ISO 16 has fixed concert pitch at A=440 Hz since 1975, yet when 3,336 classical recordings were measured, the average tuning came out at 442.38 Hz and only 50% sat between 440 and 443 Hz (Lerch, ISMIR 2006). Concert pitch is the agreed frequency of the A above middle C, the note an orchestra tunes to before it plays; hertz (Hz) simply counts vibrations per second. The most cited controlled trial behind today’s A=432 Hz claims enrolled 33 volunteers and reported a heart-rate change with a p-value of exactly 0.05 (Calamassi and Pomponi, Explore 2019). Meanwhile, listeners with top-level absolute pitch failed to notice a symphony sliding 33 cents flat (Hedger, Heald and Nusbaum, Psychological Science 2013). We aggregated data from the ISO 16:1975 standard record, a history of standard pitch built on Alexander Ellis’s measurements, peer-reviewed trials in Explore, Acta Biomedica, BMC Complementary Medicine and Therapies and Scientific Reports, absolute-pitch studies in Music Perception and Psychological Science, and ISMIR proceedings on tuning in recordings. If you arrived from our musical instruments statistics, this roundup zooms in on the one number every pitched instrument shares: the reference A.
TL;DR
- Only 50% of 3,336 classical tracks were tuned between 440 and 443 Hz, and the average was 442.38 Hz (Lerch, ISMIR 2006).
- ISO 16:1975 sets A at 440 Hz, requires tuning instruments accurate within 0.5 Hz, and was last confirmed on September 30, 2022 (ISO 16:1975 catalogue record).
- A=440 Hz was agreed at an international conference in London in May 1939, eight decades after France made A=435 Hz law in 1859 (Lynn Cavanagh, history of international standard pitch).
- In the 1830s Johann Scheibler measured the average concert A of Viennese concert pianos at 440 Hz (Whipple Museum, University of Cambridge).
- A=432 Hz sits about 31.8 cents below A=440 Hz, roughly a third of a semitone (formula from Dressler and Streich, ISMIR 2007).
- The best-known A=432 pilot had 33 volunteers and found a 4.79 bpm heart-rate drop versus 440 Hz, p = 0.05 (Calamassi and Pomponi, Explore 2019).
- In 72 root canal patients, blood pressure and heart rate changes with 432 Hz music were not significant against control, p > 0.05 (Mousavi et al., Clinical and Experimental Dental Research 2025).
- 22 male athletes ran faster after warm-up music at 440 Hz than at 432 Hz (mean speed p < 0.001, d = 0.75) (Jebabli et al., Scientific Reports 2025).
- 13 listeners with absolute pitch did not notice music detuned by 33 cents at about 2 cents per minute (Hedger, Heald and Nusbaum, Psychological Science 2013).
- Accurate absolute pitch reached 57% of students at a Kyoto conservatory versus 12% in Warsaw (Miyazaki et al., Music Perception 2018).
- One neural vocoder moved recordings’ tuning by up to 18.7 cents on average (Berendes, Maman and Mueller, ISMIR 2025).
1. Before the Standard: Four Centuries of Wandering Pitch
German organ pitch before 1600 ran from A=377 Hz to A=567 Hz, a spread of 190 Hz. Expressed musically, 567 / 377 = 1.504, and 1200 x log2(1.504) is about 707 cents, slightly more than a perfect fifth (700 cents). That is our own arithmetic on the published endpoints, and it shows why “the same note” once meant very different sounds from one church to the next.
The numbers also explain why singers became the loudest voices for a standard. When pitch crept from Mozart’s A=422 Hz toward A=450 Hz in some 19th-century houses, the written notes stayed the same while the physical demand on the voice rose by more than a semitone; our vocal range and singing statistics show how narrow comfortable ranges can be. The 440 Hz figure itself is older than the standard: a textile industrialist named Johann Scheibler had already found it as the average of Vienna’s concert pianos in the 1830s, as documented by the Whipple Museum of the History of Science.
