To start, I know very little about audio manipulation, other than splitting a track into smaller parts I haven't done much, so and ELI5 answer would be preferred. I have an audio file, and I want to separate it into multiple tracks. The file has 4 people talking and some background noise. I'd like to separate it into 5 tracks, one for each person and one for the background noise.

Is there a (preferably free) program that would allow me to do this? If so, what would I need to do in said program to accomplish this?

Thank you for any and all help!


2 Answers 2


I'm afraid what you are asking for could be the equivalent of taking a Victoria sponge cake & asking for the flour, sugar, butter & eggs back in separate bowls.

If there are clear spaces between each person speaking, you have a chance; any audio editor such as Audacity [freeware] can be used to [manually] split the audio at each gap, then move each speaker to a separate track. You could then save each out as a separate audio file, containing just that one speaker.
If everybody is speaking at once, you're back to Victoria sponge.

The background noise is in itself a difficult task. There is very expensive software such as Izotope's RX5 [& cheap stuff that is nowhere near as good, for which I have no specific recommendations] that can potentially remove some/most/all depending on how regular it is. You could potentially bounce the removed noise to a separate track if you so wished - it's never been something I've ever needed or tried.

As extreme examples, an air-conditioner would produce regular-enough noise to be able to eliminated entirely, if it's not too loud. Someone shooting a gun would be next to impossible.


I know this is old, but I recommend Audionamix Trax Pro (Mac only). I was able to separate all the voices from a mixed Take 6 (vocal group) recording into separate files after several hours of use. My guess is as long as the people's voices in your file are clearly distinguishable, have different basic pitches/tone, or were all recorded from their respective positions in the same room using one same mic you'll have a very good shot at separating the voices -after MANY hours of tedious processing. If they were all recorded from one mic, you have the best chance for great results because Trax Pro allows you to isolate sound from specific parts of the stereo spectrum. My advice: Save often, it's very buggy and separate the file into short durations because it's extremely CPU intensive.

For the noise issue, I recommend using Waves NS1 to get rid of the background noise first, then import your cleaned up audio track into Audionamix Trax Pro and cancel your plans over the next few days while you try to 'trace' each voice in the spectrogram. Good luck!

NS1 Demo

Audionamix Demo

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.