One chorus line, "Stay with me till the light", sung by the voice alone with no music behind it, once at each setting. Each clip is about 8 seconds: the line, then silence so a reverb or an echo can be heard dying away.
Nobody has listened to these: they were made and measured by a program, not heard.
This is the voice as it is on the groovebox now.
The mouth moves quicker, with more of the s, t and k sounds and a little breath.
More of the talk box colour: slower mouth, sharper tone, fewer hisses and clicks.
The line sung by one voice only.
A second voice on every longer note of the line, a touch louder.
A room of the voice's own.
Repeats in time with the song.
The chord vocoder sings the same line: the song's chord in the line's rhythm, with no single tune.
In the groovebox: tap the lips, then Voice settings at the bottom of the box. Each one is kept on the device you set it on, and all of them act while a song is playing.