Spectral findings for vowels [a] and [ã] at different velopharyngeal openings.

Abstract

Background: The size control of velopharyngeal opening is an important variable for the acoustic profile characterization of hypernasal speech.

Aim: To investigate frequency spectral aspects of F1, F2, F3, nasal formant (FN) and anti-formant, in Hertz, for vowels [a] and [ã] at different velopharyngeal openings produced in the bulb of a palatal prosthesis replica used by a patient with velopharyngeal insufficiency.

Method: Speech recordings were obtained for four words ("pato/mato" and "panto/manto") produced within a carrier phrase in 5 conditions of velopharyngeal functioning: prosthesis with no openings (control condition: CC); prosthesis with bulb opening of 10 mm² (experimental condition with 10 mm² opening: EC10), prosthesis with a 20 mm² opening (EC20), prosthesis with a 30 mm² opening (EC30), and without the prosthesis (ECO). Five speech-language pathologists made a live rating of speech nasality during the reading of an oral passage. The recordings were used for spectral analysis.

Results: F1 values were significantly higher for [a] when compared to [ã] in all conditions. F2 values for [a] in EC20 and EC30 were significantly lower than values in the other conditions, being closer to the values presented for [ã]. F3 values were not significantly different between the testing conditions. There was a relationship between FN and anti-formants, and the auditory perception of nasality for conditions EC10 and EC20.

Conclusion: Significant changes were observed in the studied spectral values according to changes in the velopharyngeal opening size.