Hearing Range¶

The condition-indexed region of sound frequency and level in which a listener or species meets a declared auditory-detection criterion, often summarized by level-specific lower and upper frequency limits.

Version: v3 · 2026-09-07 · History
Domain-specific #: 1991
Origin domain: audiology
Subdomain: auditory threshold measurement
Aliases: Audible frequency range, Auditory frequency range

Core Idea¶

Hearing range is the region of acoustic frequency and sound level over which a specified listener, population, or species meets a declared auditory-detection criterion under stated conditions. It is commonly compressed to a lower and upper frequency limit, but the complete object is a threshold curve or audibility region: sensitivity changes with frequency, and an endpoint reported at one presentation level can move when the level changes.

This is why the familiar statement that human hearing spans roughly 20 Hz to 20 kHz is only a nominal shorthand. It does not mean every human detects every tone in that interval at the same pressure. Thresholds vary across frequency, listeners, age, exposure history, ears, transducers, and test environments. ISO 226:2023 gives equal-loudness-level data only for tightly specified free-field, binaural conditions and otologically normal listeners aged 18–25^[1]; ASHA's clinical guidance uses calibrated pure-tone threshold procedures across a narrower conventional test-frequency set, with extended high-frequency testing as a special procedure.

Comparative hearing makes the conditional structure particularly visible. Heffner and Heffner define animal hearing ranges from behavioral audiograms and emphasize species differences in low-frequency limits, high-frequency limits, and absolute sensitivity. A cat range reported at 70 dB SPL and a bat range reported at 60 dB SPL are level-contour summaries, not immutable anatomical endpoints. Changing the chosen level or behavioral criterion changes the summarized range even when the underlying threshold curve is unchanged.

The abstraction therefore joins a biological detector, a stimulus space, a measurement protocol, and a criterion. It supports clinical audiograms, comparative sensory ecology, acoustic welfare decisions, stimulus design, and claims about whether a signal lies inside an organism's measured audibility. It is not merely a list of species-specific numbers.

Structural Signature¶

The abstraction has ten roles:

auditory system — an individual ear/listener, defined population, or species;
acoustic medium and geometry — air, water, free field, earphone coupling, or another declared transmission condition;
signal class — pure tone, narrow-band signal, complex sound, pulse, or other specified stimulus;
frequency coordinate — the frequencies sampled or continuously modeled;
level coordinate and unit — sound pressure, hearing level, force level, or another calibrated intensity measure;
response method — behavioral report, conditioned response, auditory brainstem response, otoacoustic measure, or another operational endpoint;
detection criterion — the response rule and probability or trial convention that counts as threshold;
threshold function — the minimum level meeting the criterion at each frequency;
range summary — lower and upper endpoints at a stated reference level or within an instrument's testable region;
condition and uncertainty record — age, ear, health, masking, background noise, calibration, sample size, variability, censoring, and test limits.

Its invariant is:

audibility is assigned only relative to a specified auditory system, stimulus, level, method, criterion, and condition; the hearing range is the resulting frequency-by-level detection region or a declared slice through it.

If T(f|C) is the threshold level at frequency f under conditions C, the audible region is the set of pairs (f,L) for which L reaches or exceeds T(f|C), subject to any upper safe or technical limit. At a fixed reference level L0, the summarized frequency range contains frequencies satisfying T(f|C) ≤ L0.

What It Is Not¶

It is not one universal human constant. “20 Hz–20 kHz” is a conventional population shorthand, not a guarantee for an arbitrary person, age, level, or method.

It is not an audiogram, although an audiogram is a primary representation from which a range can be read. The audiogram records frequency-specific thresholds; the range is the audibility region or endpoint summary derived from them.

It is not equal loudness. Equal-loudness contours compare levels judged equally loud above threshold. Hearing range concerns whether a signal is detected.

It is not a difference threshold or Weber fraction. Those concern discrimination between stimuli already detectable, whereas hearing range is anchored in absolute or operational detection.

It is not the frequency response of an earphone, microphone, or loudspeaker. Instrument bandwidth constrains what can be tested and delivered but is not biological audibility.

It is not a species' vocalization range or echolocation-call spectrum. Production and detection can be related without being identical.

It is not automatically a diagnosis of hearing loss. A clinical judgment compares measured thresholds with calibrated reference values and relevant standards; the bare existence of endpoints does not classify impairment.

It is not a safe-listening range. Audibility, discomfort, damage risk, and permitted exposure are different boundaries.

Scope of Application¶

In clinical audiology, pure-tone threshold audiometry measures an individual's hearing sensitivity at specified frequencies. ASHA recommends conventional air-conduction testing across frequencies from 250 through 8000 Hz, with 125 Hz in some cases and 9000–16000 Hz for extended high-frequency purposes^[2]. Bone-conduction, masking, transducer calibration, and test environment are separately controlled. ISO 8253-1 specifies procedures for air- and bone-conduction threshold audiometry and explicitly excludes above-threshold and electrophysiological procedures from that standard's scope^[3].

