Remove music from mp3
If you’re trying to remove music from mp3, the big problem is compression. MP3 can smear detail (especially cymbals and “S” sounds), and those artifacts confuse any background music remover—which is why MP3 often sounds “watery” after processing.
Why MP3 is harder than WAV/FLAC
MP3 removes audio detail to save space. When you later try to separate voice vs music, the tool has less clean information to work with—so you may hear:
Swirling/phasey background
Warbling on vocals
Harsh “S” sounds
Best practices that help immediately
1) If you can, start from a better source
If you have any option to get WAV/FLAC (or export from the editor before MP3), do it.
2) Separate, then reduce (don’t hard-remove)
Use the Music Separation tool or AI music separator and aim for “voice-first” balance.
3) Avoid re-encoding multiple times
Each MP3 save can add more artifacts. If you must convert, do it once, then edit.
Quick “MP3 artifact” fixes
If artifacts spike during loud music: process that section separately
If “S” sounds get sharp: gentle de-esser after separation
If voice feels thin: bring back a little music bed instead of boosting EQ
For the bigger remove-vs-reduce decision, use the main guide: background music remover for clear voice audio.
Authority resources
FFmpeg documentation (conversions)
Audacity manual (de-essing basics)