Hey, there's this song I want to use for a project of mine as some background music, but I dont want the lady singing at all. I know there is hardware that can do this: a friend of mine has a stereo that has a "voice" volume knob that if he turns it down, all you hear is the music and no voices.

I don't know specifics on how he did it, but this is the kinda of effect I want to do to a certain song.

Just for FYI, the song is "Teardrop" by the artist "Massive Attack". I like the music to this very much. It is the opening theme music to the TV show "House", and they too dont have the voice.

I am on Debian Etch with the 2.6.14 kernel if that matters at all here.

Thanks!

From http://en.wikipedia.org/wiki/House,_M.D.

That edit would be easy enough to duplicate using Audacity. I have tried several software programs that claim to be able to accomplish voice removal (sometimes), and none of them were worth the effort. If your friends stereo does a good job on it, (Hearing is believing ... I haven't.) my suggestion would be to hook up his computer to the stereo, and record it via the line in jack.

I thought about chopping it like this too, but I'm really hoping to get more of the song than that little bit. To be specific, I am making a flash intro thats the whole length of the song (5 minutes and something seconds). I've done a flash intro like this before with a Marolyn Manson song, but that song already never had words. Turned out really good.

Technicality, this was an old friend from about 10 years ago. I no longer have contact with him, so this isnt a valid option for me. :/

What are some of those programs you tried that didnt work for you? I'll give them a shot just for S&G's. I'm bored anyways.

Sound Forge, Cool Edit, and Goldwave ... all in Windows. None of them worked, or they distorted the remaining audio to an unacceptable level.

Check this out ... http://audacity.sourceforge.net/manual-1.2/faq.html#a2

the point is once a song is mixed down to a stereo recording there isn't a way to remove the vocals (or anything else) most utilities that claim to do this take advantage of the fact that in most cases the vocals are mixed dead center and compare the left and right signals to guess whats vocals. I have been able to do the reverse; ripping vocals from a movie without the background music to use as a sample in a song with some clever use of audacity's noise removal feature. but nothing will accomplish what you want with perfect results short of acquiring the original multitrack recordings (pre-mixdown). any attempt to remove anything from a mixed down recording will be a crude hack and will affect everything else in the recording in some way.

It's like trying to make a board longer after you cut it to short you may be able to glue it but the results will be less then impressive.