Nothing on this tab is computed. Each entry is something somebody carried an instrument into a building to find out, with the publication that reported it and a tier saying how far it can be leaned on. Six of them are in the Journal of the Acoustical Society of America or Antiquity.
Six chambers, one band
England and Ireland · Neolithic, roughly 5,000 years ago
Acoustic measurements inside six diverse ancient structures found that every one sustained a strong resonance between 95 and 120 Hz — despite major differences in chamber shape and size.
95–120 Hz across all six. The uniformity is the finding: differently shaped rooms of different sizes should not agree this closely by chance.
What it does not show
It does not show the builders intended it, and the authors do not claim they did. Six is a small sample chosen for accessibility rather than at random, and a resonance in that band is also what a stone room of roughly human scale will tend to produce whether anyone planned it or not. What makes it worth carrying is that the band is narrow and the chambers are not alike.
Newgrange
Brú na Bóinne, County Meath, Ireland · Built around 3200 BCE — older than Stonehenge and the Giza pyramids
A cruciform chamber under a corbelled roof at the end of a nineteen-metre passage, among the structures in the 95–120 Hz survey.
Chamber 6.5 × 6.2 m, corbelled roof reaching about 6 m, passage 19 m, passage orthostats averaging 1.5 m.
What it does not show
The famous midwinter sunrise alignment is securely documented and is not an acoustic claim; the two are often run together and should not be. Nothing about the acoustics bears on the alignment, or the alignment on the acoustics.
Published dimensions: 6.5 × 6.2 × 6 m. The chamber is cruciform with a corbelled roof, not a box. A rectangular calculation gives the scale and the trend, not the measured spectrum.
Camster Round
Caithness, Scotland · Neolithic passage grave
A survey of megalithic monuments in prehistoric Britain documented a wide range of sound effects at Camster Round, including standing waves inside the chamber.
What it does not show
The authors present the acoustics as a property that should be considered alongside visual and spatial analysis — not as evidence of acoustic intent. The honest version of this finding is that archaeology had been reading these buildings with the sound turned off.
El Castillo
Chichén Itzá, Yucatán, Mexico · Built by the Maya, roughly 9th–12th century CE
A handclap at the foot of the staircase returns a descending chirp. Diffraction modelling shows the chirp is produced by the periodic steps, through a Bragg scattering effect — the physics of the echo is settled.
The staircase acts as a periodic grating: each step returns the clap slightly later and the regular spacing sorts the return by frequency, which the ear hears as a falling chirp.
What it does not show
Whether the Maya built the staircase to do this. The resemblance to the call of the quetzal is frequently asserted and is not established; a staircase with regular steps produces a chirp whether or not anyone wanted one. The mechanism is EMPIRICAL. The intention is open, and this archive leaves it open.
Notre-Dame de Paris
Île de la Cité, Paris · Begun 1163; the measurement campaign, 2015
A team measured the cathedral in 2015 using a twenty-second exponential sine sweep, deconvolved to room impulse responses, and built a calibrated acoustic model from it. Reverberation ran about six seconds at mid frequencies. After the April 2019 fire, theirs were the only modern acoustic measurements of the building in existence — and they became the reference for the restoration.
About 6 s reverberation at mid frequencies, against roughly half a second in a living room and two seconds in a concert hall.
What it does not show
That medieval builders could calculate any of this. They could not; the theory is nineteenth and twentieth century. What they had was iteration across generations of buildings and ears that could hear the result, which is a different and in some ways more impressive thing.
Published dimensions: 127 × 48 × 35 m. Nave dimensions, as a rectangular approximation. A cathedral is not a box — the aisles, apse and vaulting all matter — so this gives scale and not spectrum.
The music was written for the room
Gothic cathedrals, northern Europe · 12th century onward
At six to nine seconds of reverberation, rapid melodic writing turns to mush while sustained lines remain intelligible. The polyphonic organum that developed in these buildings — long held tenor lines under faster upper voices — works in that decay and is difficult to appreciate in a drier space.
What it does not show
That the buildings were designed for the music. The causation runs the other way and is better evidenced: the rooms existed, and the music adapted to them. Height and stone came from structure, liturgy and ambition; the acoustics were a consequence that composers then wrote into.
The Newgrange zigzag
West sub-chamber, Newgrange · Carved when the chamber was built
Two zigzag trains carved on a corbel in the west sub-chamber were noted as having the same number of nodes and antinodes as the standing-wave pattern measured from the chamber centre out along the passage.
What it does not show
Nearly everything. It is one carving, observed after the measurement rather than predicted before it, in a decorative tradition full of zigzags — and a zigzag is what a hand draws. The archive carries it because it is genuinely striking and because burying an observation is its own kind of dishonesty, but it is an observation, not a result, and it would need a pre-registered survey across many chambers to become one.
Sources
- Noted in connection with Jahn, Devereux & Ibison, J. Acoust. Soc. Am. 99(2), 1996 · journal
Hal Saflieni Hypogeum
Paola, Malta · Cut roughly 4000–2500 BCE
A rock-cut underground complex whose "Oracle Room" is widely reported to resonate near 110 Hz. A 2020 analysis argued the frequency spectrum and geometry appear tuned.
What it does not show
The 2020 analysis is an arXiv preprint and has not been through peer review. The resonance itself is reported consistently enough to be worth stating; the word "tuned" carries an intention claim that the measurement cannot reach. Carried at HYPOTHESIS for both reasons.
