6 Steps To Releasing Your Music 


with Complete Confidence

by Ian Shepherd

Thanks for downloading this guide! As a professional mastering engineer myself, I know
the goal of every musician in the home studio is to make your music sound the best it can
possibly be – on any platform, online and everywhere else.

You probably have some strategies for home mastering already, that can get it sounding
great on your own system. But maybe they don’t work so well when you listen to your
music out in the world – in your car, on earbuds, other people’s monitors or online. Or
maybe you’re just baffled by all the standards and requirements for all the different formats
and online streaming platforms. I’m going to help you with that.

In this brief guide I’ll going to walk you through the 6 steps that every master must go
through in order to sound its best.

We all want our music to sound fantastic, so that people connect directly with the emotion
and the passion behind it.

My promise to you is that after reading this guide, you’ll know exactly what it takes to
release your music with complete confidence, and what you can to get the best possible
results, listening online – or anywhere else.

Ready to get started? Let’s dive in

Ian Shepherd
(Mastering Engineer, www.ProductionAdvice.co.uk)
Overview
There are six crucial steps to mastering.

Everyone thinks about the fun stuff when they think of mastering. Expensive EQ, compression,
limiting and more, but the truth is that processing is just one of the six steps. And without the other
five, you’ll never get the best possible results.

The good news is that each step is simple to understand, with clearly defined goals. Mastering isn’t
easy, but it is simple.


The 6 steps are:

1. Calibrate – Optimise your monitoring and listening environment

2. Compare – Identify your goals using reference material at matched loudness
3. Adjust – Level, EQ, compression, limiting and more
4. Repair – Spot and correct common technical faults
5. Preview – Hear your music through your listener’s ears – the ultimate real-world test !
6. Deliver – Export your files in the correct format for each platform

Understanding what these steps are and how they fit together will help you get better home
mastering results.


STEP 1: Calibrate
To make the best decisions for how your music will sound, you need to hear it clearly. That
means your monitoring and listening environment have to be up to scratch, and this is one
of the hardest challenges to deal with for home mastering.

Professional mastering studios have been finely tuned over many years to be as neutral
and accurate as possible. Often they’re purpose-built, with insanely expensive speakers
and amps. Most of us don’t have the luxury of doing this at home ! 



The good news though is that you don’t have to. Mastering monitoring doesn’t need to be
insanely great to get good results (although it helps!) but it does need to be good enough.

That means it needs to be capable of reproducing a wide enough frequency range, and
many affordable modern monitors are up to the job.

In a minute I’m going to suggest three guidelines, or “Monitoring Maxims”, that’ll help you
get better results with any monitoring setup. First we need to tackle an important topic,
though - and understand why it matters.
The Loudness Deception


Our ears are more sensitive to bass and treble for louder sounds. No-one knows quite why
– it might be the result of evolution, perhaps. It’s more valuable to pay attention to the
sabre-tooth tiger breathing down our neck than the one over there watching that herd of
gazelles, for example, so our brain pays more attention to louder sounds.



Whatever the reason, the practical effect is that most people think that louder sounds
better. Bigger, wider, richer, deeper.

I call this the “loudness deception”. If you do nothing else but turn something up, it sounds
better – this is the root cause of the “loudness war”. To hear the loudness deception in
action, watch this video.

The effect for mastering is critical, because in order to get the music to sound right, you
need to get the loudness right first. If you add EQ and compression to something but then
change the loudness afterwards, you’ll change the way it sounds and have to adjust the
EQ again.



All of which leads to the first monitoring maxim:

Monitoring Maxim #1 - Choose ONE fixed mastering monitoring level

Since loudness affects how we hear things, and we want to listen accurately, our
monitoring level needs to be right. If your monitors are too quiet, you’ll always end up
making things too loud. If you monitors are too loud, you won’t make things loud enough -
or you’ll damage your ears.

The goal is to find an ideal monitoring level where things sound good and loud when they
need to be, but not too loud. Finding this will take some trial and error, but once you find it,
mark the level on your amp or monitoring controller, and don’t change it.

This one simple step makes a huge difference to success in mastering. Over time you’ll
learn exactly how things should sound, and how loud they are. You’ll hear EQ and
compression differences more clearly, and make better decisions for the music you’re
mastering. It’s amazing !

(Of course, the perfect monitoring level depends on how loud you make your masters.
How loud should they be ? We’ll dig into this in Step 3)

Monitoring Maxim #2 - Make it as accurate and neutral as possible


If your monitoring has too much bass, your masters won’t have enough. If it has too little,
they’ll have too much. If they’re too toppy, your masters will be too dull everywhere else…
and so on.