| Metric | Value | Source |
|---|---|---|
| Range of organ pitch in Germany before 1600 (lowest around 1511, highest for the first simple pipe organs of the Middle Ages) | A=377 Hz to A=567 Hz | Lynn Cavanagh, citing Alexander J. Ellis (Studies in the History of Music Pitch, 1968 edition) |
| Pitch favoured by Handel and by Mozart, expressed in modern terms | A=423 Hz (Handel) and A=422 Hz (Mozart) | Lynn Cavanagh, citing Alexander J. Ellis |
| Pitch reached in some 19th-century European opera houses and concert halls | A=450 Hz | Lynn Cavanagh, history of international standard pitch |
| Average concert A of Viennese concert pianos, measured by Johann Scheibler in the 1830s | 440 Hz | Whipple Museum, University of Cambridge, Brief Chronicle of the Tuning Fork |
| Year the German academy of sciences endorsed Scheibler’s pitch (the Stuttgart pitch) | 1834 | Whipple Museum, University of Cambridge |
| Year the tuning fork was invented, by John Shore | 1711 | Whipple Museum, University of Cambridge |
| Typical deviation of the tuning frequency from 440 Hz around 1750, around 1850 and around 1950 | -50 to +30 Hz; -20 to +20 Hz; -5 to +10 Hz | Lerch, ISMIR 2006 (citing Briner, Musikinstrumentenfuehrer 1998) |
Context: the 1750, 1850 and 1950 ranges are typical bands from a reference handbook, not measurements of specific instruments. Most recent available data for the pre-1900 values: Ellis’s monographs as reprinted in 1968.
2. From Paris 1859 to ISO 16: How A=440 Became the Rule
France made A=435 Hz law in 1859, Britain recognized A=439 Hz in 1896, and only a May 1939 conference in London settled on A=440 Hz. Each step was a compromise between singers who wanted lower pitch and audiences used to the brilliance of higher pitch, according to this history of international standard pitch, which draws on a 1949 lecture to the Royal Society of Arts by a British Standards Institution representative.
The British 439 Hz figure shows how arbitrary the path was. The Philharmonic Society’s adviser assumed the French 435 Hz had been defined at 59 degrees Fahrenheit and corrected it to a 68-degree room: 435 + ((68 - 59) / 1000 x 435) = 438.915 Hz, rounded to 439 Hz. The BBC then needed a note it could generate electronically, and a crystal-divided signal produced 440 Hz cleanly where 439 Hz, a prime number, could not be. Broadcasting, not acoustics, gave the final push.
| Metric | Value | Source |
|---|---|---|
| Pitch made law in France by a government commission | A=435 Hz (1859) | Lynn Cavanagh, history of international standard pitch |
| British pitch after the Philharmonic Society’s room-temperature correction | 438.915 Hz, recognized as A=439 Hz in 1896 | Lynn Cavanagh, citing Lloyd, Journal of the Royal Society of Arts 1949 |
| International conference that set A=440 Hz as the standard | London, May 1939 | Lynn Cavanagh, citing Lloyd 1949 |
| Dates on which ISO reaffirmed A=440 Hz | November 1955 and January 1975 | Lynn Cavanagh, citing The New Grove Dictionary 1980 |
| Frequency ISO 16:1975 specifies for the note A in the treble stave | 440 Hz | ISO 16:1975 catalogue record |
| Accuracy required of instruments used for tuning and retuning under ISO 16:1975 | within 0.5 Hz | ISO 16:1975 catalogue record |
| Latest confirmation of ISO 16:1975 by ISO technical committee 43 (Acoustics) | September 30, 2022 | ISO 16:1975 listing via Genorma |
| Year the Stuttgart-based standard became global, according to the Whipple Museum | 1926 | Whipple Museum, University of Cambridge |
Context: the 0.5 Hz tolerance applies to the tuning device (a fork or electronic generator), not to the pitch an orchestra actually plays, which is the subject of section 3. The ISO 16:1975 catalogue listing shows the standard is a one-page document per language version.
3. What Orchestras and Classical Recordings Actually Tune To
The working standard of classical music is a band just above 440 Hz, not 440 Hz itself: the average across 3,336 classical tracks was 442.38 Hz, with a standard deviation of 2.75 Hz. A cent is one hundredth of a semitone (the step between two neighbouring piano keys), so the 443 Hz used in Berlin and Vienna sits 11.76 cents above the ISO value. The practical consequence lands on fixed-pitch instruments such as organs and pianos, which cannot bend to meet the players.
