Birds Filter Frequencies from Songs
May 22 2003

The pure-tone quality of bird songs may be due to a sound filtering mechanism, according to “Pure-tone birdsong by resonance filtering of harmonic overtones.” Many bird species vocalize in single frequencies, devoid of harmonic overtones. However, the mechanism for these pure-tone songs is unknown. Several theories offer an explanation for this phenomenon, varying from a lack of overtone production at the syrinx, the sound's source, to a filtering mechanism that removes harmonics as sound is projected. To test these theories, Gabriël Beckers and colleagues examined the vocalizations of two species of turtledove. By surgically implanting a series of sensors, the scientists were able to record sound signals very close to the syrinx. Each sensor was wired to a recording device that the birds wore as backpacks. The researchers found that sounds emanating from the syrinx appear to have a multifrequency spectrum, with several harmonic overtones. As the final sound produced by the doves is a single frequency, these results suggest the existence of a filtering mechanism that removes harmonics. Because humans also filter a multi-frequency sound source when vocalizing, the scientists suggest that the birdsong mechanism may be akin to human speech.

“Pure-tone birdsong by resonance filtering of harmonic overtones”
by Gabriël J. L. Beckers, Roderick A. Suthers, and Carel ten Cate, Published online before print May 22, 2003 Proc. Natl. Acad. Sci. USA , 10.1073/pnas.1232227100

Gabriël J.L. Beckers, Medical Sciences, Indiana University at Bloomington

The figure at left shows a spectrogram of a ring dove coo, with frequency (Hz) on the vertical axis and time (seconds) on the horizontal axis. Researchers often use spectrograms to study bird song. See, for example, Starling Song.

The top trace shows the coo as recorded in the room; the bottom trace shows the coo as recorded near the syrinx. Overtones visible below appear to have been filtered out, above. The hissing more noticeable in the syrinx recording is electronic noise, not part of the call.

Use the QuickTime controller at the bottom of the figure to hear first the sound of a coo recorded in the room (top trace), then at the syrynx (bottom trace).

Recordings and spectrogram courtesy of Gabriël J.L. Beckers, © 2003.

QuickTime Movie assembled by Stephen Hart.