The Taos Hum: Four Laboratories, an Engine Nobody Has Parked and a Blank Box
In 1993, Los Alamos, Sandia and two other institutions took measurements inside the homes of people who heard the hum. They found nothing capable of causing it. That is not the same as saying it does not exist.
In the early 1990s, a minority of residents of Taos (New Mexico) described a low, continuous hum comparable to a diesel engine idling. In 1993, a team from the University of New Mexico, Los Alamos, Sandia and the Phillips laboratory surveyed the population, ran audiometric tests and measured acoustic and electromagnetic fields in the homes of those affected. No signal sufficient to explain the perception was detected. In other localities with similar complaints, such as Kokomo or Windsor, measurements did identify industrial sources.
Executive summary of the case file. Everything essential before you dig deeper.
Fact sheet
OSINT - Traceability
Some noises come in through the door and some move in without knocking. The first kind can be thrown out: you locate the construction site, warn the neighbor, file the complaint at the registry and, with luck and three months, the world goes quiet again. The second kind admits no appeal, because there is no window where you can file one. It is there when you wake up, it goes away when you start the car and it comes back as soon as you switch the engine off, and the worst part is not hearing it: the worst part is that the person beside you, in the same room, hears absolutely nothing.
That is what happened to a small percentage of the residents of Taos, New Mexico, in the early 1990s. A low, continuous hum, described almost always with the same image: a diesel engine idling far away, which never quite starts and never quite leaves. A few people described it, most denied it, and it ended up getting what very few phenomena in this archive manage to get: a serious investigation, with national laboratories involved, calibrated instruments and a report.
And the report, which is where it gets interesting, did not say “there is nothing there.” It said something colder and more useful: that they found no signal capable of explaining it.
What can be verified first: who measured, with what, and what they got
In 1993, at the insistence of residents and with the backing of New Mexico’s representatives in Washington, a team was assembled of a kind that is not assembled for just anything. It included people from the University of New Mexico, Los Alamos National Laboratory, Sandia laboratories and the Air Force’s Phillips laboratory. The work was led by engineer Joe Mullins. In other words: to investigate a hum in a mountain town, the same institutions that deal with the country’s nuclear weapons physics were brought together.
The method was the right one and deserves to be told, because it is exactly what this archive always asks for. First, a survey to delimit the phenomenon: how many hear it, how they describe it, when it began, under what conditions it appears. Second, audiometric tests on those affected, to find out whether the problem was in the air or in the ear. Third, direct measurement: installing acoustic and electromagnetic field equipment in the homes of those who heard it, while they were hearing it, and recording.
The results were as follows, and they should be read slowly. Those affected were a clear minority of the population, on the order of two percent. Their audiograms showed no common pattern that set them apart from everyone else. And the instruments picked up no signal, neither acoustic nor electromagnetic, with enough intensity to produce what those people were describing as they described it.
Note what that paragraph does not say. It does not say there was nothing in the air: there were, as there are anywhere, power lines, distant traffic, machinery and the ionosphere. It says that nothing that was measured came close to accounting for this. It is the difference between a case file closed for lack of evidence and a case file closed because the evidence pointed elsewhere.
A phenomenon that is not recorded but is suffered is not a false phenomenon, only one with the wrong address: we were looking for it in the air, and perhaps it is registered somewhere else.
And before Taos there was Bristol, which is where this story really begins. In the 1970s, a British newspaper published a note asking whether anyone in the city heard a persistent low noise, and received hundreds of letters. Hundreds. People who had spent months convinced they were alone in this and who discovered, through a newspaper column, that they had neighbors in the same condition. Some of those complaints were eventually attributed to traffic and industry; others were not. What matters for our case file is the precedent: New Mexico did not invent the hum, it only gave it a proper name and a mountain backdrop.
The Bristol hum
A British newspaper collects hundreds of letters from residents who hear a persistent low noise. It is the first documented mass case.
Complaints begin in Taos
Residents of the area describe a continuous low hum, comparable to a distant diesel engine.
Joint study
The University of New Mexico, Los Alamos, Sandia and the Phillips laboratory take measurements in the homes of those affected. No sufficient source appears.
Popular publication of the results
Mullins and Kelly summarize the work in the bulletin of the Acoustical Society of America.
The Kokomo, Indiana hum
An acoustic study commissioned by the city locates low tones coming from industrial facilities, though it does not explain all the complaints.
The Windsor, Ontario hum
University measurements link the noise to industrial activity on an island on the U.S. side of the border.
What is said to have happened, and what is said to cause it
From 1993 on, and especially once the matter reached the national press, the Taos hum did what phenomena do: it grew without asking for a building permit. It was attributed to military antennas, underground installations, experiments in communicating with submarines at extremely low frequencies, infrasound weapons and, as might be expected in New Mexico, a non-human presence underground. Each hypothesis arrived with its photocopied bibliography and its secondhand witness.
