Description

VoiceForger is a library and user interface for processing the voice. It is an expansion upon the type and range of effects included in SuperVP-TRaX. VoiceForger is based largely on the analysis and processing by the SuperVP library. Unlike TRaX, which uses the existing Max/MSP objects, VoiceForger directly uses the C++ interface to SuperVP for greater flexibility. With the recent advances in the modularity of the library SuperVP, VoiceForger remains effective and allows changes in real time even for multiple analyses (F0, Voiced / Unvoiced, VUF). The development of C++ version of this library began for the project: Affective Avatars. VoiceForger includes the transformations of age and gender of the speaker (also available in SuperV-TRaX). VoiceForger uses more advanced processing concepts and is more flexible than TRaX. This list contains the transformations: targeted ages and genders, types of voice (rough whisper, sucked), emotions (happy, irritated) and in future it is hoped to target personalities.

Main Functions

For the project Affictive Avatar, a basic architecture was created for the management of modules of analysis and sythesis using the SuperVP library, as well as automatic reconfiguration of the chain of processing according the the desired changes. Improved versions of the transformations available in TRaX were added: changes to the speech style (e.g trembling), the transformation of the vocal pitch, and the transformation into breathy or whispered speech.