The drift is not new. In the late 1940s, a musicologist tuning an electronic organ for New York orchestras found it went flat in performance at 440 Hz and only matched the players at 444 Hz. This 2006 tuning study from the ISMIR proceedings also notes that for both Berlin and Vienna, the tuning frequency was higher in previous decades, so the upward creep has partly reversed.
| Metric | Value | Source |
|---|---|---|
| Tuning used by the Chicago Symphony Orchestra and New York Philharmonic, and by the Berlin and Vienna Philharmonic (Berlin and Vienna figures from orchestra archivists, 2006) | 442 Hz (7.85 cents above 440 Hz) and 443 Hz (11.76 cents above 440 Hz) | Lerch, ISMIR 2006 |
| Pitch New York orchestra musicians habitually played at in the late 1940s and early 1950s | about A=444 Hz | Arthur Mendel, quoted in Lynn Cavanagh’s history |
| Average detected tuning across 3,336 classical tracks (about 300 CDs, about 291 hours) | 442.38 Hz (standard deviation 2.75 Hz) | Lerch, ISMIR 2006 |
| Share of tracks at the single most common value, 440 Hz | about 21% | Lerch, ISMIR 2006 |
| Share of tracks tuned between 440 and 443 Hz, between 439 and 448 Hz, and below 439 Hz | 50%, 95% and about 3.3% | Lerch, ISMIR 2006 |
| Tuning distribution of 352 Beethoven piano sonata recordings made between 1935 and 2022 | peaks around 440 Hz with a slight tendency higher, some as low as 430 Hz | Berendes, Maman and Mueller, ISMIR 2025 |
Outlier note: the 2006 database is a private collection with many recordings from the end of the twentieth century; most recent available data for a large classical tuning distribution is this 2006 study, still cited as the reference figure in 2025.
4. A=432 Under the Microscope: What the Controlled Trials Found
The difference at stake is about 31.8 cents: 1200 x log2(432 / 440) = -31.8, using the standard cents formula from tuning research (Dressler and Streich, ISMIR 2007). That is about a third of the step between two neighbouring piano keys. The trials below compare the same music at both tunings, often “double-blind” (neither listener nor researcher knows which version is playing) and “crossover” (each person hears both versions, in a random order).
Read as a body of evidence, the results are thin rather than damning or convincing. Samples range from 12 to 72 people, several studies did not include a 440 Hz arm at all, and in a 2022 trial with emergency staff the comparison group, given a free activity instead of music, became calmer too. The pilot that started much of the conversation, published in Explore in 2019, reported its heart-rate result at p = 0.05, the exact edge of conventional significance, while two sports trials found the opposite direction for performance: 440 Hz warm-up music beat 432 Hz.
| Metric | Value | Source |
|---|---|---|
| Change in mean heart rate with 432 Hz vs 440 Hz music, 33 volunteers, two 20-minute sessions (blood pressure decrease not significant) | -4.79 bpm (p = 0.05) | Calamassi and Pomponi, Explore 2019 |
| Change in sleep score after daily 30-minute listening, 12 patients with spinal cord injury | +3.6 at 432 Hz (p=0.02) vs -1.50 at 440 Hz (p=0.34) | Calamassi et al., Acta Biomedica 2020 |
| Median anxiety score change in the comparison group given a free activity instead of music, 54 emergency healthcare providers | 33 vs 31 (p=0.028) | Calamassi et al., Acta Biomedica 2022 |
| Change in respiratory rate and systolic blood pressure in the 432 Hz group of the same trial | -2.714 breaths per minute (p=0.000) and -3.821 mmHg (p=0.031) | Calamassi et al., Acta Biomedica 2022 |
| Median heart-rate reduction, 43 cancer patients, 15-minute sound sessions at 432 Hz vs 443 Hz (no 440 Hz arm) | -3 bpm vs -1 bpm (p = 0.04) | Hohneck et al., BMC Complementary Medicine and Therapies 2025 |
| Blood pressure and heart rate change with 432 Hz music vs control, 72 adult root canal patients | not significant (p > 0.05) | Mousavi et al., Clinical and Experimental Dental Research 2025 |
| Mean running speed in a 3-minute all-out test after warm-up music at 440 Hz vs 432 Hz, 22 male athletes | higher at 440 Hz (p < 0.001, d = 0.75) | Jebabli et al., Scientific Reports 2025 |
| Kickboxing speed test after warm-up music, 15 men and 13 women | best with 440 Hz music, then 432 Hz, both above no music | Jebabli et al., PeerJ 2025 |
Context: two further 2025 to 2026 trials compared 432 Hz with other sounds rather than with 440 Hz. In 99 root canal patients (33 per group), both 432 Hz music and binaural beats reduced anxiety against white noise (p<0.05) (Zaki et al., JPMA 2025), and in 162 students given 20 minutes of music, the salivary marker GRP78 was elevated in the 432 Hz group (0.51 [0.47-0.58]) (Bozok et al., Brain and Behavior 2026). Neither design can show that 432 Hz beats 440 Hz. These are research findings, not treatment advice.
