By Linnea Voss · Published May 13, 2026 · Updated May 13, 2026
Last reviewed: May 13, 2026.
A woman in Taos walks me to the corner of her kitchen and asks me to stand still. She is in her late sixties. She tells me the hum is loudest here, against the south wall, near the refrigerator, in the early hours before traffic begins. She says she has heard it since the summer of 1991. Her husband cannot hear it. Her daughter cannot hear it. Two of her neighbors can, and one of them keeps a small notebook on the counter, dated entries marked with a single capital H. I do not hear anything. I write that down too.
The Hum is the name researchers give to a low-frequency rumble that a small percentage of residents in particular locales report hearing, often at night, often indoors, and almost always undetectable by the people standing next to them. It has been documented in Taos, New Mexico; Bristol, England; Windsor, Ontario; Auckland, New Zealand; Largs, Scotland; and dozens of other places over the past fifty years [1]. The folklorist’s job, when the percipient steps forward, is not to decide whether the sound is real. The job is to render what the percipient heard, where, and when, and to set that account inside the longer literature on perception, infrastructure, and the threshold of hearing.
The Direct Answer
The Hum is a persistent low-frequency rumble reported by roughly two to four percent of residents in localized regions worldwide. Most listeners cannot detect it. Current research treats it as a real perceptual phenomenon with multiple plausible causes, including otoacoustic emissions, environmental low-frequency sound from industrial sources, and tinnitus-spectrum sensitivities, rather than a single unified mystery [1][2].
Five Hums on the Map
Each documented Hum carries the name of the town where it was first reported, and each carries its own social texture. The Taos Hum entered the academic record in May 1993, when the University of New Mexico, the Los Alamos National Laboratory, Sandia Labs, and Phillips Air Force base ran an investigation at the request of the United States Congress. The team interviewed 1,440 Taos residents. About two percent reported hearing the Hum. Acoustic measurements at the homes of “hearers” did not locate a clear external source within the frequency range the team surveyed [1][2].
The Bristol Hum surfaced in the southwest of England in the late 1970s. Local residents wrote to the Sunday Mirror, which printed an appeal for letters. Roughly eight hundred replies came in. Geoff Leventhall, a British acoustician who has spent more than four decades on low-frequency noise, became one of the central figures in the British investigations [3]. He found that some Bristol cases mapped onto identifiable industrial sources. Others did not.
The Windsor Hum is the case that came closest to a confirmed external cause. Windsor sits across the Detroit River from Zug Island, a heavily industrialized peninsula on the Michigan side. Residents began reporting a rumble around 2011. A 2014 study commissioned by the Canadian government identified Zug Island steel operations as the most likely source [4]. Operations at the U.S. Steel blast furnace there were idled in 2020. Windsor residents have since reported a noticeable decrease, though not a complete disappearance, of the Hum.
The Auckland Hum and the Largs Hum sit at the other end of the evidence spectrum. Both are well attested in community reports. Neither has produced a confirmed external source. The New Zealand Herald has covered the Auckland case off and on since the early 2000s, often with the kind of careful, hedged language that long-running unexplained-phenomena stories tend to acquire.
What the Acousticians Say
Geoff Leventhall’s review of the literature for the UK Department for Environment, Food and Rural Affairs in 2003 remains a useful reference point [3]. He argues that “the Hum” is not one phenomenon. It is a label that gets applied to several distinct experiences that happen to share a common surface description, namely, a low rumble heard mostly indoors, mostly at night, by a small minority of residents.
Some of those experiences have external sources. Industrial fans, ventilation systems, distant pipeline compressors, certain transformers, and combined-cycle gas turbines can all produce low-frequency content that travels long distances through ground and structures. Buildings can act as resonators. A given room’s standing waves can amplify a frequency that is undetectable outdoors. This is well-established acoustics, and Leventhall has documented case after case where careful measurement located the offending machine [3].
Other Hum reports do not yield to that approach. In those cases the candidate explanations move into the hearer’s own auditory system. Spontaneous otoacoustic emissions, the faint sounds the inner ear itself produces, can be perceived by a small fraction of listeners under quiet conditions. Tinnitus, particularly low-frequency tinnitus, can present as an environmental rumble. Hyperacusis and somatic sensitivity to vibration round out the differential. None of these flatten the experience. They locate it.
Why Only Some People Hear It
The two-to-four-percent figure is the through-line across studies. David Deming, a geophysicist at the University of Oklahoma, surveyed Hum hearers in 2004 and found a roughly comparable prevalence [2]. The pattern is consistent enough that researchers now treat selective audibility as one of the defining features of the phenomenon, not a puzzle that disqualifies it.
Hearing thresholds vary widely below 100 Hertz. A sound that sits at 40 dB at 30 Hertz is inaudible to most listeners and clearly audible to a small minority [3]. Age, individual cochlear sensitivity, head and torso resonance, the room a person sleeps in, and even body posture can all shift the threshold. When the source is at the edge of audibility for a population, a small percentage will land above the line and the rest below it. That is enough to explain the social pattern without invoking anything exotic.
