StarDestroyer.Net BBS

Are there software that can take a sound file and separate each sound? For example, to take a song that involves a guitar, drums, keyboard and the singer and separate each one?

Any help would be great.

You can use audacity to edit the .wav image of a music file and extract individual snippits if they're easily distinguishable. Otherwise I don't think you're going to be able to do such a thing without the original master recording and a mixing board.

If you can get CoolEdit (or Adobe Audition as it's called nowadays) or something like that, you might be able to play around the frequency spectrum, but don't expect it to sound very good, because guitars and voices have lots of harmonics going on. The best would be to try to get the sound file in some multichannel format, if possible.

What exactly do you want to achieve? There might be workarounds.

salm wrote:What exactly do you want to achieve? There might be workarounds.

I have a sound file that contains at least three voices. My father in law, the neighbour and his wife. I can barely hear the neighbour over his wife's screeching but at least I can hear my father in law well. I was wondering if I could separate the three voices so I can cancel out the wife's screeching so I can clearly hear what the neighbour is saying.

Enigma wrote:

salm wrote:What exactly do you want to achieve? There might be workarounds.

I have a sound file that contains at least three voices. My father in law, the neighbour and his wife. I can barely hear the neighbour over his wife's screeching but at least I can hear my father in law well. I was wondering if I could separate the three voices so I can cancel out the wife's screeching so I can clearly hear what the neighbour is saying.

I would try running some noise filters (you can start with the ones built into Audacity) to reduce the wife's screeching. You should be able to filter out most of it.

But is there a way to separate each voice?

Enigma wrote:But is there a way to separate each voice?

A woman will typically have a higher pitched voice than a man. Therefore if you run the filter to eliminate the higher end and work your way down the spectrum you should functionally remove the woman's voice.