Mastering: How much should I limit?

Mastering is a process of preparing a mixdown for distribution. Often people focus on signal processing (such as EQ and compression), but it’s important to remember that it also includes setting times between songs (for mastering albums or EPs), fades (usually fadeouts), and (in this day and age) encoding to a lossy format such as MP3 or AAC. 

First, do no harm. 

You’re asking about dynamics, so I’ll focus on that here. 

In preparing a mixdown for distribution, you should consider dynamic range and overall level. They’re not the same thing: Dynamic range is the difference between the loudest parts and the softest parts. Overall level is (in digital) the distance below 0dBfs. 

If you’re targeting digital distribution, then you’re probably expecting your music to be added to a listener’s library (such as iTunes) and played alongside other music. If you want your music to be perceived as “normal” or “professional”, then you’ll need your music to sound as similar as possible to the other music your listener has in his/her library. 

To best achieve this, you should select two or three other songs that you think represent a similar genre to your music. The more similar (in genre, instrumentation, and overall sound), the better. The more well-known, the better. This is your reference. 

You should find a way to be able to switch quickly between the references and your mixdown. Personally, I do this by starting a new project in my DAW and loading my mixdown and my references each on their own track. By working with SOLO mode on, I can switch between tracks as fast as I can press the up or down keys to switch to the current track. 

Resist the temptation to immediately slap a limiter or mastering toolbox on your mixdown and turn it up to match the references. Instead, turn the references down to match your mixdown. Don’t use meters – use your ears. Make a note of how much you turn them down. 

Let’s pretend for now that you have two reference tracks, and you turned them down by the same amount: 18dB. 

Now you have your mixdown playing at the same overall loudness as the reference tracks. First listen to the dynamic range of your mixdown compared to the references. Are the quiet sections too quiet? Are there any loud bursts (not transients – but whole notes or sections) that are too loud? If so, you might find a compressor useful. Use it to subtly reduce the dynamic range of your mixdown. Don’t use it for “colour”, or to adjust transients. Typically I’ll start with a fast attack, medium-long release, low ratio and medium-deep threshold. If your compressor has an RMS (or similar) sensing mode, use it. It might take quite a bit of fiddling to get this right. Remember to bypass te compressor regularly to make sure you’re doing no harm. Also remember to keep the overall level of the mixdown the same. Resist the temptation to push it all louder at this stage. Typically you won’t have to do much to adjust the dynamic range of your mixdown. 

Once you’ve got the dynamic range right, it’s time to look at overall level – the distance below 0dBfs. Right now your mixdown is the same loudness as your references. However, the peaks on your references are at -18dB (that’s how much you turned them down), and your mixdown might be peaking around -9dB or higher. This is where the limiter comes in. 

What you need to do is bring the peak level of your mixdown to -18dBfs without changing the loudness as you hear it

Adjusting the peak/average level ratio this way will make it easier for you to hear when you are doing damage to your audio. The typical method of “pushing up” the level of the mix makes it difficult because you’re changing two things at once – the peak/average ratio and the overall level as you hear it. Not only does this complicate the hearing process, but it also makes it easy to ignore audio damage because the more damage you do, the louder it gets (and as you know, humans tend to perceive louder music as “better”). 

Unfortunately, most limiter plugins are configured for the above behaviour, sporting an easy-to-abuse “input level” control. To get around this, you should insert two gain plugins, so your chain looks a little like this: 


Set the limiter to limit at 0dBfs. When you switch it on, you should hear no effect. Then slowly increase the level of the first gain and simultaneously reduce the level of the second gain by the same amount[1]. Do this 1dB at a time, and listen carefully to what you’re doing to your sound. 

If you get down to -18dB (or whatever your references are at) without reducing the overall level of your mixdown as you hear it and without doing too much audible damage to your sound, then WELL DONE! YOU WIN THE GAME! Turn your speakers down, remove the second gain plugin, and render it! 

If you couldn’t get down to the level of your references, you need to adjust the settings on your limiter, or use different tools. Unfortunately I can’t give much advice here, as it depends entirely on the kind of damage the limiter is doing to your sound, and what kind of damage you’re willing to tolerate. Sometimes you might need to reduce the bass with EQ (why it’s important to match the spectral/EQ response to your references before you try to match the dynamics). Sometimes you need to add some more aggressive clipping to retain the overall level as your reduce the peaks (particularly if your reference are as loud as commercial pop). Sometimes you might need to use a compressor beforehand to bring a “spiking” element into line (such as the kick drum in my track Horse Head). I’ve heard of some people using a series of several compressors or limiters, each reducing a small amount (personally I haven’t needed to try it, it sounds far too complicated to control effectively). 

If you’ve done well, you should have your mixdown at the same overall level as your references (as it has always been), but also with the peaks at the same level as well (-18dB in this example). When you’re happy turn your speakers down, remove the second gain plugin (which will pop your mixdown up to 0dBfs), and render it! 


[1]Unnecessary technical note: What this effectively does is move the 0dB point inside the limiter plugin! This is a cool trick to do with any other dynamically-sensitive plugin, such as compressors or distortion! Blue Cat’s free gain plugins can be linked in reverse, allowing this kind of two-way gain adjustment to be done quite easily.

  1. Hey thanx heaps! really helpful advice..

  2. Hello Kim,

    Thanx for sharing all this amazing knowledge.
    I’m still very confused about compression. If the quiet parts are to soft in a mix you can also just turn them up in volume, right? This way you will also reduce the dynamic range?Or do you use compression because you are already the mastering stage?

    Greetings, Chris.

  3. @chris H
    Dynamic range is the difference between the quiet sections and the loud sections. You can use compression to control the dynamic range, but simply riding the gain (typically by using the volume fader) is another way to control dynamics.


  1. No trackbacks yet.

Comments are closed.
%d bloggers like this: