Introduction
Parabolic microphones are known for their extreme sensitivity, and the origin of their acuity isn’t difficult to guess. It is the most obvious thing about them, which can also make them a liability for field use, namely, their considerable size. Just as a large amount of weak light is captured by a telescope’s parabolic mirror and made amenable to viewing with the much smaller human eye, so too are faint sounds harvested with a reflecting dish that far exceeds the dimensions of our native auditory tools. In both cases, it is a matter of intercepting more of the signal and bringing it to a focus, and the same basic wave physics is at play.
A good overview of the early adoption of parabolic microphones for the field recording of bird sounds was provided by Wahlström (Wahlström, Sten. “The Parabolic Reflector as an Acoustical Amplifier.” Journal of The Audio Engineering Society 33 (1985): 418-429). He describes how Peter Paul Kellog and others first used a 32-inch dish to capture avian vocalizations, with a Yellow-breasted Chat having the honor to be the first species recorded with their novel device.
That a large dish may be cumbersome to transport and use in the field isn’t the only caveat. Users of parabolic microphones soon learn that these devices have a bias that favors high frequencies. Where this is most evident is in the “tinny” quality to the captured sound, which some listeners find off-putting. Others find this drawback significantly outweighed by the benefits of hearing otherwise inaudible sounds, and the boost in signal strength that comes with the high directionality.
The flip-side is that the performance drops off for low frequencies, with a cut-off point below which no benefit is derived. That is, the microphone will capture the same signal power whether the dish is present or not, which may seem curious. This cutoff frequency is given by
,
where V is the velocity of sound and a is the diameter of the dish. For a=32 inches, this works out to roughly 421 Hz. Most bird vocalizations lie well above this frequency, but we can easily calculate how this limitation will pour cold water on any hopes of wielding small, unobtrusive dishes. An 8-inch parabolic, for example, would provide no gain below 1,500 kHz. The low hooting call of a Mourning Dove would not be be amplified by such a device.
The bias favoring higher pitches can be captured by an expression for the gain as a function of the relative change in frequency. In comparing performance at a frequency f versus a reference frequency f 0 , the gain G as measured in dB is:
.
For a change of one octave, that is, for a doubling in frequency, we have the log of 2, which is 0.3, and so the gain will be 6 dB per octave (provided that f 0 is above the cutoff frequency). To put this into perspective, a 10 dB change corresponds to a doubling of perceived loudness. A 6dB change therefore will certainly produce a noticeable difference.
... continue reading