• Tachanka [comrade/them]
    ·
    edit-2
    2 years ago

    it's probably a neural network specifically. the narrator voice was best because that voice has the most existing data to draw from. Cuno was the worst because he has the fewest lines to draw from. Voice synth neural networks tend to be better the more voice clips they have. The narrator by far has the most voice clips and the largest vocabulary of anyone in the game's audio files.