5. Absolute Pitch Listeners: How Fixed Is a Fixed Reference?
Thirteen listeners with the highest grade of absolute pitch heard a Brahms symphony drift 33 cents flat, about the size of the 432 versus 440 gap, and not one reported noticing. Absolute pitch (AP) is the ability to name a note without hearing a reference note first. The detuning ran at about 2 cents per minute, well below the just-noticeable difference (the smallest pitch change listeners can detect), and afterwards the listeners rated the flat notes as more in tune than before, according to this University of Chicago study in Psychological Science.
That finding cuts both ways for the A=432 debate. It means the 440 Hz convention is partly held in place by what people hear every day, as the authors put it: “If each musical experience used something other than the ‘A = 440 Hz’ standard, AP note representations might change constantly.” It also means a gradual retuning is not reliably detected even by the listeners best equipped to detect it. Prevalence matters too, and so does adaptation: in this cross-cultural study of 1,016 music students, a key shifted a quarter-tone off standard pitch had only a limited effect on AP students’ interval judgments, which the authors suggest reflects listeners mentally adjusting.
| Metric | Value | Source |
|---|---|---|
| Typical stability of note representations in people with absolute pitch | about 5 cents within a day and 33 cents between days | Hedger, Heald and Nusbaum, Psychological Science 2013 (citing Wynn 1972 and others) |
| Listeners exposed to a symphony detuned at about 2 cents per minute until 33 cents flat, and number who reported noticing | 13 participants (7 female, 6 male), none noticed | Hedger, Heald and Nusbaum, Psychological Science 2013 |
| Mean in-tune rating of 33-cents-flat violin notes before vs after listening (scale 1 to 3) | 2.08 vs 2.31; interaction F(2, 24) = 6.57, p < .01 | Hedger, Heald and Nusbaum, Psychological Science 2013 |
| Second experiment with five-note phase music (about 45 minutes detuned): shift for notes never heard detuned | shift also appeared for unheard notes; interaction F(2, 26) = 7.87, p < .01, 14 participants | Hedger, Heald and Nusbaum, Psychological Science 2013 |
| Share of students with accurate absolute pitch (score 0.9 or above), Kyoto City University of Arts vs Chopin University of Music in Warsaw | 57% vs 12% | Miyazaki et al., Music Perception 2018 |
| Share of University of Minnesota music students at the non-absolute-pitch level | 98% | Miyazaki et al., Music Perception 2018 |
| Subgroups whose interval scores fell significantly when the key was shifted a quarter-tone off standard pitch | 1 subgroup (t(14) = -2.78, p = .02) | Miyazaki et al., Music Perception 2018 |
| Mean note-naming accuracy of absolute pitch vs non-absolute pitch participants across 160 studies (23,221 participants) | 85.9% vs 17.0% | Bairnsfather et al., Behavior Research Methods 2025 |
Context: the same 2025 review found that accuracy thresholds for classifying someone as an AP possessor ranged from 20 to 100% across studies, so prevalence figures are only comparable when the test and threshold match (Bairnsfather et al., Behavior Research Methods 2025).
6. Tuning in Pop Recordings, Vocals and Audio AI
A neural vocoder, the component of an AI music or voice system that turns a spectrogram (a picture of sound energy over time and frequency) back into audio, shifted recordings’ tuning by a mean absolute deviation of up to 18.7 cents. The same 2025 study found the outputs of two widely used vocoders clustering around 440 Hz with the peak slightly above it, which the authors suggest could reflect training data in which many recordings are tuned between 440 and 443 Hz. In other words, the real-world tuning band of recorded music appears to be learned and reproduced by machines, as shown in the ISMIR 2025 tuning bias paper from AudioLabs Erlangen.