In psychoacoustics, threshold curves and equal-loudness data characterize normal human sensitivity under laboratory conditions. ISO 226:2023 fixes signal type, geometry, listener orientation, binaural listening, listener age, and otological status. These constraints demonstrate that even a normative curve is a conditioned population construction.

In comparative biology, behavioral audiograms let researchers compare species, match communication or echolocation signals to sensitivity, and evaluate whether experimental or environmental sounds are likely audible. Heffner and Heffner's review stresses that overlapping nominal ranges can conceal substantial differences in absolute sensitivity.

Physiological methods extend the scope to subjects unable to provide a deliberate report, including infants and nonhuman animals. An auditory brainstem-response threshold or otoacoustic-emission threshold measures a physiological endpoint, not necessarily the same decision process as a behavioral threshold. Comparisons must preserve method labels.

The node covers frequency range and, when explicitly declared, the audible range of levels. It does not merge detection with discomfort or damage. A full auditory dynamic range uses both lower threshold and an upper response boundary and should be labeled accordingly.

Clarity¶

A valid hearing-range claim should answer:

Whose hearing is being characterized: one ear, one listener, a population, or a species?
In which medium and acoustic geometry was the stimulus delivered?
Which signal class and duration were used?
Which frequency and level units are reported?
Was the response behavioral, clinical, electrophysiological, or another kind?
What response rule counted as detection?
Is the result a full threshold curve or only endpoints at a specified level?
Were endpoints observed, interpolated, extrapolated, or censored by instrument limits?
Which listener conditions, age, ear, health, and background-noise controls apply?
Is uncertainty or interindividual variation reported?

ASHA defines clinical threshold as the lowest dB hearing level producing responses on at least half of a series of ascending trials, with at least two responses out of three presentations at one level^[2]. That rule is not the only possible psychophysical criterion, but it shows why “heard” needs an operational definition.

A number without its contour level is suspect. The published domestic-cat example gives a frequency span specifically for sounds at 70 dB SPL. Removing “70 dB SPL” changes the claim from a measured slice to an apparently absolute limit.

Manages Complexity¶

Auditory sensitivity is a curved, multidimensional relation. A single listener has different thresholds at different frequencies; populations add distributions; species add different ear structures and ecological specializations; procedures add criterion and calibration effects. Hearing Range packages these variables into a stable measurement architecture.

The abstraction makes comparisons tractable by separating the underlying threshold curve from its endpoint summary. Two species can share a similar lower and upper span yet differ substantially in their most sensitive frequencies. Two laboratories can report different endpoints because one uses 60 dB SPL and another 70 dB SPL. A clinical audiogram can appear narrower than a laboratory range because the audiometer samples only conventional frequencies.

It also localizes disagreement. If a behavioral audiogram and physiological curve differ, the question becomes whether the response criterion, neural endpoint, acoustic field, or sample explains the gap. Heffner's animal work shows that pinna (external ear) orientation within a sound field can affect sensitivity estimates, so procedural geometry is not administrative detail^[4].

For applied decisions, the range predicts whether a signal merits further audibility analysis. It does not by itself predict loudness, annoyance, localization, recognition, or harm. Those require additional mappings after the detection boundary is crossed.

Abstract Reasoning¶

Let f denote frequency, L sound level, C the complete condition record, and R a measured response. Define a detection rule D(R) and criterion p0. The threshold function can be represented as:

T(f|C) = the lowest L for which P(D(R)=1 | f,L,C) reaches p0.

The corresponding audibility region is A(C) = {(f,L): L ≥ T(f|C)} within the valid stimulus and safety domain. A level-specific frequency range is the projection F(L0,C) = {f: T(f|C) ≤ L0}. Its lower and upper endpoints are extrema of that set, if the set is connected and the endpoints are observed.

This formulation yields useful predictions. Raising L0 normally expands F into frequencies where thresholds are higher, so endpoints are level-dependent. Adding masking noise raises thresholds selectively and shrinks or reshapes the region. Age-related high-frequency loss can raise T at high f and lower the upper endpoint without equally affecting the low-frequency boundary. A transducer ceiling can censor T and create a false endpoint.

Method changes can shift the curve because the event being detected changes. A conditioned behavioral response includes peripheral detection, neural transmission, learning, motivation, and decision. An auditory brainstem response uses a physiological waveform criterion. The curves may correlate without being interchangeable.

The region need not be a perfect interval at every level. Local threshold peaks can create islands or gaps under a strict mathematical definition. Reporting only two extrema assumes a connected effective range and hides internal sensitivity structure; the audiogram retains it.