It pays to separate the wheat from the chaff. That military emissions exist at very low frequencies is true and documented: communication with submerged submarines is done precisely this way, because only extremely long waves penetrate water. That a human being can hear them is not. They are electromagnetic waves, not sound, and the ear converts them into nothing. Confusing a low-frequency radio emission with an audible tone is a category error of the same caliber as confusing the smell of bread with the recipe.
That there are industrial sources capable of generating low tones audible for kilometers is also true, and here there are indeed closed cases. In Kokomo, Indiana, a study commissioned by the city council at the beginning of the century located specific tones generated by factory installations. In Windsor, Ontario, measurements pointed to steelmaking activity on an island across the river. When the hum has an external source, it shows up on the instruments. That is how it behaves: real noises leave a trace, just as vehicles leave ruts.
And in Taos it left none. That is the central fact of the case file and the one neither popular version wants to accept. Neither “it was a factory and they covered it up” (there is no factory to cover up, and the instruments were inside the homes) nor “it was all psychological,” which is an elegant way of calling a liar anyone who cannot prove what they hear.
It is worth pausing on a methodological detail that is almost never told and that decides the whole case. Measuring a sound that comes and goes requires measuring when it comes, and an instrumentation campaign has a start date and an end date like any contract. If the phenomenon depends on temperature, on wind or on the shift schedule of a distant factory, a team can spend weeks recording the most impeccable silence in the world. This does not invalidate the 1993 result, but it puts an asterisk on it: what was shown is that there was no permanent, strong signal, not that there was never anything.
What it could be, with no need for Martians or secret weapons
There are three explanations that hold up without inventing anything, and the global phenomenon is most likely a mixture of the three in different proportions depending on the town.
A real sound that reaches only some ears
Low-pitched sound behaves in a counterintuitive way: it passes through walls easily, reflects poorly and accumulates in some places while vanishing in others, forming a map of strong and weak zones. Inside a room there may be one spot where the tone is clearly audible and another, two meters away, where it is not. Add that hearing sensitivity at low frequencies varies enormously from person to person and deteriorates with age differently from sensitivity to high frequencies. Result: two people in the same living room, one hearing an engine and the other hearing nothing, and both telling the truth.
Distant background noise, channeled by the atmosphere
At night, when the ground cools, a temperature inversion often forms: a layer of warm air over a cold one that acts like the ceiling of a corridor and channels sound instead of letting it escape upward. That is why in the small hours you hear the highway that does not exist by day. A compressor, a pump, a plant twenty kilometers away can become audible only under certain weather conditions, which explains the intermittent and capricious character of the phenomenon, and also why the measurement team can spend a whole week without catching anything.
And the ear, which also manufactures
The third is the one people like least and the best documented. The ear is an active organ, not a microphone: it amplifies what reaches it, and that amplification can run away and generate perception without a stimulus. This is the mechanism of tinnitus, which a not insignificant percentage of the population suffers and which, in its low-pitched variant, is described as a continuous hum. Tinnitus is a real signal generated inside the auditory system, neither imagination nor suggestion. That it does not show up on a microphone placed in the living room does not make it less true for the person who suffers it; it makes it inaccessible to the microphone, which is something else.
A fourth possibility remains, neither fish nor fowl, which probably explains a good share of the cases around the world: the mixture. A real ambient tone, weak, intermittent, at the edge of the audible, acting as a seed; and on top of it an auditory system that, once it has learned to look for it, always finds it, even when it is no longer there. The brain is terrible at ignoring what we have taught it to attend to. Anyone who has lived under an air conditioner and then kept hearing it for a while after switching it off knows exactly what we are talking about, with no need for a bibliography.
Why this case will not die
Because it is the perfect example of a phenomenon that was investigated well and not resolved, a category this archive defends tooth and nail. There was no negligence here, no cover-up, no disdain: there were four scientific institutions measuring inside people’s homes. And even so the result was an I don’t know. When someone claims that science only ignores what bothers it, it is worth reminding them of Taos: it was studied, measured, published and it did not come out.
Because it also forces us to distinguish two questions that sensationalism always mixes up. The first is “do these people hear something?”, and the documented answer is yes, their experience is real and consistent. The second is “is there a wave in the air that causes it?”, and the measured answer is that none was found. Holding both at once describes the problem precisely, with no contradiction in it. Most of the enigmas in this archive would improve a lot if that double question were applied to them.
And because it points to the road that does work: instrument the site and wait. It is what has been done for decades with the lights of the Norwegian valley of Hessdalen, and it is what was missing here in the only respect that mattered: one week in one house cannot match three years in twenty houses. It is also related to our case file on psychophonies, though for the opposite reason: there the sound is on the recording and the problem is the brain that interprets it; here the sound is not on the recording and the problem is that someone hears it anyway.
An engine nobody has parked, in a town where two percent of the people cannot stop hearing it. Sometimes the most honest case file one can write is the one that ends with the stamp applied and the box for the cause left blank.
OSINT – Source traceability
No exact range of perceived frequency is given: the figures in circulation come from subjective descriptions by those affected and not from an instrumental measurement, precisely because the measurement did not detect the signal. It is described qualitatively.