This dataset was recorded by 2 speakers with authentic pronunciation and distinct vocal qualities (1 male and 1 female) in a professional recording studio. The recorded texts span the full range of phonemes, and the annotators possess a professional linguistic background, ensuring the data meets the requirements for research and development in voice synthesis.