The 110 Hz brain study
Laboratory, UCLA · 2008
Thirty adults listened to tones at 90, 100, 110, 120 and 130 Hz under EEG. Activity in the left temporal region was lower at 110 Hz than at the other frequencies, and prefrontal asymmetry shifted at 110 Hz.
What it does not show
Almost certainly less than it is used for. One laboratory, thirty subjects, a humanities-adjacent journal, and no independent replication in a mainstream venue in the seventeen years since — a search in October 2026 found only low-tier proceedings from a single other group, which reported different effects from the ones Cook reported. Meanwhile the adjacent literature that IS well powered points the other way: a three-arm crossover trial exposing noise-sensitive adults to seventy-two hours of simulated infrasound per arm found nothing on any outcome, and a double-blind study of seventy-two people found that expectation, not exposure, produced the symptoms. None of that tests 110 Hz specifically, and absence of an infrasound effect is not absence of a 110 Hz effect. But it is the shape of the evidence around the claim, and the shape is not encouraging.
Chavín de Huántar
Ancash, Peru, at 3,180 m in the Andes · Andean Formative, roughly 3,000 years ago
The best-instrumented archaeoacoustic site in the world, and the one place on this page where the instruments and the architecture were both excavated. Twenty-one decorated Strombus shell trumpets — pututus — were recovered from the site, and a Stanford team characterised each one and measured the stone galleries they were played in: impulse responses, sound transmission between chambers, and systematic auditory localisation experiments with listeners inside the passages.
The galleries return short reverberation with dense early reflections and low inter-aural cross-correlation — the acoustic signature of a space in which a listener cannot tell where a sound is coming from.
What it does not show
That the disorientation was designed. It is a measured property of the architecture, and the researchers describe it as obscuring localisation cues rather than as an intention; the galleries are narrow, hard and irregular, which is what produces the effect whether or not anyone wanted it. What makes Chavín unusual is that the argument can be had at all — most sites offer a room and a guess, and this one offers the room, the instruments that were played in it, and listeners tested inside it.
Sources
- Kolar, M. A., "Sensing Sonically at Andean Formative Chavín de Huántar, Perú," Time and Mind 10(1), 2017 · journal
- Abel, Rick, Huang, Kolar, Smith & Chowning, "On the Acoustics of the Underground Galleries of Ancient Chavín de Huántar, Peru," Acoustics '08, Paris, 2008 · academic
- Kolar, M. A., Archaeological Psychoacoustics at Chavín de Huántar, Perú, PhD dissertation, Stanford University (CCRMA), 2013 · academic
The acoustic pots, and the chronicler who laughed
Churches across France, and further east · Documented 1432; the practice is medieval and the idea is older
Clay pots were built into the walls and vaults of medieval churches, mouths open to the room. A chronicle written at Metz in 1432 records the prior of the Célestins having them installed in the choir because he believed they would make the singing better — which makes this, so far as the record goes, the only place in this archive where a builder states an acoustic intention in his own time rather than having one inferred from his building. An acoustical survey of French churches where the pots survive measured their resonances and found them consistent with an acoustic purpose: tuned absorbers, pulling down reverberation at the frequencies the speaking voice excites most.
Valière et al. surveyed 25 French churches with pots still in place and measured the pots' resonance frequencies. The chronicle is Bibliothèques-Médiathèques de Metz, Chronique des Célestins de Metz, 1432.
What it does not show
That it worked, and the medieval record is funnier and more honest than we are: the chronicler who wrote the entry mocked the prior for believing it would. So the text attests the intention and simultaneously disputes the result, which is a more sophisticated epistemic position than most of what circulates about ancient acoustics today. Note also that the measured effect — absorption, less reverberation at voice frequencies — is not the same as the "more resonant" the prior seems to have wanted, even though clearer speech in a stone church might well be heard as better singing. The archive has not read the 1432 manuscript; it is cited here as quoted in the 2013 paper.
Göbekli Tepe
Şanlıurfa Province, Türkiye · Roughly 11,500 years ago — before pottery, before farming
Because these structures survive only to their foundations, their acoustics were reconstructed digitally rather than measured, and reverberation times calculated from the models. The result for the enclosures at Göbekli Tepe, Nevalı Çori and Jerf el-Ahmar is that they were open and undifferentiated spaces that would carry a shared experience rather than stage a hierarchy. At Göbekli Tepe specifically a speaker standing near the central pillars would be heard clearly throughout the enclosure — while the broad faces of those same tall pillars blocked the line of sight between speaker and listeners.
A voice that arrives from everywhere and from no one you can see. The paper reaches that from geometry, not from atmosphere.
What it does not show
It is a calculation on a reconstruction, not a measurement in a room — the rooms no longer have their ceilings, and the reverberation of a space depends on what closed it. The archive carried "nothing acoustic has been published on Göbekli Tepe" until October 2026, which was wrong; this entry exists because a research sweep corrected us.
The Rosslyn Chapel "musical cubes"
Roslin, Midlothian, Scotland · Chapel begun 1446; the claim, 2007
It was proposed that 213 carved cubes in the chapel ceiling encode a musical score, their patterns matching Chladni figures, and a reconstruction was performed as the "Rosslyn Motet."
What it does not show
The archive does not carry this as evidence of anything. An 1860s restoration replaced a large number of the cubes, and lithographs from 1837 show cubes already missing beforehand — so the surface being read is substantially Victorian. A music professor who attended the premiere judged the result unlike any medieval music he knew, and the reconstruction required heavy interpretive licence at every step. It is listed here so that a reader who arrives having heard it finds out why it is not in the exhibit.
