Total harmonic distortion (THD) is a measure of the extent to which a sinusoidal waveform is distorted by the presence of harmonic frequencies, which are integer multiples of the fundamental frequency, and is defined as the ratio of the root-mean-square (RMS) value of the harmonic components to the RMS value of the fundamental component, typically expressed as a percentage.[1] The standard formula for THD in voltage or current signals is:\text{THD} = \frac{\sqrt{\sum_{h=2}^{\infty} V_h^2}}{V_1} \times 100\%where V_1 is the RMS amplitude of the fundamental component and V_h is the RMS amplitude of the h-th harmonic.[1] This metric arises from nonlinearities in electronic devices or systems, such as amplifiers, power supplies, or inverters, that generate these unwanted harmonics.[2]In audio engineering, THD quantifies signal fidelity in devices like amplifiers and speakers, where low values—often below 0.1%—are essential for accurate reproduction of sound without audible coloration or degradation.[3] Related measurements, such as THD plus noise (THD+N), incorporate broadband noise alongside harmonics to provide a more comprehensive assessment of overall distortion in audio systems.[3] For example, high-end audio analyzers achieve THD+N levels as low as 0.01% or -120 dB, highlighting the precision required in professional applications.[3]In electrical power systems, THD evaluates the quality of AC waveforms, with excessive levels leading to adverse effects including overheating in transformers and motors, increased neutral conductor currents, and electromagnetic interference with communication lines.[2] The IEEE Std 519-2022 standard recommends limiting voltage THD to 5% or less at the point of common coupling to mitigate these issues, considering harmonics up to the 50th order.[4] Mitigation strategies include passive or active filters and selecting low-distortion power sources, such as those with inherent THD under 0.5%.[2] Across both domains, minimizing THD enhances system efficiency, reliability, and performance.[1]
Fundamentals
Definition
Total harmonic distortion (THD) is a measure of the harmonic distortion present in a signal, defined as the ratio of the root mean square (RMS) value of the harmonic content—excluding the fundamental frequency—to the RMS value of the fundamental frequency component, typically expressed as a percentage.[5] This is the definition relative to the fundamental component (often denoted THD_F); an alternative, less common in audio but used in some power analyses, expresses it relative to the total RMS value (THD_R). This quantification assesses how much the signal deviates from a pure sinusoidal waveform due to unwanted harmonic components.[1]The concept of total harmonic distortion, measuring the combined effect of harmonics relative to the fundamental, originated in early 20th-century electrical engineering, with discussions of harmonic distortion appearing around the 1930s in the context of audio amplification and waveform analysis to evaluate distortion in electronic circuits.[6][7] Harmonics in this context refer to sinusoidal components at integer multiples of the fundamental frequency, such as the second harmonic at twice the fundamental or the third at three times, which arise from nonlinearities in systems where the output is not directly proportional to the input.[1] These nonlinearities, often introduced by devices like power electronic converters or amplifiers, cause the generation of these higher-order frequencies that distort the original signal.[8] In electrical systems, such harmonics can propagate through voltage or current waveforms, leading to inefficiencies or interference.[9]THD is commonly applied to both voltage and current signals, with voltage THD (VTHD) evaluating distortions in the supply waveform and current THD (ITHD) assessing those in load currents, though interpretations in power systems often extend to power-based effects like reduced power factor or increased heating due to these distortions.[10] Unlike direct voltage or current measures, power-based views of THD incorporate its influence on overall system efficiency, such as through the distortion factor in power factor calculations.[11]
Harmonic Distortion Basics
