Taos Hum Investigation



The following is a paper written by Joe Mullins and Jim Kelly of the University of New Mexico November 22, 1995.

Media attention has focused recently on individuals in and around Taos, New Mexico who perceive an irritating low frequency sound that has come to be known as the "Taos hum"

There have been persistent complaints of annoying low frequency sounds in the USA and other countries but the group of "hearers" in Taos has been particularly outspoken. In 1993 they got the attention of the New Mexico congressional delegation resulting in a request for an investigation. Many Taos residents feel that the government may somehow be involved in the sound, so the investigation was open and public.

One of the authors, Joe Mullins of the University of New Mexico served as the team coordinator. The team included Rod Whitaker (Los Alamos National Laboratory), Mark Leher (Phillips Air Force Laboratory) and Horace Poteet (Sandia National Laboratories) James Kelly, a hearing research scientist from the University of New Mexico's Health Sciences Center, joined the team later to provide expertise on the human ear.

Based on interviews with hearers, we determined that any proposed explanation for the hum should take into account its principle attributes.

Hearers who are musicians identify the hum as a modulated tone near Eb (41 Hz) and those with some technical background identify it as an amplitude modulated sound with a carrier frequency ranging from 30 to 80 Hz. We decided to supplement these anecdotal descriptions of the hum with psychophysical matching tests using known hearers before we proceed with our search for the source.

The hum matching tests were made with ten hearers. The equipment was set up in the bedroom of a guest house near the Taos residence of Bob and Catanya Saltzman, two hearers who had been instrumental in getting the investigation started. We built a custom microphone from an 18 inch woofer to serve as an ultra-sensitive low frequency sound detector. (-60 dbSPL from 8 to 80 Hz).

The output of this "big ear" was fed to recording and analyzing equipment housed in an RV some distance away. To keep the tests as simple as possible, we used a modulated sine-wave source and a low frequency speaker to generate matching sounds. Eight of the ten hearers were able to generate a sound which matched the hum they were perceiving . An individual hearer matched the same sound to the perceived hum quite reproducibly. Surprisingly, many hearers reported beats between the speaker-generated sound and the hum implying the physical existence of a second tone.

The results of these tests were both intriguing and puzzling. The carrier frequencies selected by the hearers ranged from 32 Hz to 80 Hz. The modulating frequencies they chose ranged from ½ to 2 Hz. All chose amplitude levels near or below the known sensitivity of the ear in this frequency range. But the most striking thing is that hearers could still perceive the hum in the presence of a relatively flat noise spectrum when the matching signal was turned off.

Fig 1 shows a representative sample of the noise background in the Saltzman's rural home at 7,000 feet above sea level recorded with our "big ear". The noise amplitude when converted to 1/3 octave bands, is 15 to 20 db below the accepted threshold for hearing in this frequency range. This suggests that the hearers are not extracting the Taos hum stochastically from acoustic noise.

The Saltzman's residence and the Taos golf course were selected as sites to conduct acoustic, geodynamic, magnetic and electromagnetic measurements. Additional geophone and magnetometer data were collected at other locations in and around Taos to track down the origin of a strong harmonic component in the 60 Hz power grid around Taos. At first glance this provided a possible source for the hum but hearers agreed almost universally that the perceived sound was unusually intense at a very remote site near Tres Piedras were we could find little trace of the 60 Hz magnetic component.

Measurements made with magnetometers in an electromagnetically quiet area did show the Schumann resonances created by lightning-induced electromagnetic radiation trapped in the spherical wave guide formed between the ionosphere and the earth. The observed resonant frequencies of this waveguide cover the frequency range implicated in the Taos hum but they appeared to be in the normal amplitude range and we could find nothing unusual about them.

Taos is a relatively quiet seismic area so it is not surprising that geophone data showed nothing but background noise. One massive event turned out to be a determined gopher restoring a burrow obstructed by placement of the geophone. Our electrostatic field detector recorded many events but all could be attributed to lightning strikes or movements of people in the vicinity. The electromagnetic data were collected from 20 Hz to 18 Ghz using an assortment of antennas and receivers. Everything that was found that could be attributed to 60 Hz power lines, electromagnetic noise or known radio and TV transmitters.

