Is there any way to split different voices in an audio recording? Specifically, I have some taped interviews where there are three people talking. I'd like to be able to split out each voice into a separate track. The original recordings were on cassettes, which I've digitized.
I've been using Amadeus Pro, but it doesn't seem to have this capability.
Now you see why people record with muti-track with multiple mics. You really can't un-mix a recording like this.
What you can do is duplicate the tracks, pan each to a different location then mute each track when the other two are talking. It two are talking you can go 1/2 volume on each track.
If one voice has a very different pitch you might be able to remove it with EQ, or at least make a little less noticeable.
I said "when the others are talking". You can go in at high zoom level and work at high time resolution but don't make it choppy, best to work at the whole word level and ramp gain up and down.
To ramp the gain up and down you need a DAW that allows "automation" and lets you adjust parameters like gain by dragging a line. Then you listen and readjust the line. You can't do it in real time. You have to make incremental adjustmants in the gain, and EQ until the sound is right
This is the same as if all three people were photographed together and you wanted a photo of each. You notice Photoshop has no "Cut photo into three persons" function. So what to do? Duplicate the photo then, carefully using the paint tools paint out the two people you don't want using a mask layer.