Analogues

So you've made a video and you're about to put it online, only the audio just doesn't sound professional. The visual quality of your material is great, the content is engaging, but the audio just isn't up to the same standard. Perhaps you've combined camera-recorded audio with voiceover, so in some parts it's nice and loud, but in others it's too quiet. You just don't have time to fiddle around with getting it right. We're going to fix this in a few minutes using the audio mastering black-magic of dynamic range limiting.

With apologies to Wikipedia.

Dynamic Range

In the field of audio engineering, 'dynamic range' refers to the difference between the loudest and quietest parts of a passage of audio. If you're listening to something with a large dynamic range, you will probably have to turn the volume up quite a bit to hear the quiet parts, and then when the louder parts come, you'll know about it. Thus, a large dynamic range is mainly appropriate when changes in loudness are used to effect, such as in film, and orchestral music. However, it can also be very annoying when you're trying to listen to something in at low volume, in a noisy environment, or where a large dynamic range just isn't really appropriate (ie. your product demo, nostalgic film review, cat videos etc.).

What do you want from me?
I'm not an audio engineer!

I could suggest that you start by adjusting the gain (volume) of each clip making up your video, to keep peak audio levels consistent throughout, but I get it, you've got better things to do. Also, in many kinds of audio material this still won't solve the problem of having a low average volume over time. So instead what we are going to do, is effectively 'squash' down the loud parts, to be more comparable in volume to the quieter parts, then boost the whole lot. In doing so, we will have noticeably increased the average volume, and thus the listenability (sorry, dictionary) of your video. In television & radio this process is part of 'broadcast mastering', which is intended to make audio content sound good in a wide range of listening environments.

First you're going to need to get hold of George Yohng's excellent W1, a freeware brickwall limiter audio effect/plugin. If you've got a Mac, grab the Audio Unit version. If you're using Windows you want the VST (.dll), but you should get the version released by BetabugsAudio instead, as the original on George's site is a 'headless' plugin (no UI included), which can cause problems. If you need help installing either plugin, see the end of this post.

You also need an audio app to host the plugin. I'd say the ideal choices on Mac are Soundtrack Pro or GarageBand (seriously). On Windows you can really use anything that can host a VST plugin, however my suggestions are Adobe Premiere, Soundbooth or Audition. Also, apparently Sony Vegas and Sony ACID both support VST plugins, but I'm not really familiar with them. Audacity... yes, technically you can use VST plugins with Audacity but it's a pain to use for this particular task, due to its lack of real-time effects. Use something else if available.

Hitting the Brickwall

In your application of choice, apply the W1 Limiter effect. Ideally you want to apply it as a real-time/rack effect, rather than as a 'processed' effect, so that you can twiddle the knobs while listening to the audio. Again, head to the bottom of the page for specific instructions if you don't know how to do this.

Navigate to a part of the audio where you've got some voice or sound that is a bit too quiet, and play it, looped. I find it helps to play a section from a part that is too quiet into a part that is too loud. Back in the limiter plugin, bring the 'threshold' parameter down to about -10dB. If you're using the Betabugs UI, this would be about '4' on the 'thresh' knob. You should notice an increase in loudness. What you're shooting for here is the master output meter consistently peaking near or at 0dB (the top). You don't need to worry about going into the red, as that is precisely what a brickwall limiter is intended to prevent. If somehow you are seeing the red 'over' indicator, drop the 'ceiling' parameter of the limiter plugin just a little bit.

The great thing about the W1 Limiter is that it is quite transparent, so you can crank it up quite a bit without negative side-effects. I know the knob on the Betabugs version goes up past 11. Still, try to use some restraint and not limit the audio signal more than is necessary, to avoid aural fatigue for your listeners. If you overdo it, you will hear a constant, fast 'pumping' of fluctuating volume. Ease off a bit. Conversely, if you've got the threshold at -30dB and it's still not loud enough, you can turn up the volume of the video's original audio track (note: not the master track).

Also, fun fact, if you are combining background audio with voiceover, increasing the volume of the voice clips and turning up the threshold (as in, more negative dB) in the limiter will have a similar effect as 'ducking' (where the background audio is lowered during speech), which is great if you don't want to have to set up the ducking effect properly.

Final Steps

So we've fixed your audio, now to get it back into your video. Basically I would suggest you export to AIFF/Wave and just replace the audio in your original video. Obviously if you are applying the effect directly in Premiere, this is unnecessary. However, if you used GarageBand and happened to import your video from an MP4 container, you can actually choose to 'pass through' the video stream, into a new .mp4 container file, with the improved audio. That is to say, without transcoding it, which is pretty cool. 

Detailed Instructions

GarageBand

Get the W1 plugin from George Yohng's website. You want the 'AudioUnit (uni)' one. Unzip to:

/Library/Audio/Plug-ins/Components/

As I mentioned above, you can import the whole video file if it's an MP4 file, just make sure when you exported it from your video app that you kept the audio at max quality. You can recompress it in the export step anyway. Otherwise, an AIFF/Wave mixdown of the video's audio track will suffice (or if you used iMovie, the project file).

Once you've got the audio imported, head to the Master Track info, in the Track Info panel. In the empty effects slot, select W1 from the drop-down list. You can click on the Audio Unit icon to the left, to access the parameters window. From there, you can follow the guide above.

Soundtrack Pro

Get the W1 plugin from George Yohng's website. You want the 'AudioUnit (uni)' one. Unzip to:

/Library/Audio/Plug-ins/Components/

Import your audio, or your Final Cut project. In the mixer, scroll across and select the 'Master' track. In the effects tab on the left, add the W1 plugin, then click the 'Advanced...' button in the parameter column. This will bring up the parameters window. From there, you can follow the guide above.

Adobe Premiere Pro (Windows only*)

You probably want the BetabugsAudio version of the plugin, though we're not making use of the custom UI, so you could just use the one from George Yohng's site. You can put the .dll in Premiere's VST plugin folder, default:

C:\Program Files\Adobe\Adobe Premiere Pro CS3\Plug-ins\en_US\VSTPlugins

However, I suggest you put them in the default VST folder, which is:

C:\Program Files\Steinberg\VstPlugins

Just create it if you don't have it already.

Load up your project, open up the audio mixer, and look to the left, in the margin of the panel is a little triangular expand/collapse button that hides the effects rack. Back over in the Master (right-most) channel, add a 'W1 limiter' to the effects rack. Don't bother opening its custom user interface, just select the effect and use the Threshold knob that appears below the rack. From here you can follow the guide above.

Adobe Soundbooth (Windows only*)

Grab the BetabugsAudio version of the plugin, you'll need its UI in Soundbooth. Put the .dll in Soundbooth's VST plugin folder, default:

C:\Program Files\Adobe\Adobe Soundbooth CS3\Plug-ins\Common\VSTPlugins

Export an audio mixdown from your video app, and load it up in Soundbooth. From the effects menu, choose George Yohng>W1 Limiter, then in the effects panel on the left, click 'Settings...' and follow the guide above.

*Unfortunately Premiere and Soundbooth's VST support on Mac OS is a big load of bullshit and chips, and they have no AU support, so they're only good for this task on Windows.

Bonus tips, for reading this far

Use an EQ to take out frequencies below 100Hz. If you have a (potentially inaudiable) low frequency hum on the recording, this will prevent it from wasting valuable headroom. You might not even know it's there. Just make sure you put the EQ effect before the limiter in the effects chain.