This paper details and evaluates some crucial aspects of an emergent family of coders: the sinusoidal model.

We present here an Harmonic+Noise Model (HNM) for speech synthesis. The noise part is represented by an autoregressive model whose output is pitchsynchronously modulated in energy. The harmonic part of the signal is represented by a sinusoidal model. This paper compares dioeerent methods for separating these two components. We then propose a method for the estimation of the sinusoidal parameters derived from the ABS model [8] and evaluate dioeerent models for the analysis/synthesis of the stochastic part proposed in the literature. INTRODUCTION Most text-to-speech systems (TTS) use nowadays concatenative synthesis where segments of natural speech are distorted by analysis-synthesis techniques in such a way that the resulting synthetic signal conforms to desired prosodic characteristics. Classically these characteristics include melody, sounds energy and duration. Modiøcations of these characteristics by Speech Synthesis Coders (SSC) are often accompanied by distortions in other tempo...

Creak, also referred to as vocal fry or glottal fry, is a perceptually distinctive vocal occurrence that is frequently produced in running speech. The presence of creak causes problems in speech processing particularly with f0 and spectral analysis. However, as creak frequently occurs in natural spoken communication and as it often carries important segmental and paralinguistic information we believe that applying appropriate processing methods for the study of creak in spontaneous conversation is warranted. This preliminary study demonstrates trends in the use of creak by 9 Hungarian speakers talking in both formal and informal settings. Creaky segments are identified and the distributions of these segments were examined in relation to frequency of occurrence in formal versus informal settings. The study shows trends of increased frequency of creak in informal settings and an apparent convergence in speaking style on the part of the interviewer with interviewees who produced creak frequently.

Abstract. This pilot study aims at defining the notion of voice quality in Western lyrical singing and seeks significant and objective criteria to characterise it, from both cognitive and acoustic points of view. We have chosen an approach based on the semantic analysis of experts ’ discourses, as we assume that the description of the acoustic signal itself cannot fully account for the characterisation of voice quality and that this notion rather relies on the listeners ’ cognitive representations, which allow them to elaborate a meaningful judgement from acoustic properties. Therefore we started this study with a listening test, conducted with 11 singing teachers

Voice quality, defined by John Laver as the characteristic auditory colouring of a speaker’s voice, is a significant feature of speech, and it is used to signal various properties such as emotions, intentions, and mood of the speaker. While voice quality measurement techniques and algorithms have been developed, much work is needed to obtain a comprehensive view of the function and analysis of human voice in the production of different voice qualities. Two major research questions are presented in this thesis: First, how can the most important laryngeal voice quality features be analyzed, and second, how do the voice quality features affect different facets of vocal expression? To answer these questions, five separate studies of the analysis methodology and two studies regarding the voice quality behaviour were

This paper describes a new algorithm for automatically detecting creak in speech signals. Detection is made by utilising two new acoustic parameters which are designed to characterise creaky excitations following previous evidence in the literature combined with new insights from observations in the current work. In particular the new method focuses on features in the Linear Prediction (LP) residual signal including the presence of secondary peaks as well as prominent impulse-like excitation peaks. These parameters are used as input features to a decision tree classifier for identifying creaky regions. The algorithm was evaluated on a range of read and conversational speech databases and was shown to clearly outperform the state-of-the-art. Further experiments involving degradations of the speech signal demonstrated robustness to both white and babble noise, providing better results than the state-of-the-art down to at least 20 dB signal to noise ratio. Keywords: Creak, Creaky voice, vocal fry, glottal source, glottal closure instant