This dataset was recorded in quiet office/home environments, with the participation of 311 speakers, including 154 males and 157 females. All speakers involved in the recording were professionally selected to ensure standardized pronunciation and clear enunciation. The recorded texts cover news and novels.