Experiment of The Month

We are surprisingly good at locating the source of a sound. This month's simple physics experiment illuminates one of the mechanisms that we use to extract direction from the sound that we hear. Dr. Mizumski is teaching our musical acoustic course (Physics 205) for the first time this semester and, as often happens in such a situation, is developing new experiments for the laboratory component of the course.

MODIFIED STETHESCOPE

 

TAPPING THE TUBE

The figure on the left above shows an ordinary stethoscope, with the diaphragm replaced by a length of tubing. The tubing used was a hard polyethylene, Dekoron 1219 FR (flame retardant) tubing, recommended for pneumatic control applications. Because of the hard character, the tube makes a clear tap sound when struck with a pen, as shown in the picture on the right above.

When tapped as shown, the student hears the sound to be coming from his left ear, even though sound of equal loudness is reaching his right ear. If the tapping pen is moved closer to the left ear than illustrated, the sensation remains the same: The sound does not sound "closer and closer," or "more and more on the left."

If the student moves the pen to his right, he hears the sound to come from straight ahead, and then, as the pen crosses the midpoint, it quickly changes to "coming from the right." Dr. Miziumski has students measure the distance along the tube between the "left" and "right" sensations.

Throughout the experiment there is essentially no change in the perceived loudness of the sound, but the perceived direction of the sound changes dramatically. Interestingly, to change the pen tap from sounding left to sounding right, the pen is moved a distance roughly equal to the diameter of the experimenter's head.

This evidence indicates that our direction perception does not depend on loudness, at least at the relatively low frequencies generated by tapping on the tube. The tap sound is noise, far from a single frequency, so it is unlikely that the direction is extracted from frequency information. With amplitude and frequency eliminated, the remaining wave characteristic is phase. Can it be that we detect phase shifts?

The answer is evidently yes, although, it is easier to believe if we think of the time shift equivalent to a phase shift. We can understand this experiment by assuming that the ear which first receives the sound "declares" the direction. It may seem remarkable that we can detect time differences as small as this (the time for sound to go 20 cm in air being about 0.5 milliseconds), but it seems to be true. Indeed, there is evidence that humans can detect differences as small as .01 milliseconds.

An interesting variation is illustrated at the right. The tubing has been twisted so that the part being tapped is closest to the student's left ear. The fact that the sound seems to come from the left verifies that the ear is not using some other sound path, or a visual que, to decide the direction of the source. In this case, the ear is fooled, and would instruct the eyes to look to the left for the source of the sound.