TKK Aparat: An environment for voice inverse filtering and parameterization

Abstract

The study of the glottal flow, the acoustic excitation for voiced speech, provides insight into the voice signal, which is of potential benefit in many disciplines. One common method for estimating the glottal flow is inverse filtering, in which the effects of the vocal tract and the lip radiation are removed from a microphone signal. This paper presents a new inverse filtering and parameterization software package, which is available under an open-source licence. It provides a user-friendly graphical interface for rapid inverse filtering and parameterization, and the algorithms and parameters can be easily re-used in other projects. The system has already proved to be useful in algorithm development, speech science research, as well as in the study of occupational voice.