Ben Lange. Photo: Kate Callas

The didjeridu (or yidaki in the Yolngu language) provides Australia with an iconic musical sound, and acoustics with a puzzle: how is such a simple instrument – a eucalyptus trunk eaten out by termites – capable of such a huge range of timbres?

Obviously the vocal tract of the player is strongly involved, but how? To find out, we used technology invented in the Acoustics lab to make measurements of the player’s vocal tract – specifically its acoustic impedance spectrum – during playing. This requires injecting a carefully calibrated probe sound, comprising several hundred different frequencies, into the player’s mouth alongside the yidaki using a thin pipe. A microphone measures the response of the tract to these frequencies. The reason why this is difficult is that the sound level inside the player’s mouth is very loud - typically 100 decibels.

Our measurements showed that strong resonances in the vocal tract can inhibit acoustic flow through the lips. The strong formants (emphasised frequency bands) that are recognised in the sound of the instrument thus correspond to frequencies falling between the strong resonances in the tract. However, the tract resonances must be very strong, which means that players learn to keep their vocal folds nearly closed – in a position more like whispering than normal breathing, so the sound is not absorbed in the lungs.

Although vocal tract effects are more subtle in other wind instruments, they are still musically important. Consequently, this research (which was carried prominently in the journal Nature and the world press) also provides insight into brass playing, too.

One of the authors was Ben Lange (traditional name Wilamara), a member of the Mara people of Roper River in Northern Australia, where he learned to play yidaki in the traditional style. He was an undergraduate when he worked on the project as a vacation scholar.

Alex Tarnopolsky, Neville Fletcher, Lloyd Hollenberg, Ben Lange, John Smith
and Joe Wolfe