The aim of this corpus was to investigate the implementation of a high-quality TTS system using multiple voices recorded using a low-cost process (i.e. using non-professional volunteers and informal recording environments).