, working around the beginning of the eighteenth century, advanced the experimental study of partial tones and acoustic frequency. These investigations helped distinguish a sound’s perceived musical identity from the multiple vibrations present in its physical motion.

The eighteenth-century vibrating-string problem brought the mathematical issues into focus. Jean le Rond d’Alembert developed a wave-equation treatment in 1747; Leonhard Euler and Daniel Bernoulli contributed different approaches to the representation of string motion. Bernoulli emphasized combinations of normal modes. Their disagreements included questions about which initial shapes could be represented by trigonometric sums—a question requiring a clearer understanding of functions and convergence than was then available.

Joseph Fourier’s investigations of heat conduction made trigonometric expansions a general analytical method. His Théorie analytique de la chaleur, published in 1822, presented techniques whose significance extended well beyond heat. Harmonic analysis subsequently developed into a broad branch of mathematics, including Fourier transforms, orthogonal expansions, and analysis on groups. The mathematical concept is therefore wider than the study of audible tones.

Hermann von Helmholtz’s On the Sensations of Tone, first published in German in 1863, connected acoustic spectra, resonators, musical intervals, and hearing. Later electronic oscillators, filters, recording systems, and digital computation made harmonic synthesis and analysis routine. The same framework could then be applied to electrical waveforms, mechanical vibration, and electromagnetic signals, even though their physical generation mechanisms differed.

Several related mathematical terms require separate definitions. A harmonic function satisfies Laplace’s equation, ∇2u=0\nabla^2u=0, on its domain. The term does not mean that the function is simply a time sinusoid or an integer-frequency component. Spherical harmonics are angular functions used to expand fields on a sphere, with applications to gravitational and magnetic potentials and quantum angular momentum. Their labels specify angular structure rather than a series of temporal frequencies.

Likewise, the quantum harmonic oscillator is defined by a quadratic potential and has equally spaced energy levels,

En=ℏω(n+12),n=0,1,2,….E_n=\hbar\omega\left(n+\frac12\right), \qquad n=0,1,2,\ldots.

Its name derives from the corresponding classical harmonic oscillator. Equal energy spacing is not a claim that every transition emits a different integer multiple of one frequency: in the ideal oscillator, adjacent levels differ by the same energy ℏω\hbar\omega, and the transitions actually observed depend on the interaction and its selection rules.