Put another way, you want you monitors to be as flat and neutral as possible, when
mastering. “Average”, if you like. That way, when someone plays your masters on earbuds,
they’ll sound tinny and harsh – just as they expect. In a car with massive bass bins, the
low end will sound massive - just as they expect. The less your monitoring changes the
sound of what you’re listening to, the better your masters can be.

Whatever you do, don’t fall for the idea that because people are listening on earbuds and
phones, we should too. By all means double-check on phone speakers and earbuds, but
don’t use them for final decisions. To find out why not, read this. Mastering monitoring
should be as accurate, and as neutral, as possible.

Monitoring Maxim #3 - Use acoustic treatment



This is probably the single most important step there is when you’re getting your
monitoring to sound as good as possible, and also the one most people don’t bother with.



Don’t be that person. Your room is lying to you, colouring the sound - all rooms do, even
professional designed studios. Adding decent acoustic treatment to your room is the best
way to solve this problem – the best monitoring upgrade you can buy, and easily the most
cost-effective.



DON’T use acoustic foam - and don’t just put it where you think it needs to go. There are
simple straightforward guidelines to get great results, which work in almost any room.



All rooms need acoustic treatment, and almost any treatment will help the room sound
better, and make it easier to get good results. Yes, you can cope without - but it’s really
hard.



To see how I added acoustic treatment to my own “home mastering studio” (and how I
almost got it wrong) take a look at this video:



Building a home mastering studio - Part 3: Acoustic treatment



The treatment I describe in this video literally changed my home room from unusable to
sounding great.
Monitoring Maxim #4 - Learn how your monitoring sounds


This one is simple, easy and free. You don’t even have to try !



Just listen to a ton of music in your studio. That’s it !



It’s such a simple idea, but so few people do it. We only work on our own music in the
studio, so we have no idea what the other music that we like sounds like.


So just listen, all the time, whenever you’re not actually working on music. When you’re
backing up files, or mending cables, or reading PDFs, whatever – over time, you’ll learn
how the music you like should sound, and unconsciously start to make yours sound more
like it, too.


And this idea leads us naturally to…




STEP 2: Compare
One of the most important goals of mastering is to help the music translate - to make it
sound great everywhere, not just on our own monitoring.



But how do we know that ? Our monitoring is all we have !

The answer is to find songs that already translate well - that sound great to you wherever
you listen to them, and work to make your music sound more like them.

Since these reference tracks sound great to you everywhere else already, if you can make
your music sound great compared to them, it should also sound great out in the real world,
too. There are some rules you need to follow to get this to work really well, though.


Comparison Concept #1 – You Need To Be Brave


I’m going to be honest with you - comparing your mixes and masters to reference tracks is
going to hurt. 



Because the chances are, your favourite songs were performed, recorded, mixed and
mastered by legends, at the top of their game. That’s why you love them so much in the
first place.



You’re comparing yourself to the best of the best, which is the toughest comparison you
can make.

So be brave - it’ll be worth it, because you’ll learn a ton. But it’s going to hurt. The next two
Comparison Concepts will help, though !
Comparison Concept #2 – You Need To Choose Wisely

True story: someone once sent me a metal album to master. Hardcore aggressive metal -
but their references included Steely Dan, Simon & Garfunkel and Fleetwood Mac.

!!!



Don’t get me wrong, those are some great-sounding references, and they told me a lot
about the breadth of taste of the artists - but they told me nothing about how the album
should sound, because they sounded nothing like it. 



In terms of their goals for the sound, it would be much better if they had chosen reference
material in the same genre, at least.



You should do the same. Look for reference material that fits well with you music - so that
you might be included together in a playlist, for example. That way the comparisons you
make will be more relevant.

Comparison Concept #3 – ALWAYS Loudness Match when comparing

Now you know about the Loudness Deception, it should be obvious why this is important.

If one of the reference files is even a dB louder than yours, there’s a chance it’ll sound
better to you just because it’s louder.


That’s not a fair comparison, and it’s not realistic, either.




In 2017, over 85% of all music industry revenue in the US came from non-physical
formats.

That’s code for streaming and downloads, and almost all streaming services these days
use loudness normalisation. They turn down the loudest songs, to stop you being
blasted by sudden changes in loudness, because it’s the biggest source of user
complaints.



This applies by default to all the main streaming services – YouTube, Spotify, TIDAL,
Amazon Music, Pandora and Deezer. And only 15% of users bother to turn it off. On
YouTube and Pandora you can’t turn it off ! iTunes has a similar feature but is the only
platform where it’s off to begin with.



What that means is that online, most people won’t hear the file at it’s original
loudness.

And even on physical formats like CD, the first thing people do is adjust the volume to their
own preference.


So when we’re making comparisons to try and make our music sound better, there’s no
point in trying to match loudness. In fact we learn far more by taking loudness out of the
equation.