Knowledge Transfer¶

The exact abstraction transfers across humans, terrestrial animals, marine mammals, birds, and other hearing organisms when all roles remain literal: acoustic signal, auditory system, detection method, threshold curve, and condition-indexed range. It also transfers among behavioral and physiological measurement only if method identity remains explicit.

It does not transfer literally to a microphone's bandwidth, a radio receiver's passband, or the visible spectrum. Those systems have analogous input-response regions, but “hearing” requires an auditory biological system or a clinically recognized hearing pathway. Their portable structure is represented by Threshold, Measurement, Boundary, Sensitivity, and operating-range abstractions.

The domain-specific node is still useful because the scientific work lies in precisely those auditory accents: Hz and dB reference frames, air/bone/free-field delivery, audiometric criteria, ear and population identity, animal conditioning, and behavioral-versus-physiological interpretation.

Examples¶

Conventional human audiogram. Calibrated tones are presented to one ear at standard frequencies. The lowest hearing level satisfying the ascending-trial rule is recorded at each frequency. The resulting points show the tested portion of the listener's threshold curve, not a universal human range.

Extended high-frequency audiometry. Frequencies above the conventional 8 kHz clinical set are measured with an appropriate circumaural transducer and calibration. A failure to respond above an instrument limit is recorded as censored rather than silently treated as the biological endpoint.

Normal equal-loudness reference. ISO 226:2023 defines curves for young otologically normal binaural listeners in a specified free field. The lowest contour approaches hearing threshold, while higher contours address equal loudness. The population and acoustic conditions are part of the result.

Domestic cat behavioral range. Heffner and Heffner measured cat behavioral audiograms and reported a 48 Hz–85 kHz range at 70 dB SPL^[5]. The level qualification is essential: it states a slice through a threshold curve.

Jamaican fruit bat. A conditioned behavioral procedure yielded a range of 2.8–131 kHz at 60 dB SPL and best sensitivity around 16 kHz^[6]. The example shows that range breadth and best-frequency sensitivity are separate properties.

Masked listener. Adding broadband noise raises frequency-specific detection thresholds. The same auditory system can therefore have a smaller operational range under masking than in quiet.

Nonexample—microphone specification. A microphone advertised as 20 Hz–20 kHz has an engineered frequency response, not a hearing range.

Nonexample—vocal repertoire. A bat call spectrum describes sounds produced; only a measured auditory threshold curve establishes sounds detected.

Structural Tensions¶

Compact endpoints vs. full sensitivity curve. A two-number range is easy to compare; it suppresses best frequency, internal peaks, and threshold shape.
Nominal population range vs. individual variability. A standard shorthand aids communication; age, ear, health, and exposure can make it misleading for a person.
Behavioral validity vs. physiological access. Behavioral detection most directly answers whether an organism responds; physiological measures work without deliberate report but use a different endpoint.
Ecological realism vs. laboratory control. Natural sounds and fields matter to organisms; calibrated pure tones make thresholds reproducible.
Frequency breadth vs. absolute sensitivity. A species can detect a broad span only at high levels while being less sensitive than another species across their shared frequencies.
Observed boundary vs. instrument boundary. Nonresponse can reveal a biological limit or merely exceed transducer output, calibration, or test-frequency coverage.
Audibility vs. safety. Raising level can reveal endpoint frequencies; high levels can be uncomfortable or harmful and therefore ethically unavailable.
Cross-species comparison vs. method equivalence. Common units invite comparison, while conditioning, geometry, medium, and criterion may differ.

Structural–Framed Character¶

Hearing Range is structurally clear. It has a response system, two principal stimulus coordinates, an operational threshold rule, a boundary curve, projections at fixed level, and predictable changes under masking, aging, level, and method. A proposed range can be tested and falsified.

It is strongly framed. The identity requires ears or auditory pathways, acoustic calibration, species/listener biology, audiometric response procedures, sound-field or transducer conditions, and frequency- and level-specific conventions. Removing those produces a generic operating region rather than Hearing Range.

This combination supports a high-confidence domain-specific abstraction. Threshold carries the portable regime boundary; hearing science supplies the exact measured residual.

Structural Core vs. Domain Accent¶

The structural core is map a response threshold across one stimulus dimension, then treat the region beyond that curve in a second dimension as operationally available. Threshold supplies the response boundary. Measurement supplies calibrated coordinates and a repeatable procedure. Boundary supplies the region's edge.

The domain accent is auditory detection, frequency in hertz, sound level and its reference, air/bone/free-field delivery, ears and species, audiograms, behavioral conditioning, physiological responses, masking, and age-related change.

