Why Old Buildings Feel Creepy: It’s Not Ghosts, It’s Infrasound

Haunted by a Bad Vibe? Science Has a New Possible Explanation

Walking into an old building and feeling a sudden wave of unease, agitation, or dread with no obvious cause in a common feeling. Some chalk it up to atmosphere or overactive imagination, while others veer towards something more supernatural. According to a new study, the real culprit may be as mundane as infrasound, an ultra-low-frequency rumble produced by aging pipes, ventilation systems, and heating equipment that sits entirely below the threshold of human hearing.

Researchers at MacEwan University in Edmonton, Canada, found that exposing people to infrasound at roughly 18 Hz raised their cortisol levels and made them feel more irritable, even though they could not reliably tell the sound was there at all. “Consider visiting a supposedly haunted building,” said senior author Rodney Schmaltz. “Your mood shifts, you feel agitated, but you can’t see or hear anything unusual. In an old building, there is a good chance that infrasound is present, particularly in basements where aging pipes and ventilation systems produce low-frequency vibrations. If you were told the building was haunted, you might attribute that agitation to something supernatural. In reality, you may simply have been exposed to infrasound.”

Published in Frontiers in Behavioral Neuroscience, the findings add a new layer to a long-running debate about whether infrasound can affect people below the threshold of conscious detection. Prior research has been mixed, with some studies finding little effect and others reporting anxiety, sleep problems, and general discomfort.

Infrasound Exposure Tested Without Participants Knowing

Thirty-six undergraduate students were randomly assigned to one of four groups, listening to either calming instrumental music or unsettling horror-style ambient audio, with infrasound at roughly 18 Hz either switched on or left off. Two large subwoofer speakers, hidden from view, generated infrasound between 75 and 78 decibels, a range consistent with what mechanical systems commonly produce inside buildings.

Participants had no reliable way of knowing whether infrasound was present, and when asked afterward, their guesses were no better than flipping a coin. Believing the infrasound was on had no detectable impact on cortisol levels. Each participant provided saliva samples immediately before the music started and again 20 minutes later, which were analyzed to measure cortisol, the hormone the body releases under stress.

Cortisol Rose Even at Frequencies Below the Hearing Threshold

Participants exposed to infrasound showed a statistically meaningful increase in salivary cortisol compared to those who were not, regardless of which type of music they heard. It is worth noting that these results emerged under tightly controlled laboratory conditions, and whether the same effects would occur in everyday life remains an open question.

On the self-report side, participants in the infrasound condition rated themselves as more irritable, reported feeling less interested afterward, and described the music as noticeably sadder. That sadness perception was the strongest self-report finding in the study. Because these mood ratings came from item-level analyses in a small sample, the researchers treated them as signals that need confirmation, not settled findings.

Anxiety, on the other hand, did not increase. People did not report feeling more nervous or scared when infrasound was present. Instead, the dominant emotional shift was toward irritation, disinterest, and a low-grade emotional discomfort.

Notably, the cortisol increase held up even after researchers statistically accounted for the negative mood states participants reported. In a press release, first author Kale Scatterty, a PhD student at the University of Alberta, said: “Infrasound exposure had effects on both outcomes that went beyond that natural relationship” between irritability and cortisol.

The cortisol measure also appeared sensitive to the emotional tone of the music. In the no-infrasound groups, calming music lowered cortisol and unsettling music raised it; when infrasound was present, that calming effect appeared to reverse.

Could Infrasound Be an Invisible Pollutant in Everyday Environments?

Researchers pointed to a possible biological mechanism. Some fish species detect and avoid infrasound through tiny stone-like organs in their inner ear responsible for balance. Humans have similar structures, and while our hearing works differently, researchers theorize the body’s balance system, closely linked to brain regions involved in emotion, may register infrasound without conscious awareness. Direct comparisons between animal and human responses, the researchers cautioned, “cannot yet be reliably made.”

The infrasound level used, 75 to 78 decibels at around 18 Hz, was chosen to mirror documented real-world conditions near ventilation, heating, and mechanical systems. The study tested only this one frequency, so the findings should not be generalized to infrasound broadly.

With only 36 participants, mostly female undergraduates at one Canadian university, exposure lasting under five minutes, and the proposed vestibular explanation still speculative in humans, these results need replication with larger and more diverse groups before anything conclusive can be drawn.

Still, the broader question the study leaves open is hard to ignore. If a brief lab exposure can nudge cortisol upward and shift mood in a measurable direction, what might hours of undetected infrasound do in offices, old apartments, or buildings long rumored to be haunted? Schmaltz, who studies pseudoscience and misinformation, put it directly: “The next time something feels inexplicably off in a basement or old building, consider that the cause might be vibrating pipes rather than restless spirits.”

---

Disclaimer: This article is based on a peer-reviewed study, but the findings are preliminary and should not be interpreted as established medical or scientific consensus. The research involved a small sample and a single brief exposure under laboratory conditions. Results may not reflect real-world outcomes and require replication in larger, more diverse populations before broader conclusions can be drawn.

---