Limiting:
The Second Last Stage

Limiting takes your song levels to the max!

It involves the same parameters as compression, but at more extreme settings.

This is particularly true where the ratio is concerned - anything above 10:1.

The attack and release times are also much shorter because we're working on the more 'transparent' part of the audio signal. The thresholds are also more higher.

The purpose here is to put a cap on the maximum level a song file can reach, so that you can then raise your levels, closing the headroom gap.

Yeah – this is where you make your song LOUD!!

I should point out that I intend to speak more specifically about what's referred to as brickwall, because I'm dealing in the digital domain with software.

In the analog world, these devices do a similar job – give a sense of fullness to the sound - but they don't catch all the overshoots.

It's important, if you intend to raise your levels, that you never exceed 0 dB.

Technically speaking, that can't happen. But what you'll do is create clipping which introduces distortion.

If you've done a proper job of equalisation and compression beforehand, you won't need to drive your levels as much.

And as I've said elsewhere, the less processing involved, the less potential for signal degradation.

There's seems to be two types of approaches when working with this process:

One method uses equalizers, compressors, and any other enhancement devices to achieve gain increase, while the mastering limiter is there only to prevent overs and maybe add a small amount of gain increase.

The other involves using the same tools but staying at close to the song's original volume level to do enhancement, but using the limiter to give whatever gain increase is needed or wanted, while still preventing overshoots.

I tend to favour the second method. Using mastering software, it just makes more sense to stay at saner levels until the second last step in the chain.

With compression, the approach to gain reduction tends to be more audible, working on the RMS levels, and attempts to enhance the rhythms in a pleasing way.

With a stereo limiter, we're working with transients. While transients contribute much to the 'liveness' of a track, they are, for the most part, more transparent to the human ear.

You can often get away with as much as 3 dB of gain reduction before it can start to become apparent.

It's a good idea, therefore, to also keep your eye on the gain reduction meter. You want to see it 'bounce'. You want to see it return to zero frequently. It's even OK to have occasional reductions as much as 6 dB, as long as your also at zero as much, or more.

If you see little movement and are always in gain reduction mode, you're doing too much and need to back off.

'Loud' doesn't necessarily make a song better.

Too much 'loud' will quickly make your listener tune out and turn off.

You'll limit your audience!

Then where will you be?

PS: You notice that I've referred to this as the second last step. You'll find the last step here.

Return from Limiting to
Signal Processing

Return to About Audio Mastering Software