Harmonic distortion arises in electronic systems when nonlinear transfer functions in components such as amplifiers, transducers, or power devices process input signals, generating unwanted frequency components that are integer multiples of the fundamental frequency.[12] These nonlinearities, often modeled as polynomial expansions like v_{\text{out}} = a_1 v_{\text{in}} + a_2 v_{\text{in}}^2 + a_3 v_{\text{in}}^3 + \cdots, introduce higher-order terms that distort the output waveform when the input amplitude is sufficiently large.[12]Harmonics generated by these nonlinearities are classified as even or odd based on their order relative to the fundamental. Even harmonics (e.g., second, fourth) typically result from asymmetric nonlinearities, such as quadratic terms, and can produce a waveform that appears more rectified or folded. In contrast, odd harmonics (e.g., third, fifth) stem from symmetric nonlinearities like cubic terms and often dominate in scenarios like amplifier clipping, where symmetrical peak limiting flattens the waveform toward a square-like shape, emphasizing odd-order components.[12][13]The analysis of harmonic distortion relies on the Fourier series decomposition, which represents any periodic signal as a sum of its fundamental frequency component and higher harmonic components at integer multiples of that frequency.[14] This decomposition, applicable to signals satisfying Dirichlet conditions, separates the original sinusoidal input from the spurious harmonics introduced by nonlinearity, enabling the identification of distortion effects. THD serves as a key metric to quantify the total contribution of these harmonic components relative to the fundamental.[14]The presence of harmonics degrades signal integrity by extending the required transmission bandwidth to accommodate higher-frequency components without attenuation or interference.[12] Additionally, when multiple input tones are present, harmonics can mix to produce intermodulation products, further complicating the spectrum and reducing the overall signal-to-noise ratio.[12]
Mathematical Formulation
THD Formulas
Total harmonic distortion (THD) is quantified through standard mathematical expressions derived from the Fourier series decomposition of periodic signals. The primary formula, often denoted as THD_F, expresses THD relative to the fundamental component and is defined for voltage as\text{THD}_F = \frac{\sqrt{\sum_{h=2}^{\infty} V_h^2}}{V_1},where V_h represents the root-mean-square (RMS) value of the h-th harmonic voltage, and V_1 is the RMS value of the fundamental component.[15] This form is widely used in power systems analysis, as specified in IEEE standards, to assess the contribution of higher harmonics to distortion relative to the desired fundamental frequency.[16]An alternative formulation, denoted as THD_R, normalizes the harmonic content relative to the total RMS value of the signal:\text{THD}_R = \frac{\sqrt{\sum_{h=2}^{\infty} V_h^2}}{\sqrt{\sum_{h=1}^{\infty} V_h^2}}.This version, common in certain international standards like those from IEC, provides a measure bounded between 0 and 1 (or 0% to 100%), as the denominator includes the full signal power.[16] The two forms are interrelated by \text{THD}_R^2 = \frac{\text{THD}_F^2}{1 + \text{THD}_F^2}, highlighting how THD_F can exceed 100% when harmonics dominate, unlike THD_R.[16]These formulas arise from Fourier analysis, which decomposes a periodic waveform v(t) into a sum of sinusoids: v(t) = V_0 + \sum_{h=1}^{\infty} V_h \cos(h \omega t + \phi_h), where the RMS values V_h are obtained from the Fourier coefficients.[15] By Parseval's theorem, the total signal power is the sum of the powers of its frequency components, so the harmonic power relative to the fundamental power is P_h / P_1 = \sum_{h=2}^{\infty} V_h^2 / V_1^2, yielding THD_F as the square root of this ratio, since distortion is often expressed in terms of effective voltage or current amplitudes.[15]THD is typically reported as a percentage by multiplying the ratio by 100, facilitating comparison across systems.[15] In linear electrical systems, voltage THD and current THD are equivalent when resistances are constant, as current harmonics scale proportionally with voltage harmonics via Ohm's law, though nonlinear loads can introduce differences.[15]The formulations assume strictly periodic signals amenable to Fourier series expansion; aperiodic or transient signals require modifications like windowed analysis.[15] In practice, the infinite series is approximated by a finite number of harmonics (e.g., up to the 50th order), as higher-order components are typically negligible due to attenuation in physical systems.[16]
THD+N
Total harmonic distortion plus noise (THD+N) is a performance metric that quantifies the combined contributions of harmonic distortion and noise in a signal, relative to the fundamental component, typically measured over a defined bandwidth to assess overall signal fidelity in practical systems.[17] This extension of traditional THD accounts for non-harmonic components like broadband noise, which are prevalent in real-world audio and analog circuits.[3]The formula for THD+N is given by:\text{THD+N} = \frac{\sqrt{\sum_{i=2}^{n} V_i^2 + V_N^2}}{V_1}where V_1 is the RMS voltage of the fundamental signal, V_i are the RMS voltages of the harmonic components (from the second to the nth), and V_N is the RMS voltage of the noise, often expressed as a percentage or in decibels.[17] In audio applications, THD+N measurements frequently incorporate A-weighting to emphasize frequencies within the human hearing range (approximately 20 Hz to 20 kHz), filtering out inaudible low-frequency noise like 60 Hz hum while focusing on perceptually relevant distortion.[18]Noise is included in THD+N because pure harmonic distortion measurements can underestimate total impairment in low-signal or high-fidelity environments, where noise from sources like thermal effects, interference, or quantization in digital systems may dominate over harmonics, especially when the signal-to-noise ratio falls below 50 dB.[19] This makes THD+N a more realistic indicator for audio equipment specifications, as it captures the full residual after notching out the fundamental, revealing system limitations under typical operating conditions.[3]Bandwidth selection is critical for THD+N, with audio standards commonly specifying 20 Hz to 20 kHz to align with audible frequencies; narrower or wider bands alter noise inclusion, potentially inflating or deflating results by changing the effective noise power captured post-filtering.[17] For instance, using a 22 kHz bandwidth in FFT-based analysis can emphasize noise in high-resolution converters.[19] Unlike pure THD, which excludes noise and thus yields lower values in quiet systems, THD+N will always be higher or equal, providing a conservative assessment suited to noisy real-world scenarios.[3] Standards such as IEC 60268-3 for audio amplifiers define related distortion factors and endorse bandwidth-limited measurements to ensure comparable results across devices.[19]
Measurement and Analysis
Techniques
Hardware methods for measuring total harmonic distortion (THD) primarily involve specialized instruments that isolate the fundamental frequency and quantify harmonic components. Spectrum analyzers, such as those from Rohde & Schwarz, capture the frequency spectrum of the input signal, allowing direct identification and measurement of harmonic amplitudes relative to the fundamental.[20] Dedicated THD analyzers, like the Keysight E5071C with frequency-offset capabilities, automate the process by performing receiver calibration and harmonic detection through built-in filtering.[21]Notch filters represent another hardware approach, where a tunable filter attenuates the fundamental frequency to isolate harmonics for subsequent amplitude measurement, often implemented in analog distortion meters.[22]Software techniques leverage digital signal processing for THD computation, particularly through Fast Fourier Transform (FFT) analysis. In tools like MATLAB, the thd function decomposes the signal into its frequency components, calculates the RMS values of the fundamental and harmonics, and applies the standard THD formula to derive the distortion ratio.[23] Similar FFT-based methods are available in audio processing software, such as those integrated with oscilloscopes or virtual instruments, enabling post-capture analysis of recorded waveforms. These approaches are particularly useful for field measurements where hardware portability is essential.The step-by-step procedure for THD measurement begins with generating or capturing a clean sinusoidal input signal at the desired fundamental frequency. The signal is then passed through the device under test, and the output is analyzed to separate the fundamental from harmonics via filtering or spectral decomposition. RMS values are computed for the fundamental and the sum of harmonics, followed by calculating the THD as the ratio of these quantities, typically expressed as a percentage.[22] This process ensures accurate quantification in both laboratory and field environments.Calibration of THD measurement instruments follows established standards to maintain precision and traceability. Instruments compliant with IEEE procedures undergo testing with reference waveforms to verify voltage and current THD functions, using calibrated sources to ensure high accuracy.[24] Similarly, IEC 61000-4-7 provides guidelines for harmonic instrumentation, specifying measurement bandwidths and anti-aliasing requirements for consistent results across devices. For non-periodic signals encountered in real-world scenarios, windowing techniques—such as Hanning or rectangular windows—are applied during FFT processing to minimize spectral leakage and approximate periodicity.[25][26]Challenges in THD measurement include aliasing in digital sampling, where high-frequency harmonics fold back into lower frequencies if the sampling rate is insufficient, potentially inflating distortion estimates.[22] In power systems, impedance matching between the measurement instrument and the source or load is critical to avoid reflections or loading effects that alter harmonic content, requiring careful selection of probes and interfaces with matched impedances.[27]
Interpretation
Interpreting THD measurements requires contextualizing the results against established thresholds and considering influencing factors to ensure accurate assessment of system performance. In high-fidelity audio applications, THD levels below 1% are generally deemed acceptable, as higher values can introduce perceptible nonlinearities in sound reproduction.[28] For electrical power systems, the IEEE 519-2022 standard recommends voltage THD limits below 5% at the point of common coupling for systems rated up to 69 kV, aiming to prevent interference with connected equipment and maintain grid stability.[29] These thresholds provide benchmarks for compliance, but exceeding them does not always indicate failure; rather, it signals the need for mitigation strategies tailored to the application.Several factors influence the interpretation of THD values, necessitating a holistic view beyond raw percentages. The crest factor, defined as the ratio of peak to RMS value, can indicate waveform peaking due to harmonics; deviations from the ideal 1.414 for a sine wave suggest distortion levels that may stress equipment, and high crest factors (>3) often correlate with elevated THD under nonlinear conditions.[30] Load conditions play a critical role, as nonlinear loads like variable frequency drives increase current harmonics, potentially elevating THD during peak demand while remaining low under linear loads.[31]Frequency dependence further complicates analysis, since higher-frequency harmonics attenuate differently across system impedances, leading to THD variations that are more pronounced at low fundamental frequencies where measurement bandwidth captures more artifacts.[32]Error sources in THD measurements must be identified to distinguish genuine distortion from artifacts. Instrumentation limits, such as finite bandwidth or inadequate signal-to-noise ratio, can introduce floor noise that inflates THD readings, particularly at low signal levels.[33] Environmental noise, including electromagnetic interference or ambient vibrations, adds extraneous components that mimic harmonics; true distortion is verified by repeating measurements under controlled conditions, applying notch filters to isolate the fundamental, and confirming consistency across multiple trials to rule out transient artifacts.[34]Comparative metrics highlight THD's strengths and limitations in capturing overall signal integrity. THD quantifies only harmonic components relative to the fundamental, potentially understating total degradation when noise dominates, whereas SINAD (signal-to-noise and distortion ratio) encompasses both, providing a more comprehensive view for noisy environments like audio chains.[35] Similarly, intermodulationdistortion (IMD) assesses non-harmonic products from multiple tones, revealing issues THD misses, such as high-frequency nonlinearities; THD may overstate distortion in single-tone tests by ignoring IMD's perceptual harshness in complex signals.[36]The IEEE 519-2022 revisions refined harmonic limits and clarified application at points of common coupling, but no major updates to THD standards have occurred post-2024 as of 2025, maintaining focus on steady-state conditions for design guidance.[37]
In audio reproduction, total harmonic distortion (THD) introduces audible artifacts that degrade perceived sound quality, such as increased harshness and roughness in the midrange and treble frequencies, particularly when odd-order harmonics dominate. For instance, THD levels above 1% can produce noticeable intermodulation between harmonics and the fundamental signal, leading to a "fuzzy" or colored timbre that listeners often describe as fatiguing during extended playback. Psychoacoustic masking plays a key role here, where low-level harmonics (below approximately 0.1% THD, or -60 dB relative to the fundamental) are often inaudible due to the ear's reduced sensitivity to higher frequencies and simultaneous masking by the primary signal, allowing subtle distortions to go undetected in complex music.[38][39][40]Amplifier specifications typically rate THD at specific power levels and frequencies to ensure fidelity, with high-quality audio amplifiers targeting less than 0.1% THD at 1 watt output into 8 ohms, often measured at 1 kHz to reflect common listening conditions. This low THD preserves the dynamic range by minimizing noise floor elevation from distortion products, enabling signal-to-noise ratios exceeding 90 dB without audible degradation in quiet passages. Exceeding these specs, such as THD rising to 1% at higher powers, can compress effective dynamic range by introducing spurious energy that masks low-level details, effectively reducing the usable span between noise and peak signals.[35][39]To mitigate THD, negative feedback loops in amplifier designs linearize the transfer function, suppressing distortion by factors of 20-40 dB through error correction, though excessive feedback can introduce phase issues if not managed. Class-D switching amplifiers further reduce THD to below 0.01% by employing pulse-width modulation and high-frequency filtering, achieving efficiencies over 90% while maintaining low distortion across the audio band, as demonstrated in integrated circuits delivering up to 5 W with 1% THD+N (with 0.01% THD+N at lower power levels).[41][42][43][44]In radio communication systems, THD from nonlinear amplification generates harmonic sidebands that cause spectral regrowth, broadening the transmitted signal and increasing adjacent-channel interference, which can desensitize receivers or violate spectrum allocations. This distortion exacerbates out-of-band emissions, prompting regulatory controls such as FCC limits on spurious and harmonic emissions, requiring suppression to at least 43 + 10 log10(P) dB below the carrier (or 80 dB, whichever is less), where P is the transmitter power in watts, for FM broadcast transmitters to prevent co-channel interference.[8][45]Digital audio codecs like MP3 introduce nonlinear distortion through perceptual quantization and bit allocation, generating harmonic artifacts that alter timbre, especially in transient-rich signals such as percussion, where low-bitrate encoding (below 128 kbps) can produce audible "smearing" from inter-harmonic interactions. Unlike analog THD, these distortions are masked by the codec's psychoacoustic model but become perceptible at high compression levels, reducing overall fidelity despite transparent SNR metrics.[46]
Electrical Power Systems
In electrical power systems, total harmonic distortion (THD) arises primarily from nonlinear loads that draw non-sinusoidal currents, distorting the voltage waveform at the point of common coupling (PCC). Common sources include power electronic devices such as inverters in solar photovoltaic (PV) systems, light-emitting diode (LED) drivers, and electric vehicle (EV) chargers, which introduce harmonics due to their switching operations. With the increasing integration of renewable energy sources like solar inverters and widespread EV adoption by 2025, these nonlinear loads have amplified harmonic injection into distribution networks, potentially exceeding traditional levels in low-voltage grids.[47][48][49]The consequences of elevated THD in AC power distribution are significant, leading to overheating in transformers and motors from increased eddy current and skin effect losses, which can reduce equipment lifespan by up to 50% under sustained high distortion. Neutral conductors in three-phase systems experience overload due to triplen harmonics (multiples of the third harmonic) adding constructively in the neutral path, potentially causing conductor failures. Additionally, voltage regulation becomes challenging as harmonics propagate through the impedance of lines and transformers, resulting in distorted supply voltages that affect sensitive loads and overall system stability.[50][51][52][53]To address these issues, standards such as IEEE 519-2022 establish limits on harmonic distortion at the PCC, recommending voltage THD below 8% for systems with nominal voltages up to 1 kV and 5% for voltages between 1 kV and 69 kV to ensure compatibility with utility grids. The Information Technology Industry Council (ITIC) curve further defines equipment tolerance envelopes, specifying steady-state voltage deviations within ±10% of nominal and prohibiting sustained excursions that could trigger malfunctions in IT and industrial loads. Compliance with these guidelines helps mitigate risks from renewable integration and EV charging surges.[54][55][56][57]Mitigation strategies focus on filtering and monitoring to maintain power quality. Active harmonic filters dynamically inject counter-harmonic currents to cancel distortions from sources like solar inverters and EV chargers, achieving up to 95% reduction in THD while compensating for reactive power. Passive harmonic traps, or tuned filters, target specific frequencies such as the fifth and seventh harmonics common in LED drivers, providing cost-effective suppression in distribution feeders. Power quality analyzers enable real-time monitoring and assessment, allowing operators to detect THD exceedances and implement corrective actions in renewable-heavy grids.[58][59][60]
Examples
Waveform Calculations
To compute the total harmonic distortion (THD) for idealized periodic waveforms, the process begins with determining the Fourier series expansion to identify the amplitudes of the fundamental and harmonic components. The RMS value of each sinusoidal component is then calculated as the peak amplitude divided by \sqrt{2}, though for THD ratios, the scaling cancels out. The THD, specifically the distortion factor THD_F, is applied as \mathrm{THD}_F = \sqrt{\sum_{h=2}^{\infty} \left( \frac{A_h}{A_1} \right)^2} \times 100\%, where A_h and A_1 are the peak amplitudes of the h-th harmonic and fundamental, respectively. This involves summing the squared amplitude ratios for all harmonics beyond the fundamental, often using closed-form series sums for verification.[61]For a square wave, normalized to peak amplitude 1 (oscillating between -1 and 1), the Fourier series consists of odd harmonics only: s(t) = \frac{4}{\pi} \sum_{k=1,3,5,\ldots}^{\infty} \frac{1}{k} \sin(k \omega t). The fundamental amplitude is A_1 = 4/\pi, and higher odd harmonics have amplitudes A_k = 4/(k \pi) for k = 3,5,\ldots, yielding ratios A_k / A_1 = 1/k. The sum of squared ratios for harmonics is \sum_{k=2}^{\infty} 1/(2k-1)^2 = \pi^2/8 - 1, leading to \mathrm{THD}_F = \sqrt{\pi^2/8 - 1} \approx 48.3\%. This exact closed-form value confirms the infinite series summation.[61]A sawtooth wave, also normalized to peak-to-peak amplitude 2 (ranging from -1 to 1), features all integer harmonics: s(t) = \frac{2}{\pi} \sum_{n=1}^{\infty} \frac{(-1)^{n+1}}{n} \sin(n \omega t). The fundamental amplitude is A_1 = 2/\pi, with higher harmonics A_n = 2/(n \pi) (magnitudes), so ratios A_n / A_1 = 1/n. The squared ratio sum is \sum_{n=2}^{\infty} 1/n^2 = \pi^2/6 - 1, resulting in \mathrm{THD}_F = \sqrt{\pi^2/6 - 1} \approx 80.3\%. The 1/n decay across both odd and even harmonics contributes to the higher distortion compared to odd-harmonic-only waveforms.[62]The triangle wave, normalized similarly to peak amplitude 1, includes only odd harmonics with quadratic decay: \mathrm{tr}(t) = \frac{8}{\pi^2} \sum_{k=1,3,5,\ldots}^{\infty} \frac{(-1)^{(k-1)/2}}{k^2} \sin(k \omega t). Here, A_1 = 8/\pi^2, and for odd k \geq 3, A_k = 8/(k^2 \pi^2), giving ratios A_k / A_1 = 1/k^2. The sum of squared ratios is \sum_{k=2}^{\infty} 1/(2k-1)^4 = \pi^4/96 - 1, yielding \mathrm{THD}_F = \sqrt{\pi^4/96 - 1} \approx 12.1\%. This lower value reflects the faster 1/n^2 amplitude decrease, concentrating energy in the fundamental.[61]
Real-World Cases
In electrical power systems, total harmonic distortion (THD) often arises from nonlinear loads such as variable frequency drives (VFDs) and electronic devices, leading to measurable impacts on power quality. A case study conducted at the School of Electrical System Engineering, Universiti Malaysia Perlis, monitored current THD (THDi) over 30 days using a Fluke 1750 analyzer at the point of common coupling in laboratory and office buildings. Nonlinear loads including computers, air conditioners, and fluorescent lamps were identified as primary sources, resulting in maximum THDi values exceeding 123% in one block, with over 58% of the time surpassing the IEEE 519 limit of 16%; similar exceedances occurred in other blocks, up to 91.44%, causing 12-20% increases in RMS current and potential energy losses.[63]On a broader scale, an analysis of the U.S. electric grid using data from over 1,000,000 Ting sensors revealed that the majority of monitored homes experienced voltage THD below the IEEE 519 limit of 8%, but hotspots in densely populated areas and certain utilities showed higher levels. For instance, in February to October 2024, 36.5% of homes served by Commonwealth Edison Co. had THD exceeding 8%, attributed to inverter-based resources like solar panels and electronic loads such as LED lighting; solar storms in May 2024 further spiked THD to 25% in affected regions by inducing geomagnetic currents, reducing equipment efficiency and risking motor overheating.[50]In an R&D facility case study, harmonics monitoring with a power quality analyzer showed voltage THD remaining below 5%—the IEEE 519-2014 limit for systems under 1 kV—while current THD was marginally compliant, driven by 5th and 7th harmonics from 6-pulse VFDs on 480 V systems. Line reactors (1-5%) mitigated much of the distortion without operational issues, though recommendations included upgrading reactors on larger VFDs to prevent future exceedances.[64]In audio applications, THD quantifies signal fidelity in devices like loudspeakers and amplifiers, where low values ensure minimal audible artifacts. A study on a piezoelectric microelectromechanical systems (MEMS) loudspeaker evaluated THD within an IEC 60318-4 coupler using static nonlinear measurements and a state-space model, achieving agreement between simulated and measured THD below 1% difference from 40 Hz to 200 Hz at input levels up to 1.41 V, though discrepancies rose above 5% near 6.4 kHz due to frequency response variations; second and third harmonics dominated at 1 kHz, supporting its use in compact audio systems with predistortion potential.[65]Real-world headphone measurements further illustrate low THD in consumer audio gear. For the Sennheiser IE 900 in-ear monitors, THD plots on a logarithmic frequency scale showed lower distortion than the IE 300 model across audible bands, typically below perceptible thresholds when tested with an acoustic coupler, emphasizing design improvements in driver linearity for high-fidelity playback without correlating strongly to subjective sound quality.[66]