Roughly 2% of a New Mexico town hears a low drone that instruments can't reliably detect, a mystery serious enough to trigger a federal lab investigation. That is the hook, and unlike most things in this genre, every clause of it is documented.

Taos sits high in the mountains of northern New Mexico, an art town, a quiet town. Sometime around the early 1990s, residents began reporting a persistent low-frequency sound, a hum, a buzz, a drone, the kind of throb you feel in your chest more than hear in your ears. It kept people awake. It made some of them physically ill. And critically, it was not universal. The person next to you might hear nothing at all while you were being driven slowly insane by a noise with no apparent source.

By 1993 the complaints were loud enough, politically, that the New Mexico congressional delegation requested an investigation, and this is where the Taos Hum separates itself from every roadside legend. The team that responded was not a podcast crew. It was a dozen investigators drawn from the country's most serious institutions: Joe Mullins of the University of New Mexico as coordinator, Horace Poteet of Sandia National Laboratories, Rod Whitaker of Los Alamos National Laboratory, and Mark Leher of the Air Force's Phillips Laboratory. These are the people who instrument nuclear tests. Over the summer of 1993 they ran simultaneous measurements in Taos of acoustic, seismic, electric, magnetic, and electromagnetic signals.

Here is the evidence, and it is a beautiful kind of frustrating. The team surveyed roughly 1,440 residents and extrapolated that about 2 percent of the population were 'hearers.' Those hearers were consistent about the character of the sound, a low rumble in a narrow frequency band. But when the weapons-lab instruments went looking for a physical signal to match that band, they found nothing that fit. In the words of the investigators, there were no known acoustic signals that might account for the hum, and no seismic events that might explain it. They did not find a power transformer, a gas pipeline, a secret facility, or a geological resonance. They found people who clearly, sincerely, consistently heard something, and instruments that recorded the silence around them.

The skeptical-but-fair reading is genuinely the strongest part of this story, and it points inward rather than outward. 'The Hum' is now known to be a worldwide phenomenon, reported in Bristol, in Windsor, in dozens of locations, almost always heard by a small minority. The leading scientific candidate is not a hidden machine but the human auditory system itself: spontaneous otoacoustic emissions, the ear generating its own faint tones; heightened sensitivity to genuine but extremely low-level low-frequency environmental noise that most people filter out; or a perceptual loop in the brain not unlike tinnitus, latched onto a frequency. That would explain why only 2 percent hear it, why instruments parked outside the body find nothing, and why no single source has ever been located. The hum may be real and internal at the same time.

But notice what that explanation does not do. It does not actually close the file, because it cannot be confirmed either. The federal team did not conclude 'it is in their heads.' They concluded, honestly, that they could not identify the source, and a perceptual hypothesis that nobody has nailed down is still just a hypothesis dressed as a verdict. You are left with a documented physical sensation, experienced by thousands of people across the planet, that the best-equipped laboratories in the United States investigated by act of Congress and could neither record nor explain.

The unresolved question is the one that should bother you: if a sound real enough to summon Sandia and Los Alamos cannot be detected by Sandia and Los Alamos, where exactly is it, and what does it mean that 2 percent of us are tuned to a frequency the instruments swear is not there?