Pop is harder to measure than orchestral music because the lead voice bends pitch constantly. In the ISMIR 2007 circular-statistics study by Dressler and Streich, three estimation methods placed the same pop track anywhere from 436.7 to 438.9 Hz. For classification systems, tuning becomes a hidden variable: a genre classifier lost accuracy whenever audio was shifted by fractions of a semitone, and the effect was weaker in genres played on acoustic instruments, which the authors link to synthesizers usually defaulting to 440 Hz, per the ICASSP 2019 paper by Qin and Lerch. Anyone working with an AI voice generator or synthetic singing, the topic of our singing voice synthesis pitch accuracy statistics, inherits these reference-pitch assumptions.
| Metric | Value | Source |
|---|---|---|
| Tuning estimates for one pop track (pop2) from three methods, ISMIR 2004 test set of 20 pieces | 438.9 Hz, 437.3 Hz and 436.7 Hz | Dressler and Streich, ISMIR 2007 |
| Tuning estimates for one opera recording with strong vibrato (opera5) from three methods | 443.9 Hz, 442.8 Hz and 446.6 Hz | Dressler and Streich, ISMIR 2007 |
| Best genre-classification accuracy at the original pitch on GTZAN (1,000 clips) and FMA-small (8,000 clips) | about 88.0% and 61.9% | Qin and Lerch, ICASSP 2019 |
| Variability of the accuracy decrease as audio was pitch-shifted within plus or minus 5 semitones | up to approx. 10% | Qin and Lerch, ICASSP 2019 |
| Highest mean absolute tuning deviation introduced by a tested neural vocoder (HAWT) | 18.7 cents | Berendes, Maman and Mueller, ISMIR 2025 |
| Violin recordings kept after two tuning estimators agreed within 5 cents | 651 of 925 | Berendes, Maman and Mueller, ISMIR 2025 |
| Listening test participants, and participants kept after screening | 25 and 20 | Berendes, Maman and Mueller, ISMIR 2025 |
Context: we found no peer-reviewed study measuring the tuning of a large commercial pop catalogue against 440 Hz; the pop figures above come from small test sets and show estimation spread as much as tuning spread (see the Data watch below).
Summary: Concert Pitch A=440 vs A=432 Tuning Standard by the Numbers
| Metric | Value | Source |
|---|---|---|
| Frequency ISO 16:1975 specifies for the A in the treble stave | 440 Hz, tuning instruments accurate within 0.5 Hz | ISO 16:1975 catalogue record |
| Latest confirmation of ISO 16:1975 | September 30, 2022 | ISO 16:1975 listing via Genorma |
| International conference that set A=440 Hz | London, May 1939 | Lynn Cavanagh, history of international standard pitch |
| Average Viennese concert-piano A measured in the 1830s | 440 Hz | Whipple Museum, University of Cambridge |
| Average tuning and share between 440 and 443 Hz across 3,336 classical tracks | 442.38 Hz; 50% | Lerch, ISMIR 2006 |
| Tuning of the Berlin and Vienna Philharmonic (archivists, 2006) | 443 Hz | Lerch, ISMIR 2006 |
| Gap between A=432 Hz and A=440 Hz | about 31.8 cents | Formula from Dressler and Streich, ISMIR 2007 |
| Heart-rate change with 432 Hz vs 440 Hz music, 33 volunteers | -4.79 bpm (p = 0.05) | Calamassi and Pomponi, Explore 2019 |
| Median heart-rate reduction at 432 Hz vs 443 Hz, 43 cancer patients | -3 bpm vs -1 bpm (p = 0.04) | Hohneck et al., BMC Complementary Medicine and Therapies 2025 |
| Blood pressure and heart rate with 432 Hz music vs control, 72 patients | not significant (p > 0.05) | Mousavi et al., Clinical and Experimental Dental Research 2025 |
| Running speed after warm-up music at 440 Hz vs 432 Hz, 22 athletes | higher at 440 Hz (p < 0.001, d = 0.75) | Jebabli et al., Scientific Reports 2025 |
| Absolute-pitch listeners who noticed a gradual 33-cent detuning | 0 of 13 | Hedger, Heald and Nusbaum, Psychological Science 2013 |
| Rating of 33-cents-flat notes before vs after detuned listening | 2.08 vs 2.31 | Hedger, Heald and Nusbaum, Psychological Science 2013 |
| Accurate absolute pitch, Kyoto vs Warsaw music students | 57% vs 12% | Miyazaki et al., Music Perception 2018 |
| Non-absolute-pitch share of University of Minnesota music students | 98% | Miyazaki et al., Music Perception 2018 |
| Highest mean tuning deviation added by a neural vocoder | 18.7 cents | Berendes, Maman and Mueller, ISMIR 2025 |
Methodology and Sources
We ran targeted searches and fetched each primary document, then kept only figures we could read in the original paper, standard record or museum page. Where a value comes from an older source quoted by a newer one, the Source cell names both.
