Singing Voice Conversion Between Popular Music and Chinese Opera Based on VITS

Singing Voice Conversion is an audio processing technique designed to convert one singer's voice into another singer's voice, while preserving the singing characteristics and emotional expression in the original audio. Although this technol-ogy has been widely used in the field of music, attempts in the field of Chinese opera are very scarce. This paper attempts to transfer the timbre of popular singers to Chinese opera through the neural network-based vocoder model VITS to replace the original opera timbre, and finally achieve the effect of popular singers singing Chinese opera. The experimental results show that the generated audio can better maintain the timbre of popular singers, and can clearly feel the auditory characteristics of Chinese opera.