Psychophysical Scaling is related but not containing. A hearing range can be obtained from threshold measurements without fitting a stimulus-to-sensation magnitude law, and equal-loudness scaling concerns above-threshold comparison rather than mere detection.

Threshold is the strongest parent. At every frequency, a detection threshold separates presentations that meet the response criterion from those that do not; the collection of thresholds delimits the hearing region.

Measurement supplies the calibrated stimulus, response procedure, units, reference frame, and uncertainty record. It is essential machinery rather than the narrowest structural parent.

Boundary describes the lower contour of the audibility region and any level-specific frequency endpoints.

Psychophysical Scaling is a strong domain neighbor where detection or loudness data are fit across a stimulus continuum, but Hearing Range does not require a magnitude-scaling law.

The proposed DAG uses one strict composition/instantiation edge to Threshold. Other connections remain explanatory.

Relationships to Other Abstractions¶

Current abstraction Hearing Range Domain-specific

Parents (1) — more general patterns this builds on

Hearing Range is a kind of Threshold Prime

Threshold is the strongest parent.

Hierarchy path (1) — routes to 1 parentless root

Hearing Range → Threshold

Neighborhood in Abstraction Space¶

Hearing Range sits in a sparse region of the domain-specific corpus (89^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Harmonic Spectrum — 0.83
Validity Scale — 0.81
Linear Drumming — 0.80
Jhala — 0.79
Loudness — 0.78

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Absolute threshold of hearing: one frequency-specific threshold or a normative curve, not the entire condition-indexed range family.
Audiogram: the plotted record from which a tested range and sensitivity profile can be read.
Equal-loudness contour: combinations judged equally loud after detection.
Auditory dynamic range: the interval from threshold to an upper discomfort, saturation, or other declared boundary.
Speech-frequency range: frequencies important for speech, not all detectable frequencies.
Ultrasound and infrasound: frequency labels relative to conventional human hearing, not species-independent inaudibility.
Frequency response: an instrument or system output relation, not biological hearing by default.
Vocalization or echolocation range: produced signals rather than measured detection limits.
Hearing level (dB HL): a calibrated reference scale, not the hearing range itself.
Hearing loss classification: a clinical interpretation of thresholds against reference and functional criteria.
Psychophysical Scaling: the broader empirical mapping family for sensation and discrimination laws.
Weber's Law: a relation for difference thresholds relative to stimulus magnitude.

References¶

[1] International Organization for Standardization. Acoustics — Normal equal-loudness-level contours. ISO 226:2023, edition 3, 2023. ISO 226:2023, whose scope confines the equal-loudness-level contours to free-field frontal-incidence binaural listening by otologically normal persons aged 18 to 25 inclusive. registry ↩

[2] American Speech-Language-Hearing Association. Guidelines for Manual Pure-Tone Threshold Audiometry. ASHA Practice Policy (Guidelines), 2005. The ASHA manual pure-tone guideline, which sets the conventional air-conduction set at 250 through 8000 Hz, adds 125 Hz where a low-frequency loss exists, and reserves 9000 to 16000 Hz for extended high-frequency audiometry as a special-purpose procedure. The ASHA manual pure-tone guideline, which defines threshold as the lowest hearing level yielding responses on at least half of a series of ascending trials, with a minimum of two responses out of three presentations at one level. registry ↩a ↩b

[3] International Organization for Standardization. Acoustics — Audiometric test methods — Part 1: Pure-tone air and bone conduction audiometry. ISO (International Organization for Standardization), Geneva; ISO 8253-1:2010, 2^nd edition, 2010. ISO 8253-1:2010, whose scope covers pure-tone air- and bone-conduction threshold audiometry and expressly excludes procedures above threshold as well as speech and electrophysiological audiometry. registry ↩

[4] Heffner, Henry E. and Heffner, Rickye S. “Hearing Ranges of Laboratory Animals”. Journal of the American Association for Laboratory Animal Science, 2007. Heffner and Heffner's comparative review of laboratory-animal hearing ranges, which reports free-field thresholds shifting by as much as 21 dB - in reindeer - with pinna orientation toward or away from the loudspeaker. registry ↩

[5] Heffner and Heffner. “Hearing range of the domestic cat”. Hearing Research, 1985. Heffner and Heffner's behavioural audiograms of two domestic cats, which give a hearing range of 48 Hz to 85 kHz at 70 dB SPL - a level-conditioned slice, not an unqualified limit. registry ↩

[6] Heffner, Koay, and Heffner. “Hearing in American leaf-nosed bats. III: Artibeus jamaicensis”. Hearing Research, 2003. Heffner, Koay and Heffner's conditioned-avoidance audiogram of the Jamaican fruit bat, giving a range of 2.8 to 131 kHz at 60 dB SPL with average best sensitivity of 8.5 dB SPL at 16 kHz. registry ↩