- Lynn Cavanagh, A brief history of the establishment of international standard pitch a=440 hertz (draws on Ellis 1968 edition, Lloyd 1949 and The New Grove 1980): PDF
- Whipple Museum of the History of Science, University of Cambridge, Historical Notes: a Brief Chronicle of the Tuning Fork: page
- ISO 16:1975, Acoustics: Standard tuning frequency (Standard musical pitch), catalogue listing: Genorma record
- Alexander Lerch, On the Requirement of Automatic Tuning Frequency Estimation, ISMIR 2006: PDF
- Karin Dressler and Sebastian Streich, Tuning Frequency Estimation Using Circular Statistics, ISMIR 2007: PDF
- Hans-Ulrich Berendes, Ben Maman and Meinard Mueller, Tuning Matters: Analyzing Musical Tuning Bias in Neural Vocoders, ISMIR 2025: PDF
- Yi Qin and Alexander Lerch, Tuning Frequency Dependency in Music Classification, ICASSP 2019: PDF
- Stephen C. Hedger, Shannon L. M. Heald and Howard C. Nusbaum, Absolute Pitch May Not Be So Absolute, Psychological Science 2013: PDF
- Ken’ichi Miyazaki et al., Absolute Pitch and Relative Pitch in Music Students in the East and the West, Music Perception 2018: PDF
- Bairnsfather et al., Conceptual coherence but methodological mayhem: a systematic review of absolute pitch phenotyping, Behavior Research Methods 2025: DOI
- Calamassi and colleagues, three 432 Hz versus 440 Hz trials: the double-blind cross-over pilot on health effects, Explore 2019 (DOI); the sleep study in patients with spinal cord injuries, Acta Biomedica 2020 (DOI); and the anxiety study in emergency nurses during the COVID-19 pandemic, Acta Biomedica 2022 (DOI)
- Hohneck et al., Differential effects of sound interventions tuned to 432 Hz or 443 Hz on cardiovascular parameters in cancer patients, BMC Complementary Medicine and Therapies 2025: DOI
- Mousavi et al., Effects of Music and Aromatherapy on Blood Pressure and Heart Rate Among Endodontic Patients, Clinical and Experimental Dental Research 2025: DOI
- Jebabli and colleagues, two warm-up music trials comparing 440 Hz and 432 Hz: male athletes, Scientific Reports 2025 (DOI), and kickboxers, PeerJ 2025 (DOI)
- Zaki et al., Comparing binaural beats and 432 Hz music for reduction of preoperative anxiety in root canal patients, JPMA 2025: DOI
- Bozok et al., Acute Music-Frequency Exposure Modulates Salivary Stress and Neurotrophic Markers in Young Adults, Brain and Behavior 2026: DOI
- Data watch: the 442 Hz and 443 Hz orchestra figures are archivists’ statements reported in 2006, not acoustic measurements, and the 3,336-track distribution comes from a private collection analysed in 2006; the 2006 paper is the most recent available data for both, and a 2025 ISMIR paper still cites the 50% figure. Sources conflict on when A=440 became global: the Whipple Museum says the Stuttgart-based standard became global in 1926, the pitch history we cite places the decisive agreement at the London conference of May 1939, and it notes that The New Grove (1980) gave the year as 1938 for the same British endorsement. The A=432 trials are small (12 to 72 participants in the trials listed in section 4), Hohneck et al. compared 432 Hz with 443 Hz rather than 440 Hz, the 2022 emergency-staff trial saw anxiety fall in the free-activity comparison group as well, and the kickboxing study describes its music as tuned around 440 Hz. The often repeated 1 in 10,000 prevalence of absolute pitch (Bachem 1955) is described as misleading by Miyazaki et al. (2018), and a 2025 review found AP accuracy thresholds ranging from 20 to 100% across 160 studies. We found no peer-reviewed tuning analysis of a large commercial pop catalogue; the pop values here are single-track estimates that differ by method.
Last updated: November 7, 2026. We update this roundup quarterly, and the next refresh is expected when the proceedings of the next International Society for Music Information Retrieval (ISMIR) conference are published.