This pilot study explores the acoustic characteristics of two differing registers used in singing. A comparison is made of the acoustic source characteristics of sung notes in head and chest register produced by two male singers, a bass and a tenor, covering a range of fundamental frequencies. Head register tones were extracted from descending octave scales whereas examples of chest register were extracted from ascending octave scales. Samples of two different vowels at two different pitches are compared. Glides on an open back vowel in the lower pitch range of both voices were also analysed to facilitate the examination of the way in which various voice source parameters vary as a function of fundamental frequency. The principle analysis technique used inverse filtering of the sound signal to establish the voice source waveform (the differentiated glottal flow). In order to quantify the results of this inverse filtering, a model of the differentiated glottal flow (the LF model) is matched to the output of the inverse filter. Interactive software, developed in the Phonetics Laboratory in C.L.C.S., Trinity College, Dublin is used, and allows a pulse-by-pulse analysis of the voice source. Results show that the mode of vibration of the vocal folds differs in some important ways between head register and chest register in singing. The use of inverse filtering and LF model matching to analyse the singing voice has implications for many areas of voice science including singing training and high quality speech (and singing) synthesis.
|Title of host publication||The 3rd International Physiology and Acoustics of Singing Conference|
|Editors||D. M Howard, J. Brereton, H. Daffern|
|Publication status||Published - 2006|