Reviewing AI-generated song drafts on the TV/speakers via UMS – anyone else doing this?
Posted: Fri Oct 23, 2026 3:08 am
I have been experimenting with generated song drafts from my own lyrics, and I found that the best place to judge them is not the laptop speakers but the living room system, so I just drop the exported tracks into a shared folder and play them through UMS.
Before any of these drafts goes anywhere public I do a quick review pass: check that the lyrics going in were really my own, note which parts I actually directed (genre, vocal style, structure), and keep the first idea plus each revision so I can see how the song got where it is. I have been using a lyrics-to-song creation workflow for the drafts, which keeps the jobs and the generated tracks in a private library, and then UMS handles the listening part.
Two questions for people who use UMS mostly for audio:
- Do you keep a separate shared folder for work-in-progress music so it does not mix into the main library?
- Any renderer you found that is more honest for judging mixes (I am on a Sony TV + soundbar right now)?
Before any of these drafts goes anywhere public I do a quick review pass: check that the lyrics going in were really my own, note which parts I actually directed (genre, vocal style, structure), and keep the first idea plus each revision so I can see how the song got where it is. I have been using a lyrics-to-song creation workflow for the drafts, which keeps the jobs and the generated tracks in a private library, and then UMS handles the listening part.
Two questions for people who use UMS mostly for audio:
- Do you keep a separate shared folder for work-in-progress music so it does not mix into the main library?
- Any renderer you found that is more honest for judging mixes (I am on a Sony TV + soundbar right now)?