Tutorial - Mixing a Narration With Background Music

ToDo

Oct 2/09 - I've fleshed out the outline with a first cut at the text..
- Bill

This tutorial assumes that you have figured out how to connect your microphone to your computer and get the sound from the microphone into Audacity. If you haven't got that far yet, please read the Your First Recording tutorial first.

Introduction

Give yourself and hour or so to complete this tutorial.

On your first attempt:

• don't try to create a masterpiece
• don't work on something critical.

You are learning and experimenting, exploring the many tools Audacity offers you. Don't expect to get it perfect the first time.

It really doesn't matter what your narration is about, and whether you read it or ad lib. It also doesn't matter what the music is, but an instrumental track is preferable.

Save your work often!

Step 1: Record the narration

Consider where your microphone is going to be in relation to everything else in the room. Try to place it so that it picks up a minimum of computer noise (hard drives, fan). When speaking into the microphone, point the microphone at your mouth but don't point your mouth at the microphone. If you can, place the microphone so that you are talking past it - think David Letterman and that microphone he has on his desk: he's always talking over it. Setting up the microphone so that is level with your mouth but a little off to the side also works. This avoids "popped p's" and other breath effects from ruining your recording. Finally, try to set up the microphone so that it is 4 to 6 inches away from your mouth.

While you are talking, try to keep looking in one direction - moving your head left or right, up or down, while you're talking will change the tonal quality of your voice, and may be distracting to your listeners. Speak in a normal, conversational tone of voice, but also speak clearly and enunciate carefully. Your listeners can't see you so don't have the visual cues they would have if your were speaking to them in person.

So let us begin ...

Start Audacity: a new untitled project window opens. Click on File > Save and name your project.

You are now ready to record the narration. Don't worry about mistakes, pauses, coughs etc. - we can edit those out later (fix it in post). In fact you may want to make a few deliberate mistakes so you can practice editing them out later. When you stop to make a correction, go back to the start of a natural break - sentence or phrase - and start again in a normal voice - once you've tried to edit out your mistakes you will realize the importance of this.

Click on the Record button in the Control Toolbar. Record your narration.

Step 2: Edit the narration

Now you can fix all those mistakes and sound like a pro.

Cutting out the mistakes

Don't edit too tightly - the final result should sound natural. Use Audacity's Cut Preview feature to hear how each edit will sound before you make the edit. Then listen back to each edit - undo and try again if it doesn't sound natural.

The image below shows a situation where the speaker paused, cleared his throat and continued with the next sentence. We want to take out the throat-clearing plus enough space around it to make it sound natural.

Editing out a cough.

As you can see, the portion to be removed is selected. Press the C key to hear one second of audio before and after the selection - this lets you hear what it will sound like after the selection is deleted. This is called Cut Preview. If you want to hear more than one second before and after the selection go to Playback Preferences and in the Cut Preview section set the Preview Before Cut Region and Preview After Cut Region times to your liking. Adjust the length of the selection until the edit sounds natural. Various ways to adjust a selection were discussed in the Editing an Existing File tutorial.

Continue in this manner correcting the mistakes until you are satisfied with the end result.

Save your work.

Levelling out the volume

Unless you are professional narrator or voice-over specialist there are probably level (volume) variations during your narration. Remember, your listeners can't see you, so having a consistent volume for your narration is important so they'll be able to hear and understand everything you're saying.

You could go through and manually adjust the volume throughout your narration track using the Envelope Tool, but there's an easier way - use Audacity's built-in Compressor effect. Click on the track panel of your narration track to select the entire track. Click on Effect > Compressor. The Compressor is a complex but very useful effect, so let's take a moment to see how it works.

The Compressor effect works by making the loud parts quieter, then amplifying everything, which ends up making the quiet parts louder. First we'll go to the bottom of the dialog and check Make-up gain for 0dB after compressing and Compress based on Peaks. The former means that Compressor will maximize the volume of the track after it does its work. The latter means that Compressor will look at the peaks of the waveform rather than its average value. The Threshold control is the "tipping point" - the point where Compressor decides if something is "loud" (and should be made quieter) or "quiet" (in which case it leaves it alone). For our first pass we'll set Threshold to -12 dB. The Noise Floor control tells Compressor that anything below that level is noise and it shouldn't make it any louder: for now we'll set that control to -80 dB. The Ratio control tells Compressor how much quieter it should make the loud parts - set it to 6:1. Finally, set the Attack Time to 0.5 seconds and the Decay Time to 1.0 seconds - these two controls tell Compressor how fast it should respond to changes in volume. Click the OK button and let Compressor do its work. Listen back to the result. Are the quiet parts still too quiet? Undo and try again with a Threshold setting of -18 dB. Does you voice sound unnaturally squashed? Undo and try again with a Threshold setting or -6 dB. Note that once we've gone through setting everything up the first time, on subsequent tries we just change the Threshold. This make it easy to home in on the setting that works.

