Two National Labs, One Town's Hum, No Source

What We Know

Catanya Saltzman heard it first, one night in the spring of 1991: a low, droning hum outside her home near Taos, New Mexico, with no direction she could point to and no obvious source — not a truck idling somewhere, not a neighbor’s generator, nothing she could walk toward and find. About two weeks later, her husband, Bob, started hearing it too, and kept hearing it, by his own account, continuously from then on.

They weren’t alone for long. Once the Saltzmans went public with their complaint, other Taos residents came forward describing something similar: a rumble some compared to a distant truck, a steadier pulse for others, usually strongest indoors and at night. By 1993, enough people were reporting it, and pressing New Mexico’s congressional delegation hard enough, that a federal investigation got funded.

That summer, a team led by Joe Mullins of the University of New Mexico, with Sandia National Laboratories, Los Alamos National Laboratory, and the Air Force’s Phillips Laboratory all contributing equipment and expertise, set up acoustic, seismic, electric, and magnetic sensors around Taos. They tested and eliminated the obvious mundane explanations one by one: power lines, gas lines, gravel pits, desert excavation work. Their own working theory going in was geological, a tectonic tremor along the Rio Grande Rift, the fault system running directly beneath the town.

None of it panned out. The team’s equipment, capable of registering sound quieter than the human ear can detect, recorded no unexplained acoustic signal and no matching seismic event. “These people are not kooks,” Mullins told a reporter that year. “Several of us guess that whatever it is, it’s more complicated than one source and one sound.”

By one commonly cited estimate, roughly one in fifty Taos residents are hearers. Thirty-plus years later, they’re still describing the same thing, and no instrument has ever agreed with them.

The Roadblock

Every specific candidate anyone has proposed for the Taos Hum has been tested against real instruments and come up short: not weak evidence, but a clean negative from equipment built to catch exactly this kind of signal. That’s the strange part. This isn’t a case where nobody looked. A federal team with the combined resources of two national laboratories and an Air Force research lab spent a summer trying, in the open, with the town’s cooperation.

What that effort ran into is a mismatch between the sound as reported and the sound as measurable. Hearers describe something specific, directional, physically felt: headaches, ear pressure, disrupted sleep. But roughly 2 percent of a population hearing something the other 98 percent can’t is also the textbook signature of an internal, physiological effect, not an external one. Every physiological explanation tested so far, including a 2026 study built specifically to settle it, has come back partial. A separate private engineer’s oscilloscope reading in 1993 found signals nobody has since reproduced. Whatever is happening, it hasn’t held still long enough, or shown up consistently enough, for any single instrument (acoustic, seismic, or auditory) to pin it down twice.

Best Guesses

Rio Grande Rift micro-tremor

This was the investigating team’s own working theory when they arrived in Taos in the summer of 1993: a tectonic micro-tremor, too faint and too low in frequency for a normal seismograph to register, moving along the Rio Grande Rift. No matching seismic event turned up in a full summer of dedicated monitoring, and no scientific report since has revived the idea with better instruments. But the rift itself isn’t a hypothetical — it’s the actual fault system the town sits on, and the team took it seriously enough to build seismic sensors into the search from the start.

Low-frequency tinnitus

The most rigorous physiological test of Taos-style hearers to date came in 2026, when researchers recruited 28 people who report hearing a hum and ran high-resolution hearing tests alongside inner-ear emission scans most clinics don’t have access to. They ruled out two other physiological explanations in the same study: most participants had ordinary hearing, and nobody produced the kind of low-frequency inner-ear signal that might explain a phantom sound. What was left standing was tinnitus, a real, internally generated sound some auditory systems produce that the brain locates outside the body — a finding that doesn’t explain the Hum’s town-by-town clustering, when ordinary tinnitus spreads evenly through a population. Twenty-eight hearers and a control group still didn’t produce one mechanism that fit everybody.

VLF radio-wave perception

Geophysicist David Deming has spent two decades pointing at something the acoustic explanations don’t address: hum reports cluster in specific small towns rather than scattering evenly through a population the way an internal condition like tinnitus should. His proposed answer isn’t a sound at all. Very-low-frequency radio transmissions, of the kind the U.S. Navy has used for decades to reach submerged submarines, interact with the body directly and produce a sensation the brain interprets as an external drone. It’s a real, documented physical effect in other contexts, and also a minority position nobody in the acoustics or audiology literature above has taken up or tested. No VLF signal has ever been matched to when a Taos hearer’s symptoms actually start.

Unlocated industrial source

This isn’t speculation about towns in general. It’s precedent. In Windsor, Ontario, investigators eventually traced years of complaints to blast-furnace operations on an industrial island across the river; when the plant idled, the hum reportedly went with it. In Kokomo, Indiana, consultants identified two separate industrial tones, from a casting plant and a chemical manufacturer. Taos could be the same kind of case, with a source nobody’s aimed the right microphone at yet. The trouble is that the 1993 team already tested the obvious local candidates, power lines, gas lines, gravel pits, desert excavation equipment, and eliminated every one. Nobody has proposed a specific replacement since, which leaves this theory with real precedent elsewhere and nothing concrete pointing anywhere in Taos itself.

The Verdict

None of the four theories here would need much to close: a matching seismic reading for the rift; a low-frequency tinnitus diagnosis with a clean inner-ear test for hearers where the 2026 study left the door open; a Navy VLF broadcast schedule cross-referenced against decades of Taos onset reports; or, simplest of all, one unidentified low-frequency source somebody eventually points a directional microphone at and finds. Windsor took the better part of a decade to get exactly that kind of answer. Taos hasn’t gotten it yet. Every well-funded, well-instrumented attempt to find it, private and federal, 1993 and 2026, has come back the same way: real physiological effects, no external signal to match them to. The next test that could change that hasn’t been run. Bob Saltzman, by his own account to investigators in 1993, had already been hearing it without a break for two years.