Estudio de la posición de F1, F2 y F3 en situaciones de coarticulación empleando técnicas de predicción lineal

Abstract

This article shows several formant maps obtained from speech signal recording produced using articulated VCV situations. The speech samples have been computed using Linear Prediction algorithms (LPC), obtaining the three first formants. Later, three-dimensional and two-dimensional maps have been created in order to show the formant positions and evolutions. In this study we have used four speakers and more than 3000 formant position frequencies, tying to validate the original developed methodology and tools.