A quick note - I’m not telling you not to master your music loud, if you want to ! That’s a
personal and creative choice, and we’ll talk about it more in Step 3. Just don’t get hung up
on loudness. You don’t need to compete, any more. Master your music so it sounds the
best it can be, so it sounds great when the loudness is matched, and decide exactly how
loud that should actually be as a separate issue.




How to Loudness Match
Loudness matching be ear is a tricky skill that can take years to learn, but luckily there’s a
short-cut.


We just measure, using Loudness Units (Full Scale) or LUFS for short.


The LUFS method isn’t perfect - sometimes your ears will tell you something different than
the numbers, but it’s really pretty good on most material.



And importantly, it’s an international standard, which is already used by TV & radio
stations, plus YouTube, TIDAL and Amazon Music.

So the loudness of your music in LUFS determines how many people will hear it for the
first time, making it an important step in the comparison process.



There are a ton of LUFS meters available, ranging from expensive pro hardware options to
affordable and free plugins – in fact there may already be an LUFS meter in your DAW.

They all give the same results, so just pick one that suits you and your budget.



The process is then simple:

1. Play your song and reference tracks all the way through, and look for the
“integrated” or “program” loudness value. This is an overall value that can be
measured for any piece of audio – if you watch it in realtime you’ll see it changing
quite a bit to begin with, and finally settling on a single value. Play each song from
start to end and make a note of the value for each one.


2. Adjust your song and all the reference tracks to all have the same integrated LUFS
loudness - use the lowest value for now to avoid clipping distortion.

That’s it - you’re now listening to the songs at the same overall loudness in LUFS. Any
differences you hear now when you compare will be the way people hear them on many
streaming platforms.




How to Compare Masters
Comparing is easy. Listen, and decide what you do (or don’t) like about each song. Try
and figure out why you do or don’t like it, and make some notes.



The trick is to make sure you compare as many different aspects of the music as possible,
to make sure you’re listening like a mastering engineer.



Here are the four main things I listen to, and questions I ask myself about when I’m
comparing songs in mastering:



EQ Balance


What is the overall EQ of each song ? Is one brighter or darker than the other, does one
have tight, punchy bass and the other has deep, booming subs ? Does one have a harsh
or cutting mid-range, does another have open, spacious top end ?



We’re going to look more deeply into EQ and how to figure out what you’re hearing in Step
3, so for the time being just listen and pay as much attention as possible to the differences
you hear.

Dynamics

How compressed is each song ? Are some varied, open and lively, while others are dense,
controlled and intense ? Do some have big contrasts between the louder and quieter
sections, but others all sound a similar level throughout ?



Does the sound you hear match the style and emotional intensity of the song ? Are the
verses louder than the chorus, does the drop really hit you ? Do the songs maintain energy
throughout, or do some sections lose presence and impact ? Listen especially to the
drums - how much snap and clarity do they have, and which do you prefer ?

Distortion

There are many different types of distortion. Some are beautiful creative decisions, others
are a side-effect of extreme dynamics processing. Some sound soft and mulchy, some
sound harsh and aggressive.

How does the degree of distortion relate to the dynamics and intensity of the music in your
song and the reference tracks ? Think about how distorted each song is, and whether it
suits the style and the emotional impact of the music.
Stereo image

There are two aspects to stereo image for me - width, and depth.



Width simply describes the spread of the elements in the mix or master - how widely are
they panned, how much contrast is there, does the image sound very tightly focused in the
centre, or much more spacious ?


Depth is more about the “3D” nature of the audio. Does everything sound like it’s coming
from a simple left-right line between the speakers ? Or is there a sense of space in the
sound ? Do some elements sound as if they’re set further back behind the speakers, and
others are more up-front ? Sometimes sounds can even sound almost as if they’ve come
from outside the speakers, or behind you.



Again, think about whether the choices suit the music and the emotion. Is the wide stereo
image cool, or distracting ? Does the space add to your enjoyment ? Or does the sound
seem clustered and muddled ? Reverb and delays can have an important impact on the
way we hear stereo.

Time For Action

Before we move on to Step 3, I’d like to ask you to apply some of the ideas from these first
two steps to your own music. So before you go any further, take these steps:




1. Choose 3 reference tracks




These need to be in a similar genre to the song you want to master, and they need
to sound fantastic to you, wherever you listen to them.


2. Loudness-match them by measuring their integrated LUFS values




Any LUFS meter will do - just measure the three reference songs and your own,
and turn them up or down to match.



So for example, if your mix measures -18 LUFS, and the first reference song
measures -12 LUFS, you need to turn it down by 6 dB before you compare


3. Adjust your monitoring level so they all sound good and loud



This isn’t your fixed mastering level yet, we’ll be deciding that in Step 3 - we just
want to listen at a good healthy level so we can hear what’s going on clearly


4. Compare the songs, and make notes




Think about all the categories I listed above, and decide which songs sound best to
you, and why. If you want to try and improve you own mix, think about what you
might need to change in the mastering.



