Thoroughly confused-please help!


I’ve been searching for a while and haven’t seen my particular problem addressed. Forgive me if I missed it.

My issue is that I have several tracks of music and talking. The talking is way softer than the music, and to listen you have to keep turning up the volume knob to hear the talk and then you get blasted when the music starts. I want to bring up the volume of the talk (preferably w/o adding noise, but acceptable if it has to be) up to or at least close to the volume of the music. I’ve seen Amplify, Envelope tool, ACG, compressor, leveler and amplify all suggested for these types of things, but don’t know which would be best. HELP!!!

My second issue is that the left is lower than the right, but I should be able to fix that with normalize. The only question is should I normalize before or after the fix for the issue above? Does it matter?

Any assistance given will be greatly appreciated.


My issue is that I have several tracks of music and talking.

That suggests you have tracks with music and tracks with voice and they’re out of balance. That’s a snap to fix with several Audacity tools.

I suspect what you really have is at least one track with both voice and music mixed in it. There are no automatic tools to fix that. The desperation method is to magnify the track so you can see all the places where speech happens, select each speech segment and amplify it being careful not to touch the music.

If all your lucky moons and stars line up, you might be able to use Vocal Isolation to get the voice track by itself, amplify it and slide it back into the show. That’s using tools that by themselves have about a 30% success rate, so don’t bet the farm.

I don’t think Vocal Isolation is a natural Audacity tool, so we need to wait for another elf that remembers where it is.


You didn’t say if the show was high quality stereo. If it’s not, none of these tools except magnify and amplify work. Koz

Sorry, I posted earlier and it didn’t show up. It is not high quality audio, it is audio taken from reel to reel of our old band rehearsals. We would play a song, then talk or goof around a bit, and then play another song, and so on. It sounds like I will have to amplify the talking bits one by one, something I was hoping to avoid.

I will be adding labels so that I can export to CD as individual tracks. Is there a way to amplify the labeled tracks (i.e. the info between labels) easily, or will I have to highlight each part I want to amplify? I assume that I should run the normalize to bring up the left channel after this is done, correct?

Thanks for your help.