Home Mastering Headstart – Lesson #3

Why you should always use a limiter

When I first trained as a mastering engineer over 15 years ago, it was very rare to use compression in mastering at all. Everything was done using only EQ and limiting. This was mainly because average levels on CDs at the time were as much as 6 or 7 dB lower than they are today, so gentle limiting was all that was needed to achieve those levels.

And I still use limiting today – in fact a so-called “brickwall” mastering limiter is always the final piece of processing in my mastering chain, as we’ve seen. (Which doesn’t mean I’m “brickwalling” my masters, by the way !)

What is limiting ?

A limiter is in fact actually a compressor – a compressor with a very high ratio and very fast attack and release times – and provided it’s used in moderation, the effect is virtually inaudible. It’s used to prevent the audio signal exceeding -1 dBTP (True Peak) and causing clipping distortion, in the most transparent way possible.

(Why -1 dBTP ? See below…)

Less is more

If you push a limiter too hard though, it can be a pretty brutal way to affect the audio – not as harsh as pure clipping distortion, but still very aggresssive and un-musical. For this reason, I recommend you only use a limiter for 3-4 dB of gain reduction at most, and if it’s regularly going above this, listen very carefully for negative side effects.

Excessive limiting has a dulling, “blunting” effect on the audio that I really don’t like. I actually think excessive limiting is one of the major root causes of bad sound in modern mastering. People are using the “EQ and limiting only” rule that applied fifteen years ago, and pushing limiters much too hard in pursuit of the high levels of today’s masters.

It’s much better to do things in stages – perhaps use some buss compression in the mix, some gentle multi-band mastering compression and some subtle limiting. With this technique you can achieve level increases of over 6dB with almost no audible effect on the sound, and sometimes much more without doing damage.

Why use a limiter ?

The benefit of using a limiter is that it allows you flexibility with all the other processing – especially compression. As you’ll see in the next lesson, I like to use quite long attack times when using mastering compression, to keep the sound as natural and punchy as possible. But this means it’s quite likely that the transients of the audio signal will pass through un-touched. The limiter prevents clipping distortion and allows enough flexibility to experiment with radical EQ and level changes without having to worry about distortion (within reason).

Limiter settings

There are many excellent options available for mastering limiting, these days – you probably already have one in your collection. It’s beyond the scope of this course to give you individual settings for every limiter plugin, but luckily most are similar and straightforward to set up.

First:

Set the maximum output level of the limiter to -1 dBTP

This is usually called the “ceiling”, “margin” or “output gain”.

You may be wondering why I’m specifying -1 dB as the ceiling, rather than 0 dBTP, since clipping only occurs over 0 dBFS. Well actually that’s not strictly true, especially when our work is likely to be data-compressed – encoded for mp3, AAC or streaming. For this reason -1 is the maximum recommended peak level for Mastered for iTunes, and the new R128 broadcast loudness standards, as well. To see why, take a look at this video:

Preventing inter-sample peaks – and why it’s important for Mastered for iTunes

Once you’ve watched the video, you’ll also understand why it’s important to enable “inter-sample peak detection”, if your limiter has this feature. If not, 4x over-sampling will also help protect you from inter-sample clipping.

Next:

Set the Input gain to 0 dBFS and the Threshold to -1 dBFS

In most limiters, this ensures there is no overall gain applied to the signal itself – all the gain and level control will handled earlier in the mastering chain. I’ll explain the reasons behind this method in more detail in Lesson 6.

(Note – this assumes your limiter increases the signal gain as you reduce the Threshold, but some limiters work slightly differently. In these cases you may need to set the Input Gain to +1, and the Threshold to 0, instead.)

Try it out

Once you have these settings saved as a preset, try experimenting. Add a gain plugin before the limiter, and push your music up into it.

See how the sound changes – experiment with bypassing the limiter and listen to the difference between clipping and limiting. See how far different songs can be pushed without distorting, and watch the effect on the peak-to-loudness ratio using a Loudness Meter.

Other Useful Parameters

Because of the way I work, I tend not to adjust the other parameters on the limiter often – most have a fairly subtle effect.

One exception might be saturation – sometimes limiting can sound to clean and clinical, and it can be nice to add some “analogue-style” soft distortion to the signal. This can be particularly useful on very clinical recordings that need a little extra “dirt”. Make sure you keep the effect subtle, though – on the whole I prefer to use effects like this in the mix, rather than at the mastering stage.

In the next lesson we’ll look at multiband compression, and how the results sound very different from hard limiting.

P.S

Did you get here via a link from a friend, or Twitter? This lesson is part three of a 7-part free email course to give you a head start with home mastering. I give you quick, useful guidelines from a professional perspective to help you get started today, without damaging all the hard work you’ve already done. Learn more about it and sign up here.