And remember - be brave ! Your mix isn’t mastered yet, so there’s plenty of scope
 to improve it. If it doesn’t sound great yet, that’s OK. We’re going to improve it - this
stage is all about learning to listen objectively, before we start adding mastering
processing.




That’s it – you’re done. Great work ! 


Hopefully you now have some helpful notes about the way your songs sound in
comparison to the reference tracks, and how you might like to try and improve them.

For example, perhaps the reference songs have more solid, defined bass. Or a more
open, airy top end. Or maybe you don’t like them as much as you thought you did, now
you’ve heard them loudness-matched !

Loudness can hide all kinds of problems, overwhelming subtle differences with it’s sheer
brute force. But most people nowadays are hearing things first online, with matched
loudness, and this can reveal distortion and other aspects of the sound that we didn’t
notice before.

I often hear people claim that the streaming services have ruined the sound of their
masters, and it is true that the quality can suffer, especial on mobile devices at low data-
rates. But many times when I listen, I find that what’s actually happening is that the
loudness-matching of the streaming services is revealing things people don’t like about
the sound, that were there already but they hadn’t noticed.



Most likely, it’s a mix of all these things - there are probably things you prefer about the
reference songs, and things you like about your mix, too. Either way, in Steps 3 & 4 we’re
going to look at the ways you can adjust the sound in mastering, and repair common
problems.

Before we get to that though, we need to deal with the (loud) elephant in the room.

How Loud Should You Master ?

It’s the most common question I get asked.

Loudness is fundamental, and affects every other decision we make in mastering. So how
loud should you go ? How loud is loud enough, and how loud is too loud ? What’s the best
loudness online ? And so on.

In a nutshell, you have two options:

1. Match the loudness of a reference track


2. Choose a specific loudness to master at

Option 1 is easy to understand - just choose a song or artist you think sounds great and
aim for a similar loudness. This works, but it can be hard.
A lot of mainstream music is mastered super-loud at the moment, and this can be difficult
to achieve with good results. It’s quite possible, but the question is, is the effort
worthwhile ? If people hear it online, they won’t know how loud it was originally, and
research shows that music fans actually don’t care about overall loudness, even when
normalisation isn’t being used.

Option 2 on the other hand allows you to choose exactly how loud you want to master,
without competing with every other release out there – but it still doesn’t answer the
question of “how loud” that should be.

My recommendation for loudness

Make the loudest sections no louder than -10 LUFS, and balance everything else
musically



If that idea seems simple – that’s because it is ! But it’s based on over 25 years of
professional mastering experience, and it still serves me well. This strategy will get you
right in the “loudness sweet spot” – it sounds great, with excellent dynamics, and it will
work well on both CD, vinyl and online.



We’ll talk more about this again in Step 6, but if you’re interested in more detail about this
strategy, and why it works, take a look at this:



How loud ? The simple solution to optimizing loudness online – and everywhere
else



Before we move on, a few comments:


• This guideline is based on using short-term LUFS, not the overall “integrated” value.

You can find out more about this at the link above.
• I also recommend peak levels no higher than -1 dB (True Peak)

Again, there’s more detail at the link above, and we’ll look at True Peak levels again in
Step 5
• This is not super-loud, by current mainstream standards

But it will enable you to achieve a great sound, in almost any genre, and work really
well online
• You may even find it’s too loud for your tastes - in which case choose a lower value !

There are no rules on this, no standards - it’s all about finding your own “loudness
sweet spot”

Which leads us to another important point:

There are no “targets” - we DON’T want to make everything equally loud


It doesn’t make musical sense for everything to be the same loudness – an acoustic ballad
shouldn’t be as loud as a heavy rock song, for example. So even when we’ve decided how
loud we want to master, there should still be plenty of variety from song to song. They’ll
work musically next to each other, which is a crucial goal of mastering.

It also doesn’t make sense to aim for the reference levels used by online streaming
services, either. I often see people saying “aim for -14 LUFS”, since that’s a common
reference level used by YouTube, TIDAL and others. But we don’t need to aim for that
value, or any other one.

The streaming services adjust to their preferred level for us, so all we need to do is make
the music sounds as good as it can. It’s a good idea to Preview what they’ll do, so we can
make sure we’re happy with the end result - but if it gets turned down by 2-3 dB, provided
it sounds great, that’s fine. We’ll look at this in more detail in Step 6.

And don’t worry – the variety I’m suggesting still works with normalisation, too. Individual
songs may get turned down in playlists or shuffle mode, but Spotