A Discussion regarding AI backing tracks
Posted: Sat Jul 26, 2025 9:34 am
Question. I genuinely would like a discussion. I'm not here to bash (or promote) generative AI.
This week at song circle (in person, local) one of the lyricists that used to attend quite a bit came in with a Suno.ai generated track and it was received pretty darn warmly by the group. The hitch, though. She was able to sing an original lyric with an original melody, upload that recording and get a fully produced track from it. She showed us the memo recording and, obviously, the finished track (which sounded great, btw.) It was amazing (in a weirdly mix of good and off putting ways.) It was an AI voice singing the track, a full band playing a southern gospel song with the result that, over a bluetooth speaker, sounded very impressive.
Additionally, this woman usually spends about $250 per song for someone to do a demo of her music. A couple of our local songwriters use outfits in Nashville to produce pretty good sounding demos (and sometimes spend more to pitch them to artists.) She subscribed to Suno.ai for around $200 for the year (and nearly unlimited song renderings.) So economically, for her, it was a no brainer. (I'm guessing the upload demo option is a premium feature.)
So the question is: How do you feel about an original melody and lyric being put into a generative AI tool in order to create the backing track and final mix? Is it much different than the generative drums/bass/backing in some DAWs? How do you feel about replacing a vocal with an AI generated voice. Is there a difference between using these tools separately and putting them together and just uploading a demo and pushing a button?
This week at song circle (in person, local) one of the lyricists that used to attend quite a bit came in with a Suno.ai generated track and it was received pretty darn warmly by the group. The hitch, though. She was able to sing an original lyric with an original melody, upload that recording and get a fully produced track from it. She showed us the memo recording and, obviously, the finished track (which sounded great, btw.) It was amazing (in a weirdly mix of good and off putting ways.) It was an AI voice singing the track, a full band playing a southern gospel song with the result that, over a bluetooth speaker, sounded very impressive.
Additionally, this woman usually spends about $250 per song for someone to do a demo of her music. A couple of our local songwriters use outfits in Nashville to produce pretty good sounding demos (and sometimes spend more to pitch them to artists.) She subscribed to Suno.ai for around $200 for the year (and nearly unlimited song renderings.) So economically, for her, it was a no brainer. (I'm guessing the upload demo option is a premium feature.)
So the question is: How do you feel about an original melody and lyric being put into a generative AI tool in order to create the backing track and final mix? Is it much different than the generative drums/bass/backing in some DAWs? How do you feel about replacing a vocal with an AI generated voice. Is there a difference between using these tools separately and putting them together and just uploading a demo and pushing a button?