Interesting suggestion.

I would argue that a decent producer (not me lol) and with the current tools we have can actually produce a flawless vocals removing all 'natural' hiss breathes and other to have a perfect voice. No AI required. Is it good? Is it bad? Is it too much production to the point you actually kill the vocal? Maybe. But it is achievable. Merely Photoshop for audio.

As for myself, I already said before that when I don't have the ability or time to record the singer, what I would do, in order to have a prototype to present, I would sample the singer's voice and then play with that with different directions and then with this reference we would go and record when we have the result that we want and food feedbacks for the direction.

Check the previous Blue Hour release, I would love to hear you feedback on that.

 

Last edited: Mar 8, 2026 at 6:14 PM