An articulatory study of fricative consonants using magnetic resonance imaging

Abstract

Magnetic resonance images of the vocal tract during sustained production of the fricatives/s, •, f, 0, z, 3, v, 6/by four subjects are analyzed. Measurements of vocal-tract lengths and area functions, and morphological analyses of the vocal tract and tongue shapes for these sounds are presented. Interspeaker differences in area functions are found to be greater in the pharyngeal cavity than in the buccal cavity with the nonstriden: fricatives exhibiting greater differences than the strident ones. The anterior tongue body of the alveolar stridents exhibit concave cross-sectional shapes while that of the postalveolars show a relatively raised tongue body with fiat or slightly convex cross-sectional shapes. The concave tongue shapes of the alveolars result in a more abrupt area function behind the constriction when compared to that of the postalveolars. Laminality or apicality of articulation is found to be speaker dependent. Moreover, a greater degree of anterior roedial grooving and lateral lingua-palatal contact is found in apical alveolar fricatives than in laminal ones. The posterior tongue body of all fricatives shows concave cross-sectional shapes. Voiced fricatives are characterized by larger pharyngeal volumes than the unvoiced fricatives due to tongue-root advancement. Tongue-shape asyr•metries arc found to be subject and, in some cases, sound dependent. ¸ 1995 Acoustical Society of America.

3D Vocal Tract Reconstruction Using Magnetic Resonance Imaging Data to Study Fricative Consonant Production

“…Excluding the male subject for the consonant /f/, all pharynx volumes measured were superior to the oral cavity volumes. The study in [14] also found larger pharyngeal volumes in voiced fricatives by means of vocal tract lengths and area function measurements. Equivalent results were also reported in [16] where the authors observed some supraglottal volume changes of pharyngeal articulation during the production of voiced and voiceless fricative consonants.…”

“…The task of modelling the vocal tract for articulatory synthesis systems aims to obtain the complete geometrical information concerning the vocal tract measured by different techniques, especially MRI [11,14]. Currently, the use of MRI in speech research provides useful and accurate qualitative and quantitative data of speech articulation [21,22,23,24].…”

“…Several researchers also used this approach, due to the common MRI--related constraints: acquisition times, costs and fatigue of the subjects under study. In order to improve vocal tract modelling, some techniques have been proposed that combine manually [1,4,14] or even automatically [26] orthogonal image stacks.…”

Evolution of vowel production studies and observation techniques.

“…In order to establish the relationship between vocal tract shape and speech sounds, the lateral view of the vocal tract is insufficient, and the 3D vocal tract configuration must be obtained. Since magnetic resonance imaging (MRI) is capable of the 3D measurement of the body in multi-slice images, this technique has been applied to direct measurement of the vocal tract shape and area function [38][39][40][41][42]. Although the MRI provides a powerful visualization technique for speech research, it has a few problems; low signals from the teeth make toothless images, supine position of the subjects can affect natural articulation, and intense machine noise inhibits highquality audio recording.…”

A continuum mechanics representation of tongue deformation

“…MRI imaging has been used to reconstruct 3-D airway volumes for static sounds-vowels and fricatives [7,8,9]-but not to extract tongue features. The current MRI dataset consists of four midsagittal test images of the tongue: at rest, and saying /a /, /i/ and /u/.…”