When you have your narration edited to your liking, be sure to save your project. Now it's time to add the music.

Step 3: Import the background music file

Click on File > Import > Audio and open the background music file you've chosen for the project. Audacity imports the music file into your project and puts it in its own stereo track.

Click the Skip to Start button on the Control Toolbar, then click the Play button. Note that Audacity automatically mixes the music and narration for you. Click the Stop button when you've heard enough.

Look at the track panel at the left of each track. Note the Mute and Solo buttons. You can use these to control which tracks you hear when you click the Play button. Clicking the Mute button will turn that track off - it will not be included in the mix when you press the Play button. Clicking the Solo button will cause only that track to be heard when you press the Play button. The exact behaviour of the Solo button can be set in the Tracks Preferences panel.

Step 4: Time-shift the tracks

The following example assumes that you want to cut up the narration and move pieces of it around so as to fit the narration to the pace and dynamics of the music. Your goal may be different. You may want an uninterrupted narration where the background music is merely incidental. This is still a useful exercise as you will learn the power of clips and the various ways to move them to create the effect you want.

The Audacity Tracks and Clips page describes how to use the Time Shift Tool to slide clips around.

Cutting the Narration Track

Start by pressing the Solo button on the narration track - we don't want to listen to the music track while we're doing this work on the narration track.

For the purposes of this tutorial we will assume that we want to cut the narration into three separate segments. Each of these segments will begin when the music makes a dramatic change.

Find the spot between the first and second portions of the narration. Using the Selection Tool click at this point. Click on Edit > Split - a split line appears. There are now two clips on the narration track. Similarly, put a split point between the second and third portions of the narration. The narration track will now look something like this:

Narration track cut into three clips.

Moving the clips to work with the music

We need to turn off Audacity's linked tracks feature to do the work described below. When linked tracks is on, and a label track exists below one or more audio tracks, those tracks are linked and time shifting a clip in one track will time shift all the clips in all the tracks - this is not what we want. Look at the Linked Tracks button in the Edit Toolbar - it should not have a dark grey background. If it does have a dark grey background, click on it to turn it off.

Click the Solo button on the narration track to turn off the Solo function on that track, so we can hear both the narration and music tracks.

Identify the point in the music track where you want the narration to begin. Click at that point with the Selection Tool. Click on Tracks > Add Label at Selection. Type a name for the label, for example "First Narration".

Click on the track panel of the narration track to select the entire track. Click on the Time Shift Tool in the Tools Toolbar. Click and drag in the narration track to move all the clips in the track to the right. When the left edge of the first clip lines up with the label you just made, a yellow "snap line" will light up in yellow. While the snap line is lit, let go. The first clip is now lined up with the selected start point.

Identify the spot in the music track where you want the second narration clip to begin. Create a new label at that point. Using the Selection Tool, drag a selection through the second and third clips in the narration track. Switch to the Time Shift tool then click and drag the second and third narration clips until the start of the second clip lines up with the label.

Repeat this procedure to line up the third narration clip with the appropriate point in the music. Your project will look something like this:

The three narration clips are lined up with the labels.

Step 5: Adjust relative volume levels

You now have the narration clips where you want them. It's now time to adjust the volume of the background music so your listeners can hear what you're saying. There are at least two ways to do this: manually with the Envelope Tool, or automatically with the Auto Duck effect.

Using the Envelope Tool

Details on how to use the Envelope Tool are here. You may want to read that page before continuing.

Select the Envelope Tool from the Tools Toolbar. In the music track, click to create a control point a second or so before the first narration clip begins. Click to create a second control point just as the narration starts. Drag the second control point down to reduce the volume of the music track. Click in the Timeline a few seconds before the start of the first narration clip to hear the effect. Press the spacebar to stop playback. Adjust the first and second control points to get the length and depth of the fade you want. In a similar manner create the fade up at the end of the first narration clip.

An envelope is used the lower the volume of the music track during the narration.

You could continue in this manner doing the fades on the music track for each clip in the narration track. Or you could try using the Auto Duck effect. The advantage of using the Envelope tool is that you can always go back and change the fades. Auto Duck, being an effect, permanently alters the music track.

Using the Auto Duck Effect

Have a look at the Auto Duck page for an overview of how it works.

Note that to use this effect the narration track (which Auto Duck calls the control track) must be below the music track. So we need to drag the music track above the narration track.

Select a portion of the music track that overlaps one of the narration clips, with two or three seconds extra on each side. Click on Effect > Auto Duck. For now, accept the default parameters and click the OK button. After the effect completes, listen to the result. If you don't like it, click on Edit > Undo. Bring up the Auto Duck dialog again and change the parameters. Continue trying different parameters until you get a result you like.

When you have setting for the Auto Duck effect that you like, click on Edit > Undo one more time. Then click on the track panel of the music track to select the entire track. Click on Effect > Auto Duck one last time. Click the OK button to apply the Auto Duck effect to the entire music track.

The result of applying the Auto Duck effect

Note the difference between this and the use of Envelopes: Auto Duck has altered the music track, and the effect is permanent.

Step 6: Smooth fade of the background music

Unless you're very lucky (or have planned very, very well), the music probably goes on for some time after the narration is finished. In this case you'll want to fade out the music. Decide how long you want the music to continue a full volume after the narration has finished, and how long you want it to take to fade out.

Using the Selection Tool, click in the music track at the point where you want the music to be completely faded out. Click on Edit > Select > Cursor to End of Track. Press the delete key. Click at the point where you want the music to begin fading out. Click on Edit > Select > Cursor to End of Track. Click on Effect > Fade Out.

Step 7: Check your mix

If you maximized the volume of your narration track back in Step 2 there is a good chance that when you mix it with the music track the resulting mix will be too loud and cause clipping - this is a bad thing. To check for this, play back a portion of the project where there is narration and background music. Watch the playback level meters - if clipping occurs the red "clip bars" will appear at the right-hand end of the playback meters. If this happens, use the Track Gain Control to turn down the volume of both the narration and music tracks to -2 dB. Listen to the section again and look for the red clip bars on the Playback Meters. Keep turning down the volume of the narration and music tracks until no clipping occurs.

Step 8: Save and Export

The File > Save command just saves the Audacity project. You need to Export your project in order to use it with other audio programs. When you Export your project, Audacity automatically mixes it, so the exported file sounds just like what you hear when you click the Play button.

Make sure there are no extraneous bits far along the time line or your exported file will be unexpectedly long! Click on the Fit Project button in the Edit Toolbar. If the expected length of your final mix is displayed in the window, all is fine. If it looks too long, looks for the extraneous bits of audio and delete them. You may need to click just past the real end of your narration/music and delete everything from there to the end of the track(s).

Audacity can export in a variety of formats. For now the two you will be most interested in are the MP3 format (for podcasts, etc.) and WAV (for burning to CD). Note that in order to export to MP3 format you must first install the optional LAME MP3 encoder: instructions are [here].

Optional - Mix and Render

You may want to Mix and Render your project before you export it. The Mix and Render command (in the Tracks menu) will mix your project, and put the resulting mix in a new track (this is the "render" portion of the command). This gives you the opportunity to maximize the volume of the final mix, and do a final check for clipping. If you would like to try this, follow these steps:

• Click on Edit > Select > All
  • Mix and Render only mixes the selected tracks - you want to mix all of them
• Click on Tracks > Mix and Render
  • your project is mixed into a new stereo track and the previous, separate tracks are deleted
• Click on Edit > Select > All
• Click on View > Show Clipping
  • The Show Clipping command will put red vertical bars anywhere clipping has occurred in your track. If any red bars appear, click on Edit > Undo Mix and Render and go back to Step 7.

If no clipping was detected, proceed:

• Click in the track panel of the newly-created track
• Click on Effect > Normalize
  • accept the default values and click the OK button.
  • the volume of your final mix is maximized.

You exported a project in a previous tutorial. If you need a refresher on how the File > Export command works, that information is here.

Congratulations, your narration over background music is now ready to share with the world.

Bonus: Listening while recording

Listening while recording is problematic if you're using software playthrough - there is a noticeable and annoying delay. Since hardware playthrough seems to be less and less available (not on Mac, not on Vista (?)) I'd vote to eliminate this section and instead dedicate a new tutorial to overdubbing.

I've thought some more about this and come to the conclusion that overdubbing in Audacity is an advanced topic. To do it properly requires an external mixer and an understanding of how to get the output of the mixer into Audacity while at the same time getting playback from Audacity into the mixer and your headphones without sending that signal back into Audacity. This requires a mixer with separate output and monitor buses and enough understanding of signal paths inside a mixer to be able to figure out how to use them. Effective use of overdubbing also requires an understanding of latency and how to deal with it, which would be the subject of another advance tutorial.
- Bill

Preferences: play existing tracks while recording new one (is this on by default? Yes)

What if you wanted to listen to the background audio while narrating? (Give examples, like narrating over a concert recording, etc.)

Talk about wearing headphones, so that your microphone doesn't pick up audio feedback.