Infrasound in Sacred Spaces — The Physics of Awe

The resonant frequency of the human eyeball is roughly 18 Hz. A 32-foot cathedral organ pipe produces a fundamental of about 16 Hz. A typical Gothic nave resonates between 0.5 and 3 Hz. None of these frequencies are heard — they are felt, as pressure in the chest and a faint smearing of peripheral vision. The architectures of awe across continents and millennia may share, beneath their visual diversity, a coincidence of physics in the band below human hearing.

The Tandy discovery

In 1998, Vic Tandy was working late in a Coventry University engineering lab known for inducing dread and apparitions. He felt ill. A colleague had reported a shadowy figure. While preparing for a fencing competition, Tandy noticed a foil clamped in a vice vibrating without contact. He traced the source: a newly installed extractor fan generating a standing wave at 18.98 Hz, matching the room's dimensions.

That frequency sits inside the resonant band of the human eyeball. At sufficient amplitude it oscillates the vitreous humor enough to blur peripheral vision into shapes the pattern-recognition system reads as figures. Tandy published in the Journal of the Society for Psychical Research (1999) and later measured a similar 18.98 Hz standing wave in the undercroft of Coventry Cathedral. The mechanism is settled science; the question of how many "haunted" rooms share it is open.

The Purcell Room experiment (2003)

Composer Sarah Angliss and psychologist Richard Wiseman built a seven-metre infrasonic cannon and placed it behind the stage of London's Purcell Room. 750 audience members heard four short contemporary pieces; during two, 17 Hz infrasound was covertly added at 6–8 dB above ambient.

Self-reported "unusual sensations" rose 22% during infrasound conditions. The descriptors clustered around presence, sorrow, anxiety, peripheral-vision artefacts, and awe. Listeners did not know the manipulation was running. The effect was statistically significant but high-variance: some subjects spiked 50%, others nothing. Openness to Experience predicted magnitude — the same trait that predicts concept frisson response.

The study has not been cleanly replicated. The honest reading: suggestive, underpowered, and confounded by the ordinary low-frequency content of live music. It established the question, not the answer.

Cathedrals as acoustic machines

The Gothic organ is not only an instrument. A 32-foot open flue pipe produces a fundamental near 16 Hz; the rare 64-foot ranks at Atlantic City and Sydney Town Hall reach about 8 Hz. Bach-era builders added these registers knowing they were felt rather than heard. They climb in at the liturgically loaded moments — fortissimo, the cadence of the Sanctus — and arrive as pressure in the ribs.

The building itself adds a lower layer. A typical nave 80–100 m long, 25–30 m wide and 30–40 m high resonates in modes between roughly 0.5 and 3 Hz, excited by wind on the exterior, by massed voices and footfalls, and by modern urban infrasound coupling through the foundations. Stone is the structural prerequisite: wood-framed domestic interiors cannot sustain those modes.

Above the infrasound there is a second acoustic disorientation. Reverberation times of 5–10 seconds delocalize sound sources, so a chanted line "fills" the volume rather than arriving from a direction. Reflections from distant walls land 100–400 ms late, well past the 50 ms threshold at which the brain stops fusing them with the original and starts hearing them as separate events. The default-mode network, which keeps self-location and narrative continuity stitched together, runs on degraded input. The phenomenology overlaps with early meditation, the overview effect, and low-dose psilocybin.

The Paleolithic pattern

Beginning with Iegor Reznikoff and Michel Dauvois in 1988 and replicated by several groups since, surveys of painted caves in France and Spain have found that roughly 90% of paintings and engravings cluster at the most acoustically resonant locations — long echo, dome-like reverberation around 2–3 seconds, pronounced source delocalization. Acoustically dead chambers are largely unpainted.

The inference is that paint placement followed acoustic experience, not just visual access. Drumming, singing or vocalization in those spots would have animated the figures by ear before the eye — a plausible ritual technology 30,000 years old. No systematic infrasound survey has ever been conducted inside a painted cave. That is the largest unrun experiment in the field.

What's contested

Three things have to be separated, and the literature regularly conflates them.

First, Tandy's eyeball-resonance mechanism is well-grounded but explains only a narrow band of effects (peripheral-vision artefacts at 18–19 Hz). It does not, on its own, explain awe.

Second, the 2003 Angliss–Wiseman result is real but small and partially replicated. Whether infrasound carries the effect, or whether low-frequency floor vibration, electromagnetic confounds, or expectancy do, is unsettled.

Third, the leap from infrasound to "the physics of the numinous" is a hypothesis, not a finding. Reverberation, visual scale, ritual priming and crowd dynamics all suppress the default-mode network too. Infrasound is one candidate variable in a stack, and the cleanest cross-cultural study isolating it has not been done.

Why this has to do with other realms

If awe is partly an architectural artefact — a coincidence between building dimensions and the resonant frequencies of human tissue — then several philosophy-of-religion arguments shift onto physical ground. The same default-mode suppression appears in astronauts during the concept overview effect, in long-term meditators, in psilocybin trials, and apparently in worshippers under a Gothic vault. The technologies differ; the neural signature converges. That convergence is what the concept hard problem consciousness has to answer for: why a 17 Hz pressure wave and a view of Earth from orbit can produce phenomenologically similar reports.

An open question

If Paleolithic artists were choosing painting spots by ear, what frequencies were they hearing — and is anyone holding a microphone in Lascaux below 20 Hz?

Key sources

Tandy, V. and Lawrence, T. (1998), "The Ghost in the Machine," Journal of the Society for Psychical Research 62 — the original 18.98 Hz Coventry paper.
Tandy, V. (2000), "Something in the Cellar," JSPR 64 — Coventry Cathedral undercroft follow-up.
Angliss, S. and Wiseman, R., Purcell Room infrasound study, presented at the BA Festival of Science 2003 — to verify: the formal write-up has not been peer-published in a major journal.
Reznikoff, I. and Dauvois, M. (1988), "La dimension sonore des grottes ornées," Bulletin de la Société Préhistorique Française — the founding paper of cave archaeoacoustics.
Stanford "Sound, Space and the Aesthetics of the Sublime" project, Jonathan Berger (Templeton-funded, ongoing as of 2025) — to verify: project page and any published impulse-response datasets.

Abhishek