Disentangling two male speakers in one track

I’m wondering if it is possible at all to disentangle two presenters (speakers) speaking at the same time in an mp3 recording? The track I have is a stereo mp3 file recorded using my Mac’s built-in mike. There’s almost no background noise. In the mp3 file an English-speaking guy was translated simultaneously by a Russian interpreter. I need to somehow "gate out’’ or filter out the interpreter from the speaker. Is it possible at all? And, by the way, I don’t care about the resulting quality. I only need to hear the original speaker to make to transcribe his speech. Thanks.

Probably not.

Software has no intelligence. All it knows is the show has a series of sounds. At best it may be able to sense a human presenter/announcer with a particularly rich and complex voice. One presenter.

There’s a fuzzy rule of thumb that if you can’t understand the presenter in the original show, that’s the end.


I did a job where I needed to capture a room of people giving notes on a theatrical performance. The prevailing wisdom was just make a simple recording like you did and go home. I insisted on a top quality stereo capture of the review and later, it was possible to clearly hear and understand multiple conversations going on at the same time by wearing stereo headphones.

A simple mono recording, even a good quality one would have been useless.