Ok so I have a song: the instrumental part of it in the first verse is EXACTLY the same as the instrumental in the second. What I want to do is extract what those 2 tracks have in common, which is their instrumental element. SO what I do is i line the 1st verse up directly under the 2nd, invert one of them, and then i end up with just the vocals combined. i export this as an MP3. Then I undo the inversion so that i’m back to the combination of the 2 verses together with no effect on them. So I open a new project, load the track that has just the 2 sets of vocals and load the track that has the 2 unedited verses (vocals+instrumental), and line those up. INVERT the vocal-only track so that I end up with an instrumental, BUT what happens is I just end up with the first verse of the song completely unedited. What can I do to end up with JUST an instrumental aka what the 1st and 2nd verse have in common? Thanks so much for reading. Any suggestion is appreciated.
You need to stay in WAV land. MP3 compression causes damage messing up your cancellation.
Also, arethmetic cancellation like this needs two events to work. Amplitude and phase. So you slide the track back and forth and you may have to ampify up or down as well. This is one problem with Audacity not being a real-time production tool.
And when you get it exactly right you will CANCEL the sounds that are common to both.
It is not mathematically possible to isolate the sound that is common to both by summing or differencing the tracks.
ISOLATING the common sounds is much more difficult. There is a commercial VST plug-in called “Extraboy” that claims to be able to do it.
I couldnt get the Extraboy VST to work in Audacity at all, however its manufacturers produce a free standalone player “Player Boy”, (only accepts WAVs).
Ooh that’s nice - it’s free
Oh no, it’s just a free VST host and requires the non-free plug-in
Oh well, never mind.
The Extraboy VST is only $0.90 “bandwidth fee” … http://www.elevayta.net/fghdjertjy.php
but like I said you’re flogging a dead horse trying to get it to work in Audacity.
Well no-one can complain about it being expensive.
I think I would be flogging a dead horse to try and get it to work in Linux as well.
I apologise for my windows-centric suggestion.