The main goal of the work here described is the DSP implementation of innovative algorithms for real-time voice transformation. This work represents part of the procedure (developed in the framework of the RACINE-S European Project) conceived for reconstructing voice and dialogue in audio tracks of old and highly damaged film movies.