Need to make music quiet & voice loud, both on same track

Audacity 3 Help Forum Windows
1

Audacity Version 2.1.1

I have some music mixed with voice recording as a single audio file (there are no two original audio files, they were mixed automatically when both were recorded). However, what I didn’t realize is that my voice is somehow a tiny whisper (despite practically yelling into the mic) and the music sounds like a planet exploding. Is there any way, for a single track, to somehow extract my voice and/or at least make it louder or the music quieter and/or both?

If not, is there an alternate way to partially remedy this using Audacity?

2

We can’t take apart a mixed performance into individual instruments and voices, fix it, and put it back. So that’s a reshoot.

Is it a stereo show? There is one possibility of rescue and then only if you wired the microphone wrong. Drag-select about 8 seconds of the show featuring both music and voice, File > Exported Selected Audio and post it here. Scroll down from a forum text window > Upload attachment.

There’s no chance of this working, so you should start planning what you’re going to do if there is no show.

Koz

3

Is there any possibility of using Noise Reduction to remove the music/voice and then Normalize the remaining audio in order to artificially split the track? The thought just occurred to me.

4

Noise Reduction only works on constant sounds generally quieter than the desired performance. Microphone hum, air conditioner compressors, computer fan noises. It doesn’t work on conversations or music.

Also, Noise Reduction is not one thing. It’s a sliding scale. It always affects both the noise and the show. This doesn’t matter if the noise is quiet background sound, the show damage isn’t significant, but it just kills you if it’s very loud. That’s why bad microphone hum and leaving a TV on in the next room are still reasons for a reshoot.

Try it. Drag-select a sample of noise without the desired performance. Effect > Noise Reduction > Profile. Then select the whole show by clicking just above MUTE. Effect > Noise Reduction. There will be no combination of those three values that suppresses the noise but leaves the show OK.

My standard settings for voice are 6, 6, 6; 12, 6, 6 and 18, 6, 6.

Koz

5

If I could manage to gather up files for the music that was playing, would it theoretically be possible to cancel the music out? For example, playing a track and its inversion simultaneously should come out as silence (since the highs and lows cancel out). Would this affect my voice at all? If not, it sounds like it’s my best bet.

6

If I could manage to gather up files for the music that was playing, would it theoretically be possible to cancel the music out? For example, playing a track and its inversion simultaneously should come out as silence (since the highs and lows cancel out). Would this affect my voice at all? If not, it sounds like it’s my best bet.

That can work perfectly if the conditions are perfect. Audacity does have an Invert effect and if you mix an inverted copy with the original digital file, you’ll get digital-dead silence. Or, if you happen to have a mix of sounds including a digital file and some other sounds, you can subtract-out the digitl file/sound, leaving everything else untouched. …So, you can try it!

But if the levels, timing, or phase, are not identical it will not work. Usually, something gets changed during the mix… And if you have an unaltered copy of the music/background, you usually already have an unaltered copy of the voice too, so there’s no need to do this.


Here’s are some examples of what won’t work - Record yourself saying “Hello”, and then record yourself saying “Hello” again. If you subtract one from the other, you will NOT hear the SOUND difference… You’ll hear you and your imaginary twin saying “Hello” together… Because of the random timing/phase differences, it turns-out that subtraction sounds exactly like addition (in this scenario).

Or, if you subtract the Karaoke version from the original, you won’t get vocals-only. The music is different and you’ll just get a mix of the two songs.

Or, if you subtract Billy Jean form Thriller, you won’t hear the “difference” between the songs, you’ll just hear two Micheal Jackson songs playing at the same time.

…The sound of the difference is not the same thing as the difference in the sound.

(since the highs and lows cancel out).

This has nothing to do with “highs and lows”. The individual digital samples will be subtracted one-by-one. (i.e. 44,100 samples per second.)


P.S.
If you still have the music, and it’s your voice, then your “best bet” is probably to start over.