phonometry Documentation #
Full documentation for phonometry. Also available as a website: https://docs.jmrp.io/phonometry/
Guides #
Every guide in the library, grouped by the folder it lives in. Each group heading links the overview page that says what the area is for and what it deliberately leaves out; each bold line links a section overview inside it.
Before any of them, Start is the six pages meant to be read once: Getting Started installs the library and runs a first calibrated analysis, Why phonometry shows how a metric is checked against its standard, Upgrading from 3.3 to 4.0 is the map for code written against an earlier release, Support the project lists the ways to help, starting with the sources the library still lacks, and About says who maintains it, how to report an error and how to cite it.
Signal analysis #
The chain that turns a digital signal into a standards-compliant number: bands, weightings, ballistics, levels, spectra, calibration and uncertainty. Every other section consumes it.
- Build a sound level meter: an end-to-end walk-through that composes the core API into a working meter: calibrate against an IEC 60942 tone, apply the IEC 61672-1 frequency and time weightings, integrate into $L_\mathrm{eq}$, SEL and percentile levels, split into IEC 61260-1 octave bands, and verify the class of every stage
- Filter Banks: band mathematics, bank parameters, parametric EQ, band decomposition and zero-phase filtering
- Filter architecture gallery: the five filter architectures compared at the band edges, the full response gallery for 1/1 and 1/3 octave, per-architecture usage and the Linkwitz-Riley crossover
- Filter class verification: the IEC 61260-1:2014 Table 1 acceptance mask band by band, the effective bandwidth and summation of outputs of IEC 61260-2:2016, the swept time-invariance test, the grading of IEC 61260-3:2016 periodic tests, the class 0 of the withdrawn IEC 61260:1995 / ANSI S1.11-2004 edition, what a performance class buys in a measurement, and the one-page compliance fiche
- Block Processing: stateful real-time workflows
- Multichannel: vectorized multichannel analysis
- Integrated & Statistical Levels: $L_\mathrm{eq}$, $L_\mathrm{Aeq}$, $L_{10}$/$L_{50}$/$L_{90}$, $L_\mathrm{Cpeak}$/SEL, noise dose (IEC 61252), octave spectrogram
- Frequency Weighting: A, C, Z curves
- Special weightings: the ISO 7196 G-weighting for infrasound, the historical B (ANSI S1.4-1983) and D (IEC 537) curves, and AU (IEC 61012) for audible sound in the presence of ultrasound
- Time Weighting: Fast, Slow, Impulse ballistics
- Calibrated spectral analysis: the Bendat & Piersol Welch estimators with their statistical quality: PSD and cross-spectral density with the effective number of averages, normalized random errors and chi-square confidence intervals, the coherent output spectrum with the spectral SNR, constant-power 1/n-octave smoothing, colored-noise generators with an exact power-law slope, and the Harris window figures of merit for choosing the taper
- Time-frequency analysis: the calibrated STFT spectrogram in absolute dB SPL with the exact Welch-module scaling and the time-versus-frequency resolution trade-off, and the zoom FFT that resolves tones closer than a practical FFT bin
- Multiple and partial coherence: the Bendat & Piersol multiple-input/output coherence functions for several correlated sources and one output, with the Gaussian-elimination conditioning that separates a genuine cause from a source that merely correlates with it, and the partial coherent output spectra that say which source dominates each band
- Correlation, time delay and envelope: auto- and cross-correlation with the Bendat & Piersol normalizations and random errors, time-delay estimation by the direct correlator, the cross-spectrum phase slope and the Knapp & Carter GCC (Roth, SCOT, PHAT, maximum likelihood) with the Eq. 8.129 peak-location uncertainty, sub-sample impulse-response delay and alignment, and the Hilbert envelope with instantaneous phase and frequency
- Test signals and sample-rate tools: IEC 60268-1 tone bursts with exact gating (zero-crossing start, integral full periods, repetitive trains), polyphase resampling behind an explicit anti-alias specification whose designed Kaiser filter travels with the result, and band-limited fractional delay with a linear or circular boundary
- Cepstrum, echoes and the envelope spectrum: the power, real and complex cepstrum with quefrency analysis, echo detection with the reflection coefficient read off the cepstral peak, lowpass/highpass liftering of a log spectrum, the homomorphic round trip of the complex cepstrum, and the envelope spectrum that turns amplitude modulations into discrete lines
- Time synchronous averaging: extraction of a periodic waveform of known period by time domain averaging (McFadden 1987), the comb filter that describes the operation in the frequency domain with unit teeth at the harmonics and nodes between them, the square-root noise-reduction law, and the choice of the number of averages that places a comb node on an interfering order
- System measurement: Golay, shaped sweeps, inversion: complementary Golay pairs whose periodic autocorrelations sum to an exact delta and deconvolve a noiseless system to machine precision, sweeps synthesized to follow an arbitrary target magnitude spectrum by group-delay shaping (Mueller & Massarani) with a near-ideal crest factor, and the regularized spectral inversion of a measured response with Kirkeby frequency-dependent regularization, achieved flatness and a capped out-of-band gain
- Calibration and dBFS: physical SPL and digital analysis
- Compliance and verification: what a performance class asserts in IEC 61672-1 and IEC 61260-1, the public verifiers that grade weightings, filter banks and intensity spectra against their tolerance tables, how to read the numerical conformance report, and the verified scope of the pattern-evaluation and periodic-test parts (IEC 61672-2/-3, IEC 61260-2/-3) that only a laboratory can run
- Measurement uncertainty: the GUM law of propagation of uncertainty and the Monte Carlo method (ISO/IEC Guide 98-3:2008 and Supplement 1): combined and expanded uncertainty, Welch–Satterthwaite effective degrees of freedom, and probabilistically symmetric coverage intervals
- Data qualification: the Bendat & Piersol stationarity tests (reverse arrangements with the Table A.6 acceptance regions, runs about the median with the exact Wald-Wolfowitz distribution) on segment mean squares, and the Rice statistics of level crossings, apparent frequency, peak rates and the irregularity factor that places the peak-height distribution between Rayleigh and Gaussian
- Random-incidence and diffuse-field response (IEC 61183): the directivity factor of a sound level meter from readings in two planes, one plane or 38 equal-area elements with the Table A.1 adjustment factors, its random-incidence sensitivity level, and its diffuse-field sensitivity level by comparison with a reference whose Table B.1 characteristics are built in
- Free-field corrections of a sound level meter (IEC 62585): the adjustment value at the calibration check frequency, the free-field corrections for a multi-frequency calibrator, a comparison coupler and an electrostatic actuator, the uncertainty budget of Annex I with its Welch-Satterthwaite coverage factor, and the maxima of clauses 9 to 14 as a verdict and a fiche
- Periodic tests of a sound level meter (IEC 61672-3): the verdict on a laboratory's periodic-test results, clause by clause against the acceptance limits of IEC 61672-1 and the maxima of its Table B.1, what a complete test holds, and the statement of Clause 22.
- Microphone calibration by comparison (IEC 61094-5 and IEC 61094-8): a working standard microphone calibrated against a reference in a coupler (simultaneous, with the interchange of Annex C) or in a free field (by substitution against a monitor microphone), the phase of the sensitivity, the environmental, WS3 jig, different-diameter and acoustic impedance corrections with the air between two microphones, the validation of a jig or coupler, the uncertainty budgets of IEC 61094-5 Table D.1 and IEC 61094-8 Table 2 with k = 2, and the time-selective processing of IEC 61094-8 Annex B (the window, the stepped sine and the pulse of the direct impulse method)
- Microphone calibration by reciprocity (IEC 61094-2 and IEC 61094-3): laboratory standard microphones calibrated in pairs with no reference microphone, in a coupler (the acoustic transfer impedance of a plane-wave or large-volume coupler with the heat conduction of Annex A, Gerber's full solution reproducing Table A.1, and the capillary tubes of Annex B) or in a free field (the attenuation of Annex B, the dispersion of the speed of sound and the acoustic centres), the triad and the two-microphone method, the coupler and arrangement checks, and the uncertainty budgets of Table 1 with the one-at-a-time recalculation
The medium #
The fluid the sound travels through, computed from the conditions that were measured rather than assumed. A density and a speed of sound stand behind every level in the library; this is where they come from instead of being typed once.
- Humid air: the density, speed of sound, ratio of specific heats, viscosity and thermal diffusivity of air from the measured temperature, pressure and humidity (IEC 61094-2:2009 Annex F, CIPM-2007), what each condition is worth, the domain the annex states for itself, and what the model refuses to guess
Elastic solids #
The elastic constants of a solid and the three longitudinal wave speeds that follow from them, named after the shape the wave travels in. A plate table and a solver table print different numbers for the same steel; the conversion between them lives here.
- Wave speeds: the beam, plate and unbounded longitudinal speeds of a homogeneous isotropic solid, the inverse that turns each back into a Young's modulus, and the thickness-critical-frequency product that lets two books be checked against each other
Files #
Measurement audio in and out: every linear WAV a meter writes read into a
calibrated Signal with its bext provenance, long recordings streamed
block by block, BWF written with provenance and measured loudness, the
calibration sidecar, and lossless conversion. A catalogue of materials of
your own is read from a JSON file, or from a spreadsheet saved as CSV, into
the rows every published catalogue hands out, and written back. Any
matplotlib animation, and an FDTD run as it is computed, is written to a
WebM, an MP4 or a GIF through ffmpeg.
- Reading and writing measurement audio: the whole workflow on one runnable page, from the meter's WAV to the calibrated level, the lossy warning, streaming, BWF writing, the sidecar and lossless conversion
- Your own catalogues: a data sheet, a test report or a measurement of yours as a catalogue file, read into the rows every published catalogue hands out, searched with the same lookups, handed to the same models and completed in your editor with its JSON Schema
- Writing animations and FDTD runs to video: any matplotlib animation to WebM, MP4 or GIF through ffmpeg, the encoder of a WebM or MP4 found by a real trial encode, and an acoustic or elastic FDTD run written straight to a video without storing its snapshots
Hearing and perception #
What a listener makes of the sound a level merely counts: loudness, sharpness, roughness and annoyance, speech intelligibility, and what a working life in noise costs a hearing threshold.
- Loudness: the ISO 532-1 Zwicker loudness in sones with its one-page fiche, plus the equal-loudness contours (ISO 226)
- Advanced loudness: the Moore-Glasberg stationary (ISO 532-2) and time-varying (ISO 532-3) methods and the Sottek Hearing Model loudness (ECMA-418-2), with the model-choice table
- Sound Quality Metrics: sharpness (DIN 45692) and the ECMA-418-2 Sottek Hearing Model tonality, roughness and fluctuation strength
- Prominent discrete tones: the ECMA-418-1 tone-to-noise and prominence ratios that decide whether a discrete tone is prominent and justify tonal rating adjustments
- Tonal audibility of tones in noise: the ISO/PAS 20065 engineering method for the audibility $\Delta L$ of a tone above the masking threshold: the critical band about the tone, the critical-band masking level, the masking index, and the decisive and mean audibility
- Psychoacoustic annoyance & fluctuation strength: the Fastl & Zwicker annoyance $PA = N_5\left(1 + \sqrt{w_S^2 + w_{FR}^2}\right)$ from loudness, sharpness, roughness and fluctuation strength (Eqs 16.2–16.4), the closed form for AM broadband noise (Eq. 10.2) and the Osses 2016 fluctuation-strength signal model
- Speech Transmission Index: how much of the speech envelope a room or sound system preserves: the IEC 60268-16 modulation transfer function, indirect method and direct STIPA measurement
- Speech Intelligibility Index: the ANSI S3.5-1997 SII in all four band procedures (critical band, equally contributing, one-third octave, octave): band-importance weighting, self-speech and upward spread of masking, band audibility, and the index in noise and hearing loss
- Objective intelligibility (STOI & ESTOI): the correlation-based intelligibility measures for time-frequency weighted noisy speech from a clean/degraded pair: STOI (Taal et al. 2011), the clipped per-band envelope correlation, and ESTOI (Jensen & Taal 2016), the row- and column-normalised spectral correlation that tracks modulated maskers
- Hearing threshold: the age-related hearing threshold distribution (ISO 7029:2017) and the free-field/diffuse-field reference threshold of hearing (ISO 389-7:2005)
- Audiometric test methods: whether an audiometric test room is quiet enough (ISO 8253-1 Clause 11) and its sound field free, quasi-free or diffuse (ISO 8253-2), the ascending, bracketing, automatic and sweep-frequency rules that turn responses into a hearing threshold level, and its Annex A uncertainty
- Noise-induced hearing loss: the ISO 1999:2013 noise-induced permanent threshold shift (NIPTS) and its population distribution, and the combination with age into the hearing threshold level associated with age and noise (HTLAN)
- Occupational noise exposure: the ISO 9612 task-based, job-based and full-day measurement strategies and the Annex C uncertainty budget behind every $L_\mathrm{EX,8h}$ report
- Hearing protectors: the ISO 4869-1 attenuation on sixteen subjects with its uncertainty and significance test, the ISO 4869-2 octave-band, HML and SNR methods that turn it into the effective A-weighted level left at the ear, the ISO 4869-6 total attenuation of an active noise reduction earmuff, and the ISO 4869-3 insertion loss that screens an earmuff on an acoustic test fixture
Rooms and buildings #
Sound in the built environment, split along its natural line: what happens inside one room, and what passes between rooms: measured, rated and predicted.
- Room Acoustics: impulse-response acquisition (ISO 18233), reverberation and room parameters (ISO 3382-1/2), open-plan speech metrics (ISO 3382-3), reverberation-room sound absorption (ISO 354)
- Sound strength G (ISO 3382-1): how loud the hall is, against the same source at 10 m in a free field, and the three printed routes to that reference
- Spatial impression (ISO 3382-1): the early lateral energy fraction, the late lateral sound level and the interaural cross correlation, measured with a second microphone
- Stage support and uncertainty (ISO 3382-1): what a player hears of their own instrument on the platform, and what the randomness of the excitation costs a reverberation time
- Reporting and source qualification (ISO 3382-1): the printed tables, from the directivity a source must clear to the fifteen items a test report shall carry
- Room impulse response acquisition (ISO 18233): the exponential sine sweep and its deconvolution, the MLS correlation method, and the source and microphone placement rules of a valid measurement
- Spatial sound decay in workrooms (ISO 14257, ISO 11690-3): the spatial sound distribution curve of a workroom, the rate of spatial decay $\mathrm{DL}2$ and the excess $\mathrm{DL}\mathrm{f}$ over a free field, the Annex B correction for the source's own curve, and the prediction categories and detail levels that match them at design time
- Open-plan office acoustics (ISO 3382-3): the spatial decay rate $D_{2,\mathrm{S}}$ of A-weighted speech, the level at 4 m, and the distraction and privacy distances derived from STI
- Sound absorption in enclosed spaces: the EN 12354-6:2003 prediction of a room's total equivalent absorption area and reverberation time from its surfaces and objects (Clause 4)
- Reverberation-time prediction: the reverberation time from a room's volume and surface absorption by five statistical models (Sabine, Eyring, Millington-Sette, Fitzroy, Arau-Puchades), with the air-absorption term
- Image sources and the steady-state room field: the deterministic image-source room impulse response of a rectangular room (Kuttruff/Vorländer) and the statistical steady-state level with the room constant, critical distance and Schroeder frequency (Bies)
- Room-noise criteria: the ANSI/ASA S12.2-2019 room-noise ratings: the NC rating (NC-(SIL) designation with the tangency method when exceeded, Table 1) and the RC Mark II rating with its rumble/hiss/neutral spectral tag (Annex D)
- Field Insulation Measurement (ISO 16283): field airborne and impact insulation (ISO 16283-1/2), the Clause 14 test report and the measurement uncertainty that qualifies it (ISO 12999-1)
- Small Rooms: the ISO 16283 Low-Frequency Procedure: the corner procedure ISO 16283-1/-2/-3 make mandatory below 25 m³: the 25 m³ trigger with its rounding, the corner level of Formula (12), the Formula (13) combination and the 63 Hz octave reverberation time of Clause 10.4
- Laboratory Insulation Measurement: the ISO 10140 laboratory sound reduction index and normalized impact level, measured with flanking suppressed, with the background-noise correction and the accredited test fiches
- Linings, Floor Coverings, Joints and Rain (ISO 10140-1): the four application rules of ISO 10140-1:2021 that report an improvement or a new quantity: the improvement ΔR of a lining with its single numbers on the standard basic elements (Annex G), the improvement ΔL of a floor covering on the heavyweight and the three lightweight reference floors with the form of Figure H.4 (Annex H), the sound reduction index of a sealed joint per metre, corrected for the test arrangement, with the gap widths of a variable slit and the form of Figure J.7 (Annex J), and the sound intensity level of artificial rain on a roof, normalized with the reference glass pane (Annex K), with the ISO 717 reference curves as published tables
- Sound Insulation by Intensity (ISO 15186): the ISO 15186-1/-2 sound reduction index from the intensity scanned over the radiating face, for the whole element or element by element
- Sound Insulation Survey Method (ISO 10052): the ISO 10052 octave-band control method: the reverberation index and its room-class estimate, the airborne, impact, façade and service-equipment quantities and their survey reports
- Service-Equipment and Activity Noise (ISO 16032): the ISO 16032 engineering method, from the ISO/DIS 16032:2023 draft: the corner and reverberant-field positions and the ladder that decides how many, the band average, the background correction held at 2.2 dB, standardization over 50 Hz to 5000 Hz, the A- and C-weighted single numbers of Table 1, the operating conditions of Annex B with the activities of B.10, and the reproducibility of Table 2
- Laboratory Flanking Transmission (ISO 10848): the ISO 10848 junction vibration reduction index $K_{ij}$ and the flanking descriptors $D_\mathrm{n,f}$ and $L_\mathrm{n,f}$ measured on a test facility
- Insulation Ratings (ISO 717): the ISO 717-1 airborne and ISO 717-2 impact reference-curve engines with the spectrum adaptation terms $C$, $C_\mathrm{tr}$ and $C_\mathrm{I}$, the enlarged-range and one-decimal variants, and the ISO 717 fiche
- Spanish Building Code (CTE DB-HR): the DB-HR global indices $R_\mathrm{A}$, $R_\mathrm{A,tr}$, $D_{\mathrm{n}T,\mathrm{A}}$ and $D_{2\mathrm{m,n}T,\mathrm{Atr}}$ from the direct Annex A formula over eighteen bands, the four normalised spectra, the clause 2 requirement tables and the window-size correction
- Façade Sound Insulation: the building envelope measured (ISO 16283-3), predicted from its element indices (EN 12354-3) and radiating an indoor source outwards (EN 12354-4)
- Heavy and Soft Impact Sources (ISO 16283-2): the rubber ball and the bang machine, their impact force exposure level and octave-band specification, the Fast-weighted standardization of the maximum impact level and the A-weighted single number of ISO 717-2 Annex D
- Structure-borne sound power of building equipment: the EN 15657 reception-plate method with the plate-injected power level, the spatial mean plate velocity, the loss factor from the structural reverberation time, and the blocked-force / characteristic-level / free-velocity source quantities from the low- and high-mobility plates
- Installed structure-borne sound from equipment: the EN 12354-5 prediction of the receiving-room sound pressure level from service equipment: the coupling term from source and receiver mobilities, the installed structure-borne power, and the per-path transmission with its energetic total
- Predicting Sound Insulation (EN 12354): airborne and impact flanking-transmission prediction between rooms (EN 12354-1/2)
- Detailed Per-Band Prediction (ISO 12354): the per-band detailed model of ISO 12354-1/-2 with in-situ element conversion and per-path contributions
- Predicting Panel Sound Insulation: theoretical airborne insulation of single panels (mass law and coincidence, Sharp) and double walls (mass-spring-mass, Bies), transmission through slits and apertures (Gomperts/Wilson-Soroka), plate radiation efficiency (Leppington/Maidanik) and the point mobilities of infinite plates and beams (Cremer)
- Floor-Covering Impact Improvement (ISO 16251-1): the ISO 16251-1 weighted improvement of impact sound insulation given by a soft floor covering, measured on a small heavyweight mock-up
- Predicting Resilient-Layer Performance: the prediction side of resilient layers, from the tapping-machine force model to the improvement of soft coverings, floating floors and wall linings
Materials and surfaces #
Where the coefficients every room and insulation model consumes come from: the laboratory instruments that measure them, the single-number ratings that summarise them, and the models that predict them.
- Reading a material datasheet: what the numbers on a data sheet or a test report are before you type them into a catalogue of your own: the practical coefficient of ISO 11654 against the one-third-octave coefficient of ISO 354, a level declared under CE marking, the mounting, the revision, and the sound reduction index and impact improvement of a laboratory report, each rated again from its bands
- Sound Absorption Measurement and Rating: reverberation-room absorption measurement (ISO 354), the weighted rating $\alpha_\mathrm{w}$ with its classes (ISO 11654), and the measurement uncertainty (ISO 12999-2)
- Suspended ceilings: the EN 16487 test code: the specimen size, the type E mounting depth and the air-absorption cap that make one ceiling measurement comparable with another, and Table 1, the reproducibility between European laboratories
- Airflow Resistance: static and alternating determination of airflow resistance, specific resistance and resistivity (ISO 9053-1/-2), and what the resistivity feeds in the porous models
- Impedance Tube: normal-incidence absorption, surface impedance and transmission loss by the standing-wave-ratio, transfer-function and transfer-matrix methods (ISO 10534-1/-2, ASTM E2611), plus the virtual FDTD tube
- Porous and Multilayer Absorbers: the Delany-Bazley, Miki and Johnson-Champoux-Allard porous models, the transfer-matrix multilayer solver with perforated, microperforated (Maa) and membrane layers, and the random-incidence Paris integral
- Metamaterial Absorbers: the critical-coupling condition for perfect sound absorption and the slow-sound slit panel loaded by Helmholtz resonators (Jiménez et al. 2016/2017), with the transfer-matrix model and the design solver
- Diffusers and Their Coefficients: random-incidence scattering (ISO 17497-1), the free-field diffusion coefficient (ISO 17497-2), Schroeder quadratic-residue design and the Fraunhofer far-field prediction that grades a well-depth sequence before it is built
- Metadiffusers: deep-subwavelength Schroeder diffusers from slits loaded by Helmholtz resonators (Jiménez et al. 2017): slow sound, per-well reflection phases, ternary sequences and the published quadratic-residue design evaluated end to end
- Dynamic stiffness of resilient materials: the EN 29052-1:1992 dynamic stiffness per unit area of floating-floor resilient layers from the load-plate resonance, and the floating-floor natural frequency
- In-situ Road-Surface Absorption: in-situ road-surface absorption by the extended-surface subtraction technique (ISO 13472-1) and the spot method (ISO 13472-2), with the Adrienne window, the sampled-area radius and the plane-wave limits of the spot tube
Vibration and structure-borne sound #
Vibration as a source of sound, as a human exposure in its own right, and as the diagnostic signature of a machine.
- Mechanical mobility and the FRF family: the ISO 7626-1:2011 family of motion-per-force frequency-response functions (receptance, mobility, accelerance and their reciprocals, Table 1), conversion through the receptance pivot, and the single-degree-of-freedom reference resonator (Annex A)
- Mobility by impact excitation: the ISO 7626-5:2019 measurement with an impact hammer: the energy spectral density of the pulse, the force and exponential windows (8.5), the average over impacts with its coherence (8.6, 9.1), the damping the window adds taken away by Annex A, and verdicts on double hits, the force spectrum, overload, the response decay and the channel match
- Dynamic transfer stiffness of resilient elements: the ISO 10846 dynamic transfer stiffness $k_{21}$ of vibration isolators: the level $L_k$ re 1 N/m and loss factor, the direct and indirect (transmissibility) determination methods, the one-third-octave band average every part asks for, the adequacy checks of the rig and the effective blocking mass (ISO 10846-4), the driving-point method with its upper limiting frequency and Annex B budget (ISO 10846-5), and the Annex-A relation to mechanical impedance and effective mass
- Bending-wave transmission at plate junctions: the wave-approach (Cremer/Craik/Hopkins 5.2.1.3) frequency-independent bending-wave transmission coefficients for rigid X, T, L and in-line plate junctions, their diffuse-field angular average, and the derived coupling loss factor and vibration reduction index $K_{ij}$
- Vibration damage to structures: the DIN 4150-3:1999-02 guideline values for the peak particle velocity at a foundation, in the topmost floor plane and on a buried pipeline, by building class, frequency and how often the vibration occurs, with the bending stress of Formula (1)
- Predicting a building's own frequency: the ISO 4866:2010 Annex D empirical predictors of the fundamental translation frequency, the coefficient ranges the national codes span, the f = 46/h fit to 163 measured buildings with the ± 50 % it admits, and the damping range D.4 reports
- Human Vibration: whole-body and hand-arm frequency weightings (ISO 8041-1), weighted r.m.s. acceleration, running r.m.s./MTVV/VDV and crest factor (ISO 2631-1), vibration in buildings (ISO 2631-2), vibration total value and daily exposure $A(8)$ (ISO 5349-1/-2), and the exposure action/limit values of Directive 2002/44/EC
- Multiple-shock whole-body vibration: the ISO 2631-5:2018 spinal-response model: the seat-to-spine transfer function, the acceleration and daily dose from the response peaks (Clause 5), and the compressive stress, stress variable $R$ and Weibull probability of lumbar injury (Annex C)
- What a seat does to the vibration: the ISO 10326-1:2016 laboratory method, with the SEAT factor of Formula (2), the three runs that must agree within 5 %, the correction to the input the test intended, and the transmissibility at resonance of the damping test
- Verifying a human-vibration meter: the ISO 8041-1:2017 tolerances on the frequency weightings, with the four transition frequencies of Table 4, the three distinct tolerance pairs Table 5 prints across them, the reference conditions of Table 1 and the indication tolerance of Table 2, giving a vibration meter the kind of verdict IEC 61672-1 gives a sound level meter, and the personal vibration exposure meter of ISO 8041-2:2021, graded with the same burst tables, its own two-row Table 2 and its own uncertainty limits
- Evaluating machine vibration (ISO 20816-1): grading a machine from one broad-band measurement: the four evaluation zones, the frequency-shaped velocity criterion of Figure 9 with the zone factors of Annex C.2, the typical boundary ranges of Table C.1, the vector reading of a change that a magnitude comparison misses, and the printed zone boundaries and ALARM and TRIP settings ISO 10816-3 gives for industrial machines
- Machine fault frequencies: the kinematic fault-frequency families of rotating machinery (Norton & Karczub Section 8.4) overlaid on a measured envelope spectrum: rolling-contact bearing BPFO, BPFI, BSF and cage frequencies, gear-mesh frequency and sideband families, induction-motor supply, slip, pole-pass and rotor-slot harmonics, and fan, blower and pump blade-passing frequencies with the lobed interaction patterns of a ducted axial fan
- Measuring vibration immission (DIN 45669-1): the instrument DIN 4150-2 and DIN 4150-3 presuppose, with the band limitation and KB weighting of Formulae (3) and (4), the weighted vibration severity of Formula (1) and the clock maximum r.m.s. of Formula (2) with its two counting rules, the tolerance bands of Tables 2 and 3, the printed check values of Table 9 and of Table 8 as Corrigendum 1 redrafts it, the Annex E assessment velocity that judges a building against one number instead of a curve, and the numbers of DIN 45669-2 a measurement with it is planned by: the loose-mounting limits, the mass loading and the instrument confidence limits
- Vibration next to a railway (DIN 45672): one train passage reduced the way DIN 45672-2 reduces it, from the three stretches of Clause 5 and the running r.m.s. to the event value of an hour of traffic, the third-octave spectra of Figure 6 and the narrow-band density with Table 1, and the ground constants of DIN 45672-1 from two wave speeds, with the two formulas of Clause 4.5 that are printed wrong
- Vibration and people in buildings (DIN 4150-2): the guide values of Table 1 by area, time of day and source, the procedure of 6.2 as a verdict with the criterion that decided it, the assessment vibration severity of Formulae (4) and (5) with the rest-time weight and the exposure A_r allows, the rules of 6.5 for rare events, roads, railways and construction sites with Table 2 and the interpolation of Figure 3, the railway classes of Annex A with the interval their spread puts on the result, the trains an hour of Figure D.1, and the estimate of Clause 7 from a peak velocity, with Formula (A.1b) in the corrected form registered in the errata
- Railway vibration by category of train (E DIN 4150-2): the draft of August 2023 that is to replace DIN 4150-2, with one clock interval per passage, a KB_Fmax that is 1,5 times the r.m.s. of a category, the weighting factors of Table 2 by kind of train and alignment, the night-time upper value of a new line, the 25 % rule of an altered one, and the rest of the draft's changes to Table 1, the procedure and the sources behind an edition switch
- Predicting railway vibration (E DIN 45672-3): the additive chain of Formula (1) from an emission spectrum through the ground, the foundation, the floor and a mitigation to a third-octave spectrum on a floor, the speed rescaling and the ground transmission of Clauses 5.2 and 5.3, the six transfer tables of Annex A, the KB values of Clause 7 formed from the spectrum, and Annex B on a train as a line of point sources, with what the Annex C print gets wrong
- Predicting vibration before measuring (DIN 4150-1): the far field of Formula (1), the decay of Formula (2) with the exponents of Figure 1 and the damping of the ground, the building on its ground and its floors of Clause 4.3, the blast, the falling mass and the machine hall of Formulae (5) to (7) with the nomogram of Figure 3 read off the page, and what a track excites at
Environment and transport #
The source-path-receiver problem stretched over open ground: emission models for traffic and turbines, outdoor propagation, and the ratings and legal limits a noise map is drawn against.
- Outdoor Sound Propagation: atmospheric absorption $\alpha(f)$ (ISO 9613-1) and the ISO 9613-2 general method: divergence, atmospheric absorption, ground effect and barrier screening
- Spherical ground effect and advanced barriers: the Weyl-Van der Pol spherical-wave reflection coefficient over a finite-impedance ground, and wave-theoretic barriers (Kurze-Anderson Fresnel number, exact rigid half-plane, thick barriers and the coherent four-path barrier on the ground)
- Atmospheric refraction: rays and the GFPE: effective sound-speed profiles (linear and logarithmic), ray tracing through a refracting atmosphere (curved paths, turning points, closed-form curvature radius and shadow-zone distance) and the Green's Function parabolic equation (GFPE) for the relative-level field over the range-height plane, anchored to the spherical ground effect in the homogeneous limit
- Sound reflection of a noise barrier in situ: the sound reflection index of EN 1793-5 measured in front of a road noise barrier where it stands, with the signal subtraction, the Adrienne window, the single number DL_RI and the low frequency limit a sample's size sets
- Quality of outdoor sound software (ISO 17534-1): a program's results against the limits of its certified test-case results (the TRC form of Annex B), the 0,1- and 0,9-quantiles of the level differences a faster modified configuration introduces (Annex C, Table C.1 and Formulas (C.1) and (C.2)), where that sample is taken, and the precision of a method across programs from a round robin
- Wind-turbine noise: apparent sound power & tonal audibility: the IEC 61400-11 apparent sound power level referred to the rotor centre and the tonal-audibility chain (Zwicker critical band, masking-noise level and audibility criterion)
- Wind turbine sound at a dwelling (IEC TS 61400-11-2): the receptor side of wind turbine acoustics: wind speed between heights and the binning wind speed, bin averages with their uncertainty and the background correction, the low frequency level indoors of Annex C, the emergence of Annex J, the amplitude modulation rating of clause 13 checked against the IOA reference code, and the rating level of Annex A
- CNOSSOS-EU railway source emission: the common EU method for railway noise emission (Directive 2002/49/EC Annex II, section 2.3 and Appendix G): rail and wheel roughness with the contact filter and the transfer functions, impact noise, curve squeal, traction, aerodynamic noise, bridges, the source directivity and the two equivalent source lines at 0,5 m and 4,0 m
- Railway rolling-stock noise: the ISO 3095 type test: the stationary, constant-speed, starting and braking tests of a railway vehicle (ISO 3095:2013), the pass-by level and the transit exposure level of ISO 3095:2005, the reference track judged by the rail roughness of EN 15610 and the track decay rates of EN 15461, the small deviations of Annex C and the uncertainty budget of Annex G
- Road-surface noise: the statistical pass-by method: the ISO 11819-1 regression of the maximum level of each passing car and heavy vehicle on the logarithm of its speed, the vehicle sound levels at the reference speeds of Table 1, the Statistical Pass-By Index and its difference from a reference surface
- CNOSSOS-EU road traffic source emission: the common EU road source of Annex II to Directive 2002/49/EC (section 2.2 and Appendix F): rolling and propulsion sound power per vehicle category with the corrections for road surface, air temperature, studded tyres, gradient and junctions, and the directional sound power per metre of source line
- Environmental levels: $L_\mathrm{den}$, $L_\mathrm{dn}$ and the composite rating levels of ISO 1996-1, with the ISO 1996-2 tonal adjustment, residual-noise correction and measurement uncertainty budget
- Spanish Noise Regulation (RD 1367/2007): the corrected level $L_\mathrm{Keq}$ with its tonal, low-frequency and impulsive corrections, the evaluation periods split into noise phases, the acoustic quality objective and immission limit tables, and the Article 25 compliance check of an activity
- Impulsive-sound prominence: the NT ACOU 112:2002 predicted prominence of impulsive sounds (onset rate and level difference) and the graduated adjustment $K_\mathrm{I}$ added to $L_\mathrm{Aeq}$
- Soundscape analysis (ISO/TS 12913): how people hear a place, by ISO/TS 12913-2 and ISO/TS 12913-3: the questionnaire and its scale values, the median and range per site, the pleasantness and eventfulness of Figure A.1, the Spearman and Pearson correlations with their probability values, the Method B statistics, the binaural metrics of Table D.1 at both ears and the minimum reporting record
- Impulsive sound exposure statistics (ISO 13474): how often a distant blast is how loud, the ISO 13474:2009 distribution of the single-event sound exposure level over replica atmospheres, with the long-term and rating levels, the spread for turbulence and the exceedance levels
Aircraft noise #
The two internationally negotiated families: certification, which fixes one number per aircraft type at reference points around a runway, and contour methods, which predict what a whole airport does to the land around it.
- Aircraft noise: Effective Perceived Noise Level: the ICAO Annex 16 Vol. I Appendix 2 EPNL (perceived noisiness and PNL, the tone correction by the slope method, and the 10 dB-down duration correction), the IEC 61265 measurement-system verifier, the SAE ARP 5534 one-third-octave-band atmospheric absorption (SAE Method), and the ECAC Doc 29 noise-power-distance (NPD) event-level interpolation
- Airport Noise (ECAC Doc 29): the noise-power-distance engine, the per-segment single-event chain of lateral attenuation, engine-installation and duration effects, and the ground-grid SEL contour of a full departure or approach
- The ANP fleet database: the EASA/EUROCONTROL Aircraft Noise and Performance database shipped with the package, its noise-power-distance curves and default fixed-point trajectories per aircraft type, and the wiring that runs the ECAC Doc 29 single-event level and ground-grid contour straight from an aircraft identifier
- NPD data for the air at the airport: the ECAC Doc 29 Vol. 2 Appendix D recalculation of the ANP noise-power-distance curves for the temperature, humidity and pressure of a study, from the SAE AIR-1845 rates and each aircraft's spectral class, with SAE ARP 5534 or the SAE ARP 866A absorption of ISO 3891 Annex A, and the event level and contour in that air
- Rotorcraft noise: the hemisphere method: the ECAC Doc 32 / NORAH2 helicopter noise hemisphere source model, its propagation adjustments (spherical spreading, ISO 9613-1 / Table 4 atmospheric absorption, Chien-Soroka ground effect over CNOSSOS impedance ground), the flight-condition interpolation and track kinematics, the single-event SEL/LASmax/EPNL with ground-grid contours, and the terrain machinery (mean ground plane, log-mean flow resistivity, rubber-band screening, digital elevation models)
Underwater acoustics #
The same physics on a different scale and a different reference: levels re 1 µPa, a medium that refracts sound into channels carrying it for kilometres, and the marine mammals that hear it.
- Underwater acoustics: radiated noise & pile driving: the ISO 18405 reference levels (SPL, SEL, peak re 1 µPa), the ISO 17208 ship radiated noise level and equivalent monopole source level via the Lloyd's-mirror correction, and the ISO 18406 single-strike, peak and cumulative pile-driving sound exposure
- Underwater sound propagation: closed-form propagation loss (geometrical spreading plus volume absorption by Francois-Garrison, Ainslie-McColm or Thorp), Weston's four shallow-water propagation regimes with their transition ranges, the speed of sound in sea water (UNESCO/Chen-Millero, Del Grosso, Mackenzie, Medwin) with the sound-speed profile, the passive/active sonar equation and the detection-range inversion, seabed reflection loss (Rayleigh) and the ocean ambient-noise spectrum (Wenz wind/thermal plus JOMOPANS-ECHO ship-traffic source levels)
- Underwater propagation solvers: the numerical solvers of the range-independent stratified ocean (Jensen et al.): the finite-difference normal-mode sum, Runge-Kutta ray tracing with Snell's invariant, Gaussian beam tracing on those same rays for a field that stays finite at caustics, and the split-step Fourier parabolic equation, each validated against an exact closed form, plus the guidance for choosing between them and the closed forms
- Marine-mammal noise exposure: the hearing side of underwater noise: the Southall et al. group audiograms and Ainslie's orca audiogram, the regulatory auditory weighting functions with the guidance version selectable (NMFS 2024 v3.0, NMFS 2018 v2.0, Southall et al. 2019), the TTS and injury onset criteria with their dual impulsive metric, and the weighted cumulative exposure of a pile-driving campaign against them
Sources and devices #
Emission: numbers that belong to the device rather than to the room or the distance it is heard at, from sound power to loudspeaker distortion to duct silencers.
- Sound Intensity (p-p): two-microphone intensity and field indicators
- Sound Power: choosing the determination method for a source and declaring the noise emission (ISO 4871)
- Sound power by pressure methods: the enveloping surface of ISO 3744/3746 and the precision anechoic grade of ISO 3745, and the sound energy level of a burst over the ISO 3744/3746 enveloping surface
- Free field and reference sources: the ISO 26101 inverse-square-law qualification of an anechoic or hemi-anechoic room against the Annex A that Amendment 1:2017 wrote into ISO 3745, with the fit of $b$ and of the origin inside the source, and the calibration and clause 5 requirements of an ISO 6926 reference sound source
- Sound power in the reverberation room: the direct and comparison methods of ISO 3741, for sound power and for the sound energy of a single event
- Sound power in small test rooms: the ISO 3743-1 comparison in a hard-walled room and the ISO 3743-2 direct and comparison methods in a special reverberation room, for small movable sources
- Sound power in situ by comparison: the ISO 3747 comparison against a reference sound source where the machine works, with the sound energy level of an impulsive source
- Sound power in a duct: the ISO 5136 in-duct method for fans, with the sampling-tube flow and modal correction $C_{3,4}$ of Annex A, the plane-wave relation and the Table 2 reproducibility
- Sound power of multisource industrial plants: the ISO 8297 contour round a whole plant, with the positions, heights and directions of the microphones, the area, proximity and air absorption terms of clause 10, the uncertainty of Table 1 and one verdict on the arrangement
- Sound power in the 16 kHz octave: the four ISO 9295 methods for the octave band above ISO 3741 and ISO 3744, with the room constant from the measured reverberation time or from the air absorption of Annex A, the comparison for broadband noise and for tones, and the free-field correction beyond 2 m
- Sound power of steam turbine sets: the IEC 61063 test code for a steam turbine set and its driven machinery in its own hall, with the stepped measurement surface 1 m out and its area, the key positions of Figure 2, the stepped background correction of Table 2 against the $K_{1\mathrm{A}}$ of ISO 3746, the environmental correction of Figure A.3 and the report of clause 10
- Sound power by intensity scanning: the on-site scanning of ISO 9614-2 and the ISO 9614-3 precision grade
- Emission sound pressure at a work station: the ISO 11200 group, which turns a reading at the operator's position into what the machine alone puts there, with the background and local environmental corrections and the grade each earns
- Sound power from surface vibration: the ISO/TS 7849 estimation of a machine's radiated airborne sound power from its surface vibratory velocity and a radiation factor: the velocity level and calibration, the surface mean, and the Part 1 upper limit ($\varepsilon = 1$) versus the Part 2 engineering value
- Electroacoustics: distortion & frequency response: the IEC 60268-3 distortion set (THD, nth-order harmonic, THD+N and SINAD via AES17, SMPTE and CCIF intermodulation, dynamic intermodulation and weighted THD), the AES17 dynamic range and idle channel noise, and the Bendat & Piersol H1/H2 frequency-response estimators with the ordinary coherence $\gamma^2$
- Loudspeaker characterisation (IEC 60268-5): the sensitivity conventions of a loudspeaker datasheet, the radiating piston (radiation impedance and directivity) behind its polar response, and the IEC 60268-5 characteristics fiche
- Microphone characterisation (IEC 60268-4): the IEC 60268-4 sensitivity references, the directional patterns and directivity index, the inherent noise in dB(A) and dB(468), and the IEC 60268-4 characteristics fiche
- Swept-sine distortion and phase utilities: harmonic separation from one exponential sweep (Farina 2000) with the synchronized swept-sine of Novak et al. 2015 for coherent harmonic phases, THD as a function of the excitation frequency, and minimum phase from $|H|$ (real cepstrum), group delay and excess phase
- Audio-frequency induction loops (IEC 60118-4): the hearing loop for telecoil listeners, the 400 mA/m reference and the Biot-Savart field of a loop, the loop as a load reproducing Table B.1 of IEC 62489-1, the true-RMS meter, the pink noise and the combi signal, the verdicts on a room, a refuge and a counter as Amendment 1 writes them, the overload test of Table 4, the amplifier's compliance voltage, AGC range and quadrature error, and the neck loop with the two types of the draft Amendment 2
- Headphones and earphones (IEC 60268-7): the simulated programme signal of IEC 60268-1 from Table II and from the filter of Figure 2, the classification code, the rated impedance, the characteristic voltages with the power summation of Figure 3, the powers and working levels, the clipped signal of the limiting voltages, the protective device, the coupler, comparison and ear canal responses, the crosstalk, the distortion test signals and the probe microphone of Annex B
- Programme loudness & true peak: the ITU-R BS.1770-5 programme loudness (K-weighting, gated 400 ms blocks, channel weights including the Annex 3 positions) and the oversampled true-peak level in dBTP, with the EBU R 128 −23 LUFS practice, the Tech 3341 EBU Mode momentary/short-term/integrated meters and the Tech 3342 loudness range, validated against the official EBU test signals
- Quasi-peak programme meter (ITU-R BS.468-4): the psophometric quasi-peak detector of clause 2: eleven tone-burst acceptance windows where the Recommendation prints no time constant at all, the clause 2.6 calibration that makes a reading dBqps, and the three fitted time scales those windows identify only to within a factor of 1.61 to 2.09
- Silencers: reactive silencers by the four-pole transmission-matrix method (the closed-form expansion chamber, Helmholtz, quarter-wave and extended-tube resonators) with transmission and insertion loss, the independent FDTD cross-check, and the design trade-off against dissipative linings (Bies, Hansen & Howard; Munjal)
- Measuring a silencer: the substitution measurement a catalogue insertion loss comes from, its reverberation correction, the octave fold done on the transmitted energy, the three-or-five microphone rule, and the reproducibility both standards publish (ISO 7235, ISO 11691)
- Enclosures and cabins measured: what a built enclosure is actually worth, as the difference of two sound power determinations, with Table 1 deciding what may be declared from which method and Annex C asking whether the room is good enough at all; and the cabin, the inverse problem, whose in-situ source-position count is read off the spread of its own answer (ISO 11546-1, ISO 11546-2, ISO 11957)
- Silencers, screens and barriers in situ: the silencer measured in the duct it was fitted to, the screen wheeled away for one of the two runs, and the outdoor barrier that was built before anyone thought of measuring it, with the twenty installations of Figure 1, the reference microphone that normalises the source, and the three different corrections for background noise that must never be merged into one (ISO 11820, ISO 11821, ISO 10847)
- Duct-borne noise: fan to room: the end-to-end HVAC duct-noise calculation (fan sound power from the operating point, duct runs, elbows, splits, flexible duct, splitter silencers and plenums, the regenerated noise of silencers and air terminal devices, the room effect and the NC verdict) anchored on Long's worked Table 14.9 sheet, with the higher-order-mode cut-on that limits every plane-wave method (Long; Norton & Karczub; ASHRAE; AHRI 885)
- HVAC noise the German way (VDI 2081): the same fan-to-room chain by the German guideline: the assembly-type fan model of Equation (13), the attenuation and flow noise of each duct element, the splitter silencer and its own noise, and the room step of Equation (36), against the worked sheet of Part 2 Table 1 (VDI 2081 Blatt 1:2001-07; Blatt 2:2005-05)
- Control valve noise (IEC 60534-8-3): the whole of Clause 5 for a valve with standard trim: the five flow regimes and the pressure ratios that define them, the valve style modifier and jet diameter that set the peak frequency, the acoustical efficiency of each regime, the internal level at the pipe wall and its third-octave spectrum, the pipe transmission loss shaped by the ring and coincidence frequencies, and the A-weighted level one metre outside, against the seven worked examples of Annex A (IEC 60534-8-3:2010)
- Hydrodynamic valve noise (IEC 60534-8-4): the same chain for a liquid: the differential pressure ratio and the characteristic pressure ratio that decide whether the valve cavitates, the turbulent and cavitating acoustical efficiencies that add up to the sound power, the two peak frequencies they radiate at, the pipe transmission loss anchored at the ring frequency, and the level one metre outside, against the three worked examples of Annex A (IEC 60534-8-4:2005)
- Room to room: partition, receiving room, criterion: the composed room-to-room chain (the reverberant source-room level from a sound power level and the room constant, the partition transmission loss, the equivalent absorption area of the receiving room, the received spectrum and its NC verdict), the transmission loss a partition needs to meet a criterion curve and the same inverse for a lined machine enclosure, anchored on the worked octave-band answers of Norton & Karczub problems 4.16, 4.18 and 4.21
- Industrial noise control: HVAC & enclosures: the HVAC duct methods (end reflection, elbows, plenums and flow-generated noise) and machine-enclosure insertion loss from a supplied panel $R$ and the interior room constant (Bies, Hansen & Howard)
Wave simulation #
Where the rest of the library predicts a number, this one computes the wave field itself on a finite-difference grid.
- 2D FDTD wave simulation: the deterministic staggered-grid pressure-velocity FDTD solver (Attenborough & Van Renterghem 2021, chapter 4) with Gaussian, tone and arbitrary-signal sources, pressure probes, rasterised obstacles, rigid/impedance/absorbing boundaries, the near-to-far-field Kirchhoff-Helmholtz chain, and a result object with probe histories and field snapshots
- Elastic waves and fluid-solid coupling: the elastic P-SV companion solver (Virieux 1986) on the same staggered grid, with stress-imaging free surfaces and Rayleigh waves, fluids as the shear-free limit, and fluid-solid coupling validated against Brekhovskikh & Godin: mode conversion, Scholte interface waves and immersed-plate transmission
- Where a simulation runs: NumPy, CuPy and torch: both solvers on NumPy, on an NVIDIA GPU through CuPy or on PyTorch, chosen with a
Computerequest, ause()block orPHONOMETRY_COMPUTE, with the fallback to NumPy, the provenance every result records, the bit-identical CuPy fields and the published torch bounds, and decimated snapshots - Sending a simulation to a remote node: either solver's run on another machine over SSH, on a Docker node built from the repository's image or in a Python environment, with or without a GPU: the hosts file that names a node, what each side checks, the errors and the fallback to NumPy, progress and timeouts, security in both directions, and when a run there gives the numbers of a run here
- Where heavy work runs: CPU or GPU, here or on a node: the map of where simulations and videos run, on this machine or on a node over SSH, on a CPU or a GPU: the devices a run steps on, a node in the hosts file, the CPU and GPU encoders of a video, and the page that explains each
Reference #
- API Reference: curated quick table of every public function and class
- Generated API reference: one page per public module, generated from the source docstrings (
make api-docs). English only; the Spanish site serves it via locale fallback - Theory: standards, math and design decisions, split by domain
- Signal analysis: filter banks, weightings, time integration, level and exposure metrics, sound intensity, GUM uncertainty
- Perception and hearing: equal-loudness contours, loudness models, sound quality, tone prominence, STI/SII, hearing statistics
- Rooms and buildings: room acoustics, noise criteria, insulation and ratings, flanking prediction
- Materials and surfaces: scattering and diffusion, in-situ road-surface absorption, absorption ratings, airflow resistance, impedance tube
- Environment and transport: environmental descriptors, impulsive adjustment, outdoor propagation, occupational exposure, sound power
- Vibration: human vibration weightings and metrics, multiple-shock spinal model
- Glossary: every quantity the guides compute, grouped by domain, each with its symbol, a one-sentence definition, its unit, the standard and clause that defines it and the guide that implements it, plus the table of symbols that collide across domains
- Bibliography: the books and papers behind the guides, grouped by domain, every entry with a verified DOI or official publisher link
- Published catalogues: the material and fluid values this library reads rather than computes, one page per catalogue, each row naming the book, the table, the PDF page and the printed folio it was read on. Generated from the packaged data files (
make catalogue-data), so it has no hand-written mirror here, the same as the generated API reference - Report fiches: every PDF fiche
.report()renders, grouped by area, each linking to the fiche and to the guide that explains the measurement behind it. Built from the<ReportPreview />each guide already carries beside the fiche it shows, so it has no hand-written mirror here either: a second list would be a second thing to keep in step with the guides, which is the one failure this page is built to prevent - Conformance report: auto-generated numerical validation: every check pins a standard clause's expected value against the library's computed value, regenerated in CI
- Standards errata: defects found in the published standards themselves during implementation: misprints, examples contradicting their own normative text, ambiguous wording, each with evidence and the library's disposition; a Spanish edition, translated entry for entry, lives in ERRATA.es.md
- Same bits on every machine: why floating point drifts between hosts, what the deterministic weighting design pins, and the measured evidence that a designed filter is byte-identical on any CPU, BLAS or thread count, and what still separates the platforms
Development #
Run the test suite with pytest tests/, the full quality gate with make check
(ruff + mypy + bandit + tests), and regenerate the documentation images with
make graphs. See CONTRIBUTING.md.
Contributing #
- Contributor style guide: how phonometry names things, shapes its API, documents and comments its code, handles numbers, tests itself, documents and draws, and words its prose, for the people and the coding agents who change the library; every rule has an identifier a review can cite and names the check that enforces it. It lives only in this repository, with no page on the site