Removing music and keeping vocals

I’m sure this has been asked before. I tested the Center Pan Removal and got just the music so I know that works. Now I’m wondering if there’s a way to just hear the voices themselves.

The reason for this is I have a movie that does not have subtitles, so I’m making my own. There are a few scenes where the music is much louder than the dialogue. I’d like to be able to silence the music. I’ve already been able to silence the vocals, so the opposite could certainly be done couldn’t it?

A qualified yes … How about removing the stereo pops from mono records? - #2 by Trebor
[Centre pan isolation can’t be done using the inversion, a.k.a. destructive interference method, of center pan removal.
another method, spectral subtraction, is required, which is what the Extraboy and Kn0ck0ut plug-ins use ].

shades of ross perot
i am all ears

IF they subtracted the vocal and have only the music
why cant you invert the music signal and add it to the original
to get the vocals only?

thanks , sort of
last time i read that topic
it was double talk

how about somebody, anybody, giving a straight answer in english

i know what that thread *** claims ***
but i see other forum on the net that say otherwise

i would like to see a valid explanation not just claims

anybody ???

I’ll have a go …

Left audio = L
Right audio = R
Centre audio = 2C

Left channel = L+C
Right channel = R+C

Remove centre audio: (L+C)-(R+C) =L-R (mono)

Subtracting the centre pan removed mono from the original stereo in an attempt at centre pan isolation gives …

Left channel (L+C) and Right channel (R+C) minus centre pan removed mono (L-R)
=> (L+C) + (R+C) - (L**-** R)
=> L + C + R + C - L + R {double negative gives “+R”}
=> 2C + 2R

centre isolated audio is “2C“, so “2C+2R” is not centre audio.

If you define “Remove centre audio“ as : (R+C)-(L+C) =R-L (mono)
=> 2C + 2L, (which still not “2C” on its own).

Here was my original post on the subject of “Why vocal isolation cannot be achieved through the same or similar techniques as centre pan removal”.

That post does not “claim” anything. It poses a simplified model of a mixing technique and invites the reader to think about it and come to their own conclusion.

Were they written by someone claiming to be Napoleon Bonaparte?

written by a recognised professional
published in a reputable print publication
and on their web

You are saying that a “recognised professional” (in what field?) has said what?

professional audio

says that IF you can extract either the music or the voice alone
you can invert it and combine with original to get voice or music alone

I can’t imagine that any respected audio professional would ever make such a basic mistake. If you wish to quote a “recognised audio professional”, please post a link to where they have published this information.

My concern here is not whether or not you have the last word, or whether you believe me or not, but whether other users will come to this forum looking for help, read your post and go away with the impression that there is some magical way to invert and mix tracks to isolate the centre panned audio. That would be an unnecessary waste of their time.

Unfortunately it is not possible to isolate centre panned audio through inverting and mixing.
There are some techniques (such as spectral subtraction) that are able to produce some degree of attenuation of centrally panned audio (as discussed previously). Audacity does not currently have this ability and requires additional plug-ins to do so. See here for more information:

This question has come up many times before, and you insist on arguing the point every time it comes up. Please check your facts.
If you believe that you are able to substantiate your point with an actual methodology that accomplishes this, then please do so, but if not, please do not persist with posting on this subject.


I said IF you could separate the vocals or music
then you could invert and get the music or vocals

I never said that you could separate it in the first place.
And the nice man never said how he did it. Just that if you had it you could invert it and combine to get the other part.

I stand by my answer. And the gentleman was correct.

IF you had one you can invert it and subtract to get the other.

If you can explain why (A+B) -B is not equal to A then I am all ears.

regret that i do not have time to cross index every website and all their pages for instant recall to show somebody where I saw something.

Obvious, but unhelpful in the context of this thread.

But that’s the crux of this thread. Just how do you separate the vocals and dump everything else?

The (constant) problem is that people hear that they can “remove” centre-panned vocals (through inversion of one channel then summing the channels), and then think they can isolate the vocals by subtracting that “non-vocal” version from the original. As Trebor and Steve have pointed out, it can’t be done.

Actually, you would invert then add.

You’re quoting an authority but you can’t tell us who that authority is. And from what you’ve said, this “authority” simply stated the obvious [(A+B)-A=B], and did not reveal how (or if) he actually managed to isolate the vocals from a stereo recording.

– Bill

going by memory

he , the best that i can recall, implied strongly that you could andor he did remove the vocals, and then got the music alone.

i cannot say that he did, and if he had claimed that i could not confirm that he actually did it.

there are commercial programs that claim to do this
realtek has its karaoke voice remover (at least on my version)

unlike most of our visitors who post questions, i try to be accurate complete and precise. do i sometimes fall short. maybe.
but i wont blame others should i ever might make a mistake.

Perhaps if this myth comes up a lot it might be worth adding a “Case 3” Two channel track with center vocals removed - can those vocals be isolated by inverting this against the original, so as to remove the non-vocals? to .

I’ve put it on my list to do, but would be glad if anyone beat me to it.