>>50317633>kilia fwmanon isn't monotone when asking for cumMany generated audio are like that in current year, AI tech is progressing at a scary pace. You'd easily find those posted somewhere like /gif/ or that thread in which that mori horse audio spawned. Even if that was outdated, some had decent cadence already at that time.
That being said, regarding picrel, it's too perfect in generating audio that Spectralayers shows me that there are vocal harmonics at the 12k to 14k range.
Most clear VOs only has it clear up to 8k to 10k, depending on their audio setup and beyond that it's just splotches.
Recording software and vocal chords, no matter how good it is, can't just replicate vocal harmonics detectable in a spectrogram at a scary 14k Hz.
Also, it pronounced face as fes.
