Track selection from single source during recording?

I’m asking this in the audiobook forum because it seems like a related problem space. Please let me know if I should be elsewhere.

I have a friend who directs children in plays. The children need help learning their parts. So it would be great to make a recording of a reading of the play, except without an individual part. Then, the child can listen to that and “insert” their part in the right places, helping them to learn their lines.

However, doing a separate recording “missing out” each part individually would take ages and ages. So what one really needs is a recording where each part is a separate track - and then you can do N mixes, one for each track you want to leave out, and give them to the children. This would be fine if we had access to a professional recording studio with 8 isolatable microphones and a multitrack recorder, but we don’t. We can do a group of people sitting round an omnidirectional mic. This is fine, because quality is not vital here - it’s not for broadcast, only for line learning.

So I think what I’m looking for is either a) another way to solve this problem I haven’t thought of, or b) an addon for or mode in Audacity where you can record from a single source to multiple tracks, switching the target track in real time using e.g. the number keys.

So the process would be to get everyone together around the mic with their scripts, and start reading. In the gaps between each line, the person recording would press the number (1, 2, 3 etc.) of the person about to speak, and Audacity would move the recording to be recording on that track. The result would be a multi-track recording of the play from a single microphone source, which could then be used to make memorisation aids for the children.

Is this possible?


As a rule, Audacity will not do production in real time. So, no.

We can do the split management thing. If you put the individual tracks one above the other, you can easily switch between them with the MUTE and SOLO buttons on the left.

I believe you can File > Export that way and the resulting sound file will follow the buttons. MUTE nothing and the show will sound like a normal performance because Audacity will play all visible tracks at once, and Export them that way, too.

Remember, Audacity will not Save a sound file. You have to Export one, and it won’t Export an MP3 without special software.

I can think of one problem with that idea. It’s perfectly possible to go into a speech at exactly the right time, but not come out. So you will probably never get the “holes” the right size. Hollywood has elaborate studios set up for actors to do this: fit a speech into existing holes.


My stupid-but-simple solution would be to record one track with all performances, then save it as a wav file.

Next, make a new project with 8 tracks and import the complete performance on each track.

After that, replace one performer in each track with silence.

Export each track separately as a wav or an mp3.

cyrano: thanks for that idea. It would be more efficient than doing 8 different recordings, but it would still require going over the file 8 times, muting a different person each time. i was hoping to avoid that.

kozikowski: I’m not sure what you mean by “do production in real time”. Audacity can record a single source to a single track, right? And (with the right hardware) it can record multiple sources to multiple tracks at once. And it can play back multiple tracks at once. It doesn’t seem to me like recording a single source to a chosen one of multiple tracks is very different from a combination of those three abilities.

I understand that once you get the parts into individual tracks, exporting a WAV of your mix is easy. That’s not the problem. It’s finding a time-efficient way to get a multi-track recording if you only have a single source. I guess it wouldn’t have to be at time of recording - but I want a one-pass solution rather than an N-pass solution.

replace one performer in each track with silence.

That’s manually editing, but it does avoid recording the show a million times.

When you select each voice, silence the selection with Control-L or Command-L. Don’t just delete it. Deleting will throw off the timing of the show.