how to remove vocal bleeding?

need help to remove some vocal bleeding. What i mean by this is that the deep frequency from the vocals are peaking through into the backtrack and it ruins the track entirely. I have tried to use noise reduction but it does not work. I’ve searched around through youtube to see if users have used tricks to remove this but no luck. Please help thanks

I have tried to use noise reduction

Noise Reduction seems to be a get-out-of-jail card, but it only works on sounds that don’t change over time. It’s perfect, for example, if you have some microphone hiss in the whole show. You select some “clean” hiss as the Profile, and Noise Reduction will try to remove that exact hiss from the whole show. It’s helpless in the face of sounds that move or change over time, such as conversations, music or traffic.

peaking through into the backtrack

To rephrase that, you’re trying to make a backing track from a song so you can sing to it instead of the original singer. You used one (or more) of the Vocal Reduction tools and the original voice didn’t vanish.

That’s normal. The Vocal Reduction tools, or the split and flip technique, depends on the lead singer being in the exact, left-right, dead center of a song and no other sounds or instruments are in there with them. The singer can’t have stereo effects added to their voice, because the stereo effects will always bleed through. That’s what happened to this clip.

Did you try Effect > Vocal Reduction and Isolation? That tool has additional adjustments and settings to help retain bass and drums which usually vanish, too.

So what you have is normal. Not all songs work with the vocal reduction tools. If you got this technique from a YouTube video, please note the publishers of the instructions never do it to more than one song.

Some song publishers offer karaoke versions of popular songs. Check your music supplier or Google.


I think this way of removing these vocal bleeding parts are 100% impossible, i used a different software known as phonic mind which uses an AI to remove these vocals, it does good job keeping the vocals as a separate audio file and keeps the song without the vocals in good check in stereo or mono, but that issue is what is being brought up here, i was asking to see if you guys knew how to get rid of this issue. the other issue is the audio is a mono file.

other issue is the audio is a mono file.

Then that’s the end of the world. As far as I know, the Audacity tools need a real, clean, uncompressed stereo file.

i used a different software known as phonic mind

If you look back at your original posting, you suggest strongly that you found a shortcoming in the Audacity tools. No mention of a second software package. We only go with what you tell us.

Thanks for the software name.


One other point. The tools start falling apart if your original is compressed or has processing artifacts—such as an MP3. MP3’s job is to produce small file sizes by damaging the original in ways difficult to hear. That damage hangs on like a headcold messing up post production and master file deliveries.


i messed up badly didn’t i?