Since the US Navy ELF stations in the Michigan peninsular and in Wisconsin were widely suspected to be the source of the hum, we looked carefully at the 65 to 75 Hz region of the electromagnetic spectrum, the frequency range where ELF signals would be prominent. We could find nothing above the noise. Considering the spread spectrum of this broadcast, its limited antenna length and the distance from the transmitting site, it is unlikely that the signal would be above the noise level in northern New Mexico.

As a result, we are left with a mystery. There are no acoustic signals that might account for the hum nor are there any seismic events that might explain it. There are no unusual lines at suspect frequencies in the electromagnetic spectrum recorded near Taos. In fact, other than the signals generated by the power grid or in the case of the golf course, a power generator located at the course headquarters, we found no clear lines at all in this spectrum.

At a more fundamental level, there is no known mechanism whereby weak electromagnetic signals might even be transduced selectively by the human ear and perceived as sound. Our next step, therefore, was to shift our emphasis from the physical environment to the hearers themselves.

We first wanted to determine what fraction of the people in the Taos region are affected by the hum and to probe the characteristics of their perception. To this end we sent survey forms concerning the hum to 8,000 residents in Taos and Ranchos de Taos and the neighboring communities of Tres Piedras, Questa, Eagle Nest and Penasco. These communities represent respectively points to the west, north, east and south of Taos on a circular perimeter roughly 40 miles with Taos at its center.

A total of 1,400 residents in these areas responded to the survey and 161 respondents were classified as hearers. Most of the hearers (72%) are between 30 and 59 years of age and there are approximately equal percentages of male (52%) and female (47%) hearers. The hum is perceived by most hearers (62%) between 8 pm and 9 am. About 80% of the hearers perceived the hum at least one a week and, as shown in fig 2 a hearer's first experience with the hum usually occurs in the immediate vicinity of Taos. It is noteworthy that 19% of the hearers perceive the hum at sites more than 50 miles from Taos.

If we assume that all hearers responded to the survey, we can estimate that the lower limit of the fractional number of hearers is about 161/8000 or approximately 2% of the population. If this ratio hold for other populations in remote or quiet rural areas we would expect that the number of hearers is very substantial in this country.

It has been proposed that the hum might represent some kind of low frequency tinnitus arising from a disturbance in the cochlea of the inner ear. One problem with such explanations is that tinnitus itself is not well understood. Most individuals with tinnitus match the tone they perceive between 3 to 6 kHz and rarely if ever does a tinnitus sufferer match to a tone below

1 kHz. Why should such a phenomenon skip from regions of the cochlea where 3 to 6 kHz tones are represented over intermediate zones to the extreme apical end of the cochlea where the lowest frequencies of sound are represented ? Furthermore, many hearers insist that they can hear beats between their perceived tone and the objective tone generated in matching experiments. Beats have not been demonstrated conclusively in matching experiments using subjects with tinnitus.

Our next goal is to compare low frequency hearing thresholds and other aspects of auditory function in hearers and non-hearers from the Taos region. Informal testing of one hearer provided a tantalizing hint that the sensitivity of hearers in the frequency range from 20 to 100 Hz may be significantly more acute than is implied by some of the widely used sensitivity curves. There we are currently developing equipment for low frequency audiomety as well as equipment to measure distortion product otoacoustic emissions generated by the ear at low frequencies.

The measurement of these emissions has become standard clinical practice at frequencies above 1 kHz but to date no one has reported success in detecting evoked emissions or spontaneous emissions in the frequency rang we are investigating. The problem is that the noise floor for the usual insert microphones is very poor at low frequencies so we are using electrodynamic elements which have orders of magnitude better noise performance.

In summary, our interest remains high in finding a source for the elusive Taos hum. The character of the Taosenos and their unique town only add to the flavor of this mystery in acoustics and human hearing.

Back to Contents