FROM AGPEDIA — AGENCY THROUGH KNOWLEDGE

Psychoacoustics

Psychoacoustics is the scientific study of how people perceive sound: the relationship between physical sounds and the brain's interpretation of them.[1:1] It is a branch of psychophysics, which studies how sensations relate to the stimuli that cause them.[1:2] Quantities such as frequency, sound level and spectrum can be measured with instruments, whereas pitch, loudness and timbre are perceptions; psychoacoustics studies these perceptions and how they relate to the physical quantities.[2:1] The field draws on psychology, acoustics, electronic engineering, physics, biology, physiology and computer science.[2:2]

Hearing does not track the physical properties of sound in a simple way. Two sounds of equal physical intensity can differ greatly in loudness depending on their frequency;[3:1] the ear behaves as if it splits sound into overlapping frequency bands, the auditory filters;[4:1] one sound can make another inaudible, an effect called masking;[5:1] and listeners hear the pitch of a sound's fundamental frequency even when that frequency has been removed.[6:1] Such findings underlie perceptual audio coding, the basis of formats such as MP3 and AAC, which discards what the ear cannot hear,[7:1] and an international standard for measuring the loudness of broadcast programs.[8:1]

Some of the traditional concerns of psychoacoustics are the perception of pitch, loudness and timbre; more recent work deals with higher-level matters such as auditory scene analysis, how the auditory system groups sounds into separate sources and streams.[1:3] Until recently the field gave more attention to the peripheral auditory system (the ear and the early stages of the auditory nerve) than to cognitive processing.[1:4]

History

Gustav Theodor Fechner defined psychophysics in 1860, in Elemente der Psychophysik (Elements of Psychophysics), as an exact science of the functional relations between body and mind, or between the physical and the psychical world.[9:1] He named Weber's law after E. H. Weber, who had first stated it with some generality and supported it with experiments:[9:2] equal relative increases in a stimulus correspond to equal increases in sensation. Fechner regarded it as fundamental to the measurement of sensation.[9:3] The book already treated hearing: it reported experiments by Renz and Wolf, under Vierordt's direction, on the ear's sensitivity to differences in loudness, using the ticking of a watch placed at different distances from the ear.[9:4]

Hermann von Helmholtz's On the Sensations of Tone (German original 1863) set out to join physical and physiological acoustics with musical science and aesthetics.[10:1] Helmholtz regarded as proved a proposition of Georg Simon Ohm: that the ear perceives only sinusoidal ("pendular") vibrations as simple tones and resolves all other periodic sounds into a series of them.[10:2]

Lord Rayleigh set out in 1907 how the two ears locate sounds at low and high frequencies (see Spatial hearing).[11:1] Harvey Fletcher and W. A. Munson published loudness contours in 1933,[3:2] and the same paper first published the idea of the critical band.[12:1] S. S. Stevens proposed scales of subjective loudness (1936) and pitch (1937).[13:1][14:1] J. F. Schouten's 1940 papers on the "residue" offered an explanation for the pitch of sounds whose fundamental frequency is missing.[15:1]

Measuring perception

One basic measure is the difference limen (DL), or just-noticeable difference: the smallest change in a sound that a listener can detect. Stevens and Hallowell Davis defined a listener's sensitivity to changes in intensity as the reciprocal of the DL.[16:1]

Weber had proposed that the ratio of the DL to the intensity at which it is measured, the Weber fraction, is constant for any sense.[16:2] Stevens and Davis wrote in 1938 that "Weber was mistaken about its constancy" for hearing:[16:3] in the data they reviewed, the relative DL for a tone approached a constant above about 50 dB but grew rapidly as the sound neared the threshold of hearing.[16:4] Later work refined this. Magdalena Wojtczak and Neal Viemeister report that for broadband noise, Weber's law holds for the just-noticeable difference in intensity; for pure tones, thresholds improve as the level rises, a pattern known as the "near miss" to Weber's law, which they attribute to the nonlinear spread of excitation in the inner ear.[17:1][17:2]

Beyond thresholds, psychoacousticians have tried to measure the size of sensations directly. Stevens opened his 1936 paper on loudness by noting that a scale for "the subjective magnitude of sensation" had long been sought.[13:2]