It also explains why couples disagree at the breakfast table. The woman in Taos hears the Hum. Her husband does not. He is not dismissing her. His threshold sits a few decibels above hers in the frequency band where the sound lives.
The Longue Durée
The contemporary Hum has a deeper history than its 1970s emergence suggests. The Hessdalen lights in Norway, documented since 1981, came with persistent reports of ambient low-frequency sound around the valley [5]. Early-modern accounts of “demonic music” heard at night in monasteries and farmhouses share the surface description of the modern Hum, though the explanatory frame around them belonged to a different cosmology. Folklore archives in Scandinavia, the British Isles, and Appalachia carry hundreds of variants of the night-rumble motif. The percipients were not lying. They were locating an experience inside the world they had.
What the modern record adds is instrumentation. We can now measure what earlier listeners could only describe. We can place a microphone in the woman’s kitchen and run a frequency-domain analysis. Sometimes the spectrum analyzer agrees with her. Sometimes it does not. Either outcome is data.
The Researcher’s Posture
When I sit at the woman’s kitchen table I take her seriously. She has kept records for thirty-three years. She has compared notes with two of her neighbors. She has read the 1993 University of New Mexico report. She is, in the technical sense of the word, a careful informant. The folklorist’s discipline asks me to render her account with the same exactness she has brought to it.
The acoustician’s discipline asks me to bring measurement. Both disciplines belong here. The UK Council for Acoustic Comfort, the British acoustical engineering community Leventhall helped seed, has been doing exactly this kind of joint work for decades, treating low-frequency noise complaints as both technical problems and social ones. Their case files contain examples where the source turned out to be a neighbor’s heat pump, examples where it turned out to be a ferry terminal twelve miles away, and examples where no source was ever found and the resident learned, with help, to live alongside a perception that was not going away.
There is a temptation, when a phenomenon resists a unified explanation, to call it a mystery and stop. The Hum is not one mystery. It is a cluster of overlapping causes that happen to produce a similar surface report. Sorting them is patient work. The woman in Taos is not asking us to romanticize her experience. She is asking us to attend to it.
The Social Pattern of Hearing
Selective audibility produces a social pattern as recognizable as the acoustic one. In Taos in the 1990s and in Windsor in the 2010s, residents who heard the Hum found each other. They formed informal mailing lists, then online forums. They compared the time of day the rumble felt strongest. They compared the rooms in their houses where it amplified. They learned, in some cases, to recognize the timbre of one another’s accounts and to flag descriptions that drifted away from the shared phenomenology toward something different. That self-selecting pattern matters methodologically. It improves the signal in the percipient-report record, and it also raises the risk of in-group reinforcement that an outside researcher has to weigh.
Sociologists of science describe this kind of community as an epistemic public, a group of laypeople doing real observational work outside formal institutions. The Taos hearer network and the Windsor complaints register both qualify. Their notebooks, sleep logs, and weather-correlated entries became part of the data the formal studies later drew on. The folklorist’s training warns against romanticizing such communities. The acoustician’s training warns against dismissing them. Held together, the two warnings produce something better than either alone, which is the disciplined humility of treating witnesses as collaborators in an investigation neither side can finish alone.
Where the Field Goes Next
Three lines of work look most promising. First, longitudinal community studies that combine continuous acoustic monitoring with structured percipient diaries. The Windsor case showed that when industrial operations changed, the Hum changed, and a study designed to catch that kind of correlation in real time is feasible now in a way it was not in the 1990s. Second, audiological screening of self-identified hearers, looking for otoacoustic emission patterns, low-frequency tinnitus profiles, and hyperacusis markers. Third, the slow folkloric work of cataloguing variants across cultures and centuries, because the night-rumble motif is older than any of the named Hums and the comparative record matters.
For the woman in Taos, none of this will end the question of what she has been hearing for thirty-three years. The point of careful work is not to end her question. The point is to keep the question in good company.
Sources
[1] Mullins, Joe and Kelly, James P. The Taos Hum: An Investigation. Joint study by the University of New Mexico, Los Alamos National Laboratory, Sandia National Laboratories, and Phillips Air Force Base, 1993. Britannica overview of the Taos Hum investigation.
[2] Deming, David. “The Hum: An Anomalous Sound Heard Around the World.” Journal of Scientific Exploration 18, no. 4 (2004): 571-595.
[3] Leventhall, Geoff. A Review of Published Research on Low Frequency Noise and Its Effects. Report for the UK Department for Environment, Food and Rural Affairs, 2003. UK government archive.
[4] Howe Gastmeier Chapnik Limited and Western University. Windsor Hum Study. Commissioned by Health Canada and the Ministry of the Environment, 2014. Canadian government Windsor Hum reference.
[5] Teodorani, Massimo. “A Long-Term Scientific Survey of the Hessdalen Phenomenon.” Journal of Scientific Exploration 18, no. 2 (2004): 217-251.
For more on related unexplained perceptual phenomena, see the pillar guide at esovitae.com paranormal and supernatural phenomena.
