Take music out of audio
To take music out of audio for podcasts and lessons, speed matters: you’re usually processing lots of short clips. A good background music remover workflow is: trim → separate → quick A/B → export. If you try to perfect every second with EQ, you’ll lose hours and often lose voice quality too.
Fast podcast/lesson workflow (that doesn’t overcook the voice)
1) Trim to what you’ll actually use
Cut intros/outros, long silences, or repeated takes with Audio Cutter. This also makes separation cleaner and faster.
2) Separate first, then reduce
Run separation with AI music separator and bring the background down until speech is clear. For podcasts, “reduce” usually sounds more natural than “remove.”
3) Do a 10-second A/B check
Pick one tricky section (music + speech together), then:
Listen on phone speaker
Listen on earbuds
4) Keep loudness consistent
Podcasts feel “professional” when volume doesn’t jump between clips. Normalize gently after separation.
What to do when results still feel messy
Process in sections: quiet parts and loud parts separately
If the voice is distant, reduce music instead of removing it
If the music is very loud, try to get a better source audio first
If you’re unsure whether you should remove or reduce in the first place, follow the main overview: background music remover for clear voice audio.
Authority resources
NPR Training (audio basics)
EBU R128 overview (speech loudness)
FFmpeg documentation (exports)