If the voice is buried or overpowering, it's a balance problem between the vocal and the backing track.
How to fix the balance
Prompt for vocals up front, pick a lighter instrumental, or use stem separation to control the levels yourself.
- Ask for vocals forward in the mix.
- Use a less dense backing.
- Use stem separation to rebalance.
Balance is set by the arrangement, not by a volume slider
There is no vocal fader in a text-to-music model. The relative level of the voice is decided by the arrangement it is placed in, which means the way to bring a vocal forward is to ask for a sparser mix rather than a louder singer.
Prompts that stack many instruments — layered synths, big drums, full strings — naturally bury the voice. Asking explicitly for prominent or intimate vocals, and naming fewer instruments, moves the balance more reliably than any wording about volume.
- Ask for a sparser arrangement to bring vocals forward.
- Say 'prominent vocals' or 'intimate vocal' in the prompt.
- Fewer named instruments means more room for the voice.