CandyVoice's voice imitation, which allows multiple Target Voices to be imitated by multiple Source Voices, works in real time. This process requires the recording in a studio of a corpus of about a hundred sentences by the interpreters of the Source Voice and the Target Voice. The Target Voice can also come from an archival audio file.

CandyVoice creates supervised voice models of Source and Target Voices from studio recordings, or from an archival audio file. The quality of the voice patterns depends on the quality of the recordings (pronunciation of sentences and acoustic environment), or, if applicable, on the quality of the archival audio file.

CandyVoice's technology allows multiple Target Voices (including personality voices) to be imitated in real time by multiple Source Voices (and vice versa!). This technology is used, for example, in entertainment and video games, where the player can animate the game character with his own voice in real time.

Voice imitation by CandyVoice