A method for changing speech quality and its application to pathological voices
نویسنده
چکیده
Speech quality is an interesting and very important aspect not only from linguistic and/or phonetic viewpoint but also from the viewpoint of speech technology. This study has been conducted from the latter point of view. A method has been developed based on the analysis-synthesis technique which enables to control voice quality by independently manipulating the voice source and the vocal tract resonant characteristics. Through this method, it is possible to investigate the amount of contribution of individual acoustic parameters to a certain voice quality including voice individuality. Formant frequencies and their bandwidths are used as the acoustic parameters to characterize the vocal tract configuration and the pitch frequency as the voice source. These acoustic parameters extracted from a natural speech are modified or changed to some extent and then a is synthesized making use of the modified acoustic parameters. Speech intelligibility and voice individuality are found to be controlled by this method. An application to a pathological voice has also been made to control the voice quality. It has been found that the method is capable of improving the so-called “roughness” or “hoarseness” of the pathological voice to a certain extent.
منابع مشابه
A Book to Scare and Stop Singing A critique of Larynx as Instrument and the Challenges to Vocal Production
The production of scientific resources in the field of recognition and application of sound is one of the most urgent necessities for Iranian society of voice training, speech and singing. In a society where the voices of its voice actors and speakers are unfortunately often harsh, piercing and, artificial, and where these types of voices have become a trend, the necessity of research in this a...
متن کاملMeasuring vocal quality with speech synthesis.
Much previous research has demonstrated that listeners do not agree well when using traditional rating scales to measure pathological voice quality. Although these findings may indicate that listeners are inherently unable to agree in their perception of such complex auditory stimuli, another explanation implicates the particular measurement method-rating scale judgments-as the culprit. An alte...
متن کاملJitter, Shimmer, and Noise in Pathological Voice Quality Perception
Although jitter, shimmer, and turbulent noise characterize all voice signals, their perceptual importance has not been established psychoacoustically. To determine which of these acoustic attributes is important in listeners’ perceptions of pathologic voices, listeners used a speech synthesizer to adjust levels of jitter, shimmer, and noise so that synthetic voices matched natural pathological ...
متن کاملLimited domain synthesis
This work presents a reliable and efficient method for building limited domain speech synthesis voices. By constructing databases close to the targeted domain of the speech application, unit selection synthesis techniques can be used to reliably give very high quality synthesis within domain. In addition to a high quality result we include the techniques and processes required to build such voi...
متن کامل[The effect of anchor voices and visible speech in training in the GRABS scale of perceptual evaluation of dysphonia].
INTRODUCTION Perceptual evaluation of voice quality remains a key standard for judgment of vocal impairment. The GRABS method has become a commonly-used scale for rating severity of dysphonia, but it has no published, standardised protocol to follow. Training is important for reaching good interrater agreement for its parameters; however, the references most often cited for the GRABS provide no...
متن کامل