Loudness

Fletcher and Munson defined loudness as "a psychological term used to describe the magnitude of an auditory sensation."[3:3] It does not depend on intensity alone: they gave the example of a flute and a bass drum placed so that they sound equally loud, where the drum's intensity at the ear would be many times the flute's.[3:1]

They defined the loudness level of a sound as the intensity level of a 1000 Hz reference tone judged equally loud, and drew contours joining tones of equal loudness level across frequency.[3:4][3:2] For tones from 300 to 4000 Hz, over much of the practical range, loudness level was close to the physical intensity level.[3:5] The unit of loudness level is the phon: the international standard ISO 226 defines a sound's loudness level in phons as the sound pressure level in decibels of a frontally arriving 1000 Hz tone judged as loud as it.[18:1]

A loudness level says which tone a sound matches, not how loud it is relative to another. Stevens proposed in 1936 a unit for a scale of loudness itself, the sone: the loudness of a 1000 Hz tone 40 dB above threshold, heard with both ears.[13:1] He had first suggested the name "phone", but dropped it because in Germany "phon" was already in use for a different unit.[13:3]

Equal-loudness contours

Curves of equal loudness across frequency, the equal-loudness contours, are described in ISO 226 as expressing "a fundamental property of the human auditory system."[18:2] The first international standard for these contours was based not on Fletcher and Munson's contours but on loudspeaker measurements made in the United Kingdom by Robinson and Dadson in the 1950s.[19:1] These contours were widely accepted from 1956.[20:1]

Errors in that standard were reported in 1985, and by about 1990 new data from Germany, Denmark and Japan differed from it by more than 10 dB below 1 kHz.[19:2][19:3] Yôiti Suzuki and Hisashi Takeshima combined the results of 12 studies from the mid-1980s on into a new set of contours. They found that the Robinson–Dadson contours differed from the new ones most below 500 Hz, often by as much as 14 dB, while Fletcher and Munson's contours were broadly similar to the new ones in the mid-frequency range up to 60 phons.[20:2][20:3] After 18 years of work, the standard was fully revised as ISO 226:2003.[19:2] It applies to otologically normal listeners aged 18 to 25.[18:3] The 2023 edition differs from the 2003 one by at most 0.6 dB.[19:4]

Frequency selectivity and masking

Masking is the process by which one sound is made inaudible by the presence of another.[5:1]

Critical bands and auditory filters

Fletcher, following Helmholtz, suggested that the peripheral auditory system behaves as if it contained a bank of bandpass filters with overlapping passbands, now called the auditory filters.[4:1] In the resulting power-spectrum model of masking, a tone is masked only by the components of a noise that pass through the filter centered on the tone.[4:2] In a classic experiment, the threshold of a tone in noise rises as the noise's bandwidth grows, up to about 100 to 200 Hz, and then stays constant: noise outside that band no longer contributes.[21:1]

Fletcher's 1940 paper "Auditory patterns" is the one usually cited for the critical band, but Jont Allen notes that it was a tutorial and review, first given as a dinner address in December 1938. The idea had first been published by Fletcher and Munson in 1933, and the 1940 paper was the first report of the band-widening experiment.[12:2][12:1][12:3] Allen also notes that the quantity Fletcher himself called the "critical band" was a level difference in decibels, which Allen calls the critical ratio.[12:4]

Two scales express frequency in units of these bands:

Scale Basis Notes
Bark scale The classical critical band Defined so that each critical band is one Bark wide; runs from 1 to 24 Bark, the first 24 critical bands.[22:1][22:2] Proposed by Zwicker and coworkers.[4:3]
Equivalent rectangular bandwidth (ERB) Auditory filter shapes measured with notched noise The ERB of the normal auditory filter is 24.7(0.00437F + 1) Hz at center frequency F, about 132 Hz at 1 kHz.[4:4] It is narrower than the classical critical band: about 11% of the center frequency at high frequencies, leveling off at about 25 Hz at low ones.[22:3]

The ERB comes from the notched-noise method, in which the threshold of a tone is measured in a noise with a gap (a spectral notch) around the tone's frequency, as the width of the gap changes.[23:1][22:4] Brian Glasberg and Brian Moore used such data in 1990 to propose new formulae for how the auditory filter varies with frequency and level.[23:2]

Masking over time

Masking also occurs between sounds that do not overlap in time. In forward masking a faint sound follows the masker; in backward masking it comes just before.[4:5] Forward masking decays to zero 100 to 200 ms after the masker ends, whatever its initial amount; practiced listeners often show little or no backward masking.[4:6]

Pitch

Ohm's proposition holds only in part. Schouten argued in 1940 that "Ohm's law of subjective sound analysis" applies to the lower harmonics of a periodic sound only: the higher harmonics cannot be heard separately as pure tones, but are heard together as one component.[15:2]

This bears on the missing fundamental. A sound is heard with the pitch of its fundamental frequency even when the fundamental itself is absent.[6:1] In Schouten's demonstration, removing the 200 Hz fundamental from a periodic pulse left the pitch of 200 Hz unchanged.[6:2] He called the extra perceived component a "residue", and argued that its pitch follows the periodicity of the combined waveform of the harmonics that make it up.[6:3][6:4] In Moore's account, filtering out the fundamental leaves the pitch unchanged, with only a slight change in timbre.[4:7]

Stevens, J. Volkmann and E. B. Newman built a scale of pitch in 1937 from listeners' settings of the tone that sounded half as high as another. They assigned the 1000 Hz tone a pitch of 1000 units, called mels.[14:1][14:2] Measured on this scale, musical intervals such as octaves grew larger as their frequency rose, except in the two highest audible octaves.[14:3]

Consonance and roughness

R. Plomp and W. J. M. Levelt reviewed explanations of musical consonance in 1965 and found support for Helmholtz's hypothesis that the difference between consonant and dissonant intervals is related to beats between neighboring partials (the component frequencies of a sound).[24:1] In their experiments with pairs of pure tones, intervals were judged consonant when the two frequencies were more than a critical bandwidth apart, and most dissonant at a separation of about a quarter of a critical bandwidth.[24:2] The separation giving the most roughness was not a constant number of hertz, as Helmholtz had supposed, but a constant fraction of the critical bandwidth.[24:3] They concluded that the special status of intervals with simple frequency ratios is due to interference between adjacent partials, heard as roughness.[24:4] An analysis of chords by Bach and Dvořák suggested to them that critical bandwidth plays a part in actual music.[24:5]

Spatial hearing

Sound localization relies largely on differences between the signals at the two ears. In experiments reported in 1907, Rayleigh found that listeners keeping their heads still could tell left from right but could not say whether a pure tone came from in front or behind.[11:2] He noted that a difference of intensity between the ears easily explains left-right judgments of high sounds, but that this explanation becomes less adequate as pitch falls: at 128 Hz the difference in intensity was barely perceptible, yet left and right were as easy to tell apart as before.[11:3] He concluded that at high frequencies the judgment is based on differences in intensity, and at low frequencies on differences of phase (timing) between the ears.[11:1][11:4] This account is known as the duplex theory of binaural hearing.[25:1]

In modern terms, the cues are the interaural time difference (ITD) and the interaural level difference (ILD). A 2024 review by Alessandro Carlini and colleagues reports that the smallest perceptible differences are about 10 microseconds for ITD and about 1 dB for ILD; ITD works best below about 1500 Hz and ILD above about 4000 Hz, and localization accuracy is poorest between about 1000 and 3000 Hz.[25:2][25:3] Reflections from the shoulders, head and outer ear filter the sound reaching each eardrum in a way that depends on the source's position, described by head-related transfer functions (HRTFs). Judgments of elevation rely mainly on these spectral cues from the outer ear, and need sound energy between about 4000 and 16,000 Hz.[25:4][25:5]

In rooms, reflections reach the ears shortly after the direct sound. The precedence effect is the perceptual dominance of the spatial information carried by the first-arriving sound.[26:1] The term was introduced by Wallach and colleagues in 1949; similar phenomena were described by Cremer (1948) and Haas (1949, 1951), as the "law of the first wavefront" and the "Haas effect".[26:2] When a sound and a delayed copy arrive from different directions with a short enough delay, listeners hear one sound, locate it at or near the first source, and are insensitive to features of the second.[26:3] At delays under about 1 ms they hear a single image between the two sources, called summing localization.[26:4] The delay at which the echo becomes audible as a separate sound can grow from 5–10 ms to 15–30 ms or more after the sound pair is repeated several times, a phenomenon called buildup.[26:5]

Applications

Perceptual audio coding

Perceptual audio coders compress sound by removing what listeners cannot hear. In 1988 James Johnston of AT&T Bell Laboratories described a transform coder that, rather than modeling the sound source, used a model of human hearing to compute a short-term masking curve from the signal's spectrum.[27:1] The resulting threshold estimates the most noise that can be inserted into the signal without being heard, so a coder can place its quantization noise (the error from storing the signal with fewer bits) beneath it.[27:2] In a subjective listening test, his coder at 128 kbit/s could not be distinguished from the original.[27:3] Johnston named the lower bound on bit rate for such transparent coding perceptual entropy, estimated at about 2 bits per sample;[27:4] a 2000 review by Ted Painter and Andreas Spanias cites measurements suggesting about 2.1 bits per sample for a wide variety of CD-quality material.[5:2]

Karlheinz Brandenburg of the Fraunhofer Institute describes perceptual encoding as a lossy technique that uses knowledge from psychoacoustics to make compression inaudible.[7:1] The encoder's perceptual model estimates a masking threshold that varies with time and frequency, and the spectral components are quantized so as to keep the noise below it.[7:2] In MPEG-1 Layer-3 (MP3), defined in 1991, the coder's frequency partitions roughly match the critical bands; MPEG-2 Advanced Audio Coding was developed as its successor.[7:3][7:4]

Loudness measurement

Recommendation ITU-R BS.1770, published by the International Telecommunication Union, sets out a single algorithm for estimating the loudness of broadcast programs, on the grounds that listeners want loudness to be uniform across sources and program types and that loudness control in program exchange needs one recommended algorithm.[8:1] The algorithm weights the signal's frequencies (K-weighting), averages its power in each channel, sums the channels with weights, and ignores quiet passages by gating 400 ms blocks against an absolute and a relative threshold.[8:2] The first stage of the K-weighting filter models the acoustic effect of the listener's head as a rigid sphere.[8:3] Results are given in LKFS (loudness, K-weighted, relative to full scale), a unit that behaves like a decibel.[8:4] The recommendation notes that measured loudness is an estimate of subjective loudness, with some uncertainty depending on listeners, material and listening conditions.[8:5]

  1. ^ ↗ definition ^ ↗ branch-of-psychophysics ^ ↗ traditional-and-contemporary ^ ↗ peripheral-emphasis Psychoacoustics. Handbook for Acoustic Ecology. Simon Fraser University. https://www.sfu.ca/sonic-studio-webdav/cmns/Handbook5/handbook/Psychoacoustics.html.
  2. ^ ↗ subjective-vs-objective ^ ↗ interdisciplinary Zhang, Peter Xinya (2008). Psychoacoustics. Handbook for Sound Engineers. Routledge. https://doi.org/10.4324/9780080927619-10 https://www.taylorfrancis.com/books/9780080927619/chapters/10.4324/9780080927619-10.
  3. ^a ^b ↗ not-intensity-alone ^a ^b ↗ contours ^ ↗ loudness-definition ^ ↗ loudness-level-definition ^ ↗ midrange Fletcher, Harvey; Munson, W. A. (1933-10). Loudness, Its Definition, Measurement and Calculation. The Bell System Technical Journal. https://doi.org/10.1002/j.1538-7305.1933.tb00403.x https://archive.org/details/bstj12-4-377.
  4. ^a ^b ↗ auditory-filter-bank ^ ↗ power-spectrum-model ^ ↗ erb-number-scale ^ ↗ erb-formula ^ ↗ forward-backward-masking ^ ↗ forward-masking-decay ^ ↗ missing-fundamental Moore, Brian C. J. (2007). Psychoacoustics. Springer Handbook of Acoustics. Springer, New York. https://doi.org/10.1007/978-0-387-30425-0_13 https://math.ucdavis.edu/~saito/data/auditory/moore_psychoacoustics.pdf.
  5. ^a ^b ↗ masking-definition ^ ↗ perceptual-entropy Painter, Ted; Spanias, Andreas (2000-04). Perceptual coding of digital audio. Proceedings of the IEEE. https://doi.org/10.1109/5.842996 https://www.ee.columbia.edu/~dpwe/e6820/papers/PaintS00-pcpcod.pdf.
  6. ^a ^b ↗ missing-fundamental ^ ↗ fundamental-cancelled ^ ↗ residue-named ^ ↗ pitch-from-periodicity Schouten, J. F. (1940). The residue, a new component in subjective sound analysis. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen. https://dwc.knaw.nl/DL/publications/PU00017418.pdf.
  7. ^a ^b ↗ perceptual-encoding ^ ↗ masking-threshold ^ ↗ partitions-critical-bands ^ ↗ aac-successor Brandenburg, Karlheinz (1999-09). MP3 and AAC explained. Proceedings of the AES 17th International Conference: High-Quality Audio Coding. Audio Engineering Society. https://aes2.org/publications/elibrary-page/?id=8079.
  8. ^a ^b ↗ purpose ^ ↗ four-stages ^ ↗ head-model ^ ↗ lkfs-unit ^ ↗ estimate-only International Telecommunication Union, Radiocommunication Sector (2023-11). Recommendation ITU-R BS.1770-5: Algorithms to measure audio programme loudness and true-peak audio level. ITU, Geneva. https://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1770-5-202311-I!!PDF-E.pdf.
  9. ^ ↗ defines-psychophysics ^ ↗ names-webers-law ^ ↗ webers-law-statement ^ ↗ renz-wolf-watch Fechner, Gustav Theodor (1860). Elemente der Psychophysik. Erster Theil. Breitkopf und Härtel, Leipzig. https://archive.org/details/elementederpsych001fech.
  10. ^ ↗ joins-acoustics-and-music ^ ↗ ohms-acoustic-law Helmholtz, Hermann L. F. (1895). On the Sensations of Tone as a Physiological Basis for the Theory of Music. Longmans, Green, and Co., London and New York. https://archive.org/details/onsensationsofto00helmrich.
  11. ^a ^b ↗ duplex-statement ^ ↗ front-back ^ ↗ intensity-fails-low ^ ↗ phase-fails Strutt, John William, 3rd Baron Rayleigh (1907-02). On our perception of sound direction. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science. https://doi.org/10.1080/14786440709463595 https://bsd.neuroinf.jp/w/images/4/4d/Rayleigh_1907.pdf.
  12. ^a ^b ↗ critical-band-first-published-1933 ^ ↗ 1940-paper-review ^ ↗ band-widening-first-reported ^ ↗ critical-ratio Allen, Jont B. (1996-04). Harvey Fletcher’s role in the creation of communication acoustics. The Journal of the Acoustical Society of America. https://doi.org/10.1121/1.415364 https://jontalle.web.engr.illinois.edu/Public/Allen-pdf/Allen96_FletcherComm.pdf.
  13. ^a ^b ↗ sone-proposed ^ ↗ scale-sought ^ ↗ phone-name-dropped Stevens, S. S. (1936-09). A scale for the measurement of a psychological magnitude: loudness. Psychological Review. https://doi.org/10.1037/h0058773 https://archive.org/details/sim_psychological-review_1936-09_43_5.
  14. ^a ^b ↗ half-pitch-method ^ ↗ 1000-mels ^ ↗ octaves-grow Stevens, S. S.; Volkmann, J.; Newman, E. B. (1937-01). A scale for the measurement of the psychological magnitude pitch. The Journal of the Acoustical Society of America. https://doi.org/10.1121/1.1915893.
  15. ^ ↗ explains-missing-fundamental ^ ↗ ohms-law-lower-harmonics Schouten, J. F. (1940). The residue and the mechanism of hearing. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen. https://dwc.knaw.nl/DL/publications/PU00017508.pdf.
  16. ^ ↗ dl-defined ^ ↗ weber-fraction ^ ↗ weber-mistaken ^ ↗ relative-dl-above-50-db Stevens, S. S.; Davis, Hallowell (1938). Hearing: Its Psychology and Physiology. https://archive.org/details/in.ernet.dli.2015.145142.
  17. ^ ↗ holds-for-noise ^ ↗ near-miss Wojtczak, Magdalena; Viemeister, Neal F. (2008-04). Perception of suprathreshold amplitude modulation and intensity increments: Weber’s law revisited. The Journal of the Acoustical Society of America. https://doi.org/10.1121/1.2839889 https://pmc.ncbi.nlm.nih.gov/articles/PMC2394195/.
  18. ^ ↗ phon-definition ^ ↗ contours-fundamental ^ ↗ listeners International Organization for Standardization (2003-08-15). Acoustics — Normal equal-loudness-level contours. ISO, Geneva. https://www.iso.org/standard/34222.html.
  19. ^ ↗ based-on-robinson-dadson ^a ^b ↗ revised-2003 ^ ↗ off-by-10-db ^ ↗ 2023-practically-same Suzuki, Yôiti; Takeshima, Hisashi; Kurakata, Kenji (2024). Revision of ISO 226 “Normal Equal-Loudness-Level Contours” from 2003 to 2023 edition: The background and results. Acoustical Science and Technology. https://doi.org/10.1250/ast.e23.66 https://www.jstage.jst.go.jp/article/ast/45/1/45_e23.66/_pdf/-char/en.
  20. ^ ↗ robinson-dadson-accepted ^ ↗ twelve-studies ^ ↗ differences Suzuki, Yôiti; Takeshima, Hisashi (2004-08). Equal-loudness-level contours for pure tones. The Journal of the Acoustical Society of America. https://doi.org/10.1121/1.1763601 https://pubmed.ncbi.nlm.nih.gov/15376658/.
  21. ^ ↗ band-widening Moore, Brian C. J. (2008). Basic auditory processes involved in the analysis of speech sounds. Philosophical Transactions of the Royal Society B: Biological Sciences. https://doi.org/10.1098/rstb.2007.2152 https://pmc.ncbi.nlm.nih.gov/articles/PMC2606789/.
  22. ^ ↗ bark-definition ^ ↗ bark-24 ^ ↗ erb-narrower ^ ↗ erb-notched-noise Smith, Julius O., III; Abel, Jonathan S. (1999). Bark and ERB bilinear transforms. IEEE Transactions on Speech and Audio Processing. https://doi.org/10.1109/89.799695 https://ccrma.stanford.edu/~jos/bbt/bbt.pdf.
  23. ^ ↗ notched-noise-method ^ ↗ new-formulae Glasberg, Brian R.; Moore, Brian C. J. (1990-08). Derivation of auditory filter shapes from notched-noise data. Hearing Research. https://doi.org/10.1016/0378-5955(90)90170-T https://pubmed.ncbi.nlm.nih.gov/2228789/.
  24. ^ ↗ helmholtz-beats ^ ↗ critical-bandwidth ^ ↗ proportional-to-critical-band ^ ↗ small-integer-ratios ^ ↗ role-in-music Plomp, R.; Levelt, W. J. M. (1965-10). Tonal consonance and critical bandwidth. The Journal of the Acoustical Society of America. https://doi.org/10.1121/1.1909741 https://pure.mpg.de/rest/items/item_66382_5/component/file_468040/content.
  25. ^ ↗ duplex-theory ^ ↗ thresholds ^ ↗ frequency-ranges ^ ↗ hrtf ^ ↗ elevation Carlini, Alessandro; Bordeau, Camille; Ambard, Maxime (2024-07-10). Auditory localization: a comprehensive practical review. Frontiers in Psychology. https://doi.org/10.3389/fpsyg.2024.1408073 https://pmc.ncbi.nlm.nih.gov/articles/PMC11267622/.
  26. ^ ↗ definition ^ ↗ named ^ ↗ three-components ^ ↗ summing-localization ^ ↗ buildup Brown, Andrew D.; Stecker, G. Christopher; Tollin, Daniel J. (2015-02). The Precedence Effect in Sound Localization. Journal of the Association for Research in Otolaryngology. https://doi.org/10.1007/s10162-014-0496-2 https://pmc.ncbi.nlm.nih.gov/articles/PMC4310855/.
  27. ^ ↗ auditory-model ^ ↗ max-inaudible-noise ^ ↗ transparent-128k ^ ↗ perceptual-entropy Johnston, James D. (1988-02). Transform coding of audio signals using perceptual noise criteria. IEEE Journal on Selected Areas in Communications. https://doi.org/10.1109/49.608 https://www.ee.columbia.edu/~dpwe/papers/Johns88-audiocoding.pdf.

Wikidata: Q557399

Available in