Abstract: In this paper we describe the process of building a new Text-to-Speech (TTS) system in Spanish from the materials provided for the Albayzin 2010 challenge using the deployed Microsoft TTS technology in Spanish. The architecture of the system is described as well as the process of compiling a new voice font and producing a new synthetic voice. The main issues found during this process are presented and discussed.
Index Terms: Text-to-Speech, HTS, Spanish, voice quality, evaluation.