Voice transformation algorithms with real time DSP Rapid Prototyping tools

The main goal of the work here described is the DSP implementation of innovative algorithms for voice transformation in real time. This work represents part of the procedure conceived for the reconstruction of voice and dialogue in audio tracks of old and highly damaged film movies that has been developed within the framework of the RACINE-S European Project. In addition to the implementation of a set of methods, as LPC/ VC (Linear Prediction Coding/ Voice Conversion) techniques, innovative non-linear algorithms for improving the quality of synthesized speech have been proposed. This entire set of operations represents a particular implementation of the so-called "Virtual Dubbing" procedure. The main steps of the complete project include: a method for high-quality voice transformation, designing a suitable algorithm in Matlab/Simulink and, finally, translating it into Digital Signal Processor target code by means of a rapid prototyping approach. The original code was developed on Matlab and so we have used the Mathworks MATLAB's Real Time Workshop (RTW) DSP platform for rapid prototyping. In fact, with the advent of the Real Time Workshop, is now possible to compile, load and execute graphically designed Mathworks Simulink models into a real DSP platform. RTW supports many Developer Starter Kit DSP boards, including the pretty fairly Texas Instruments C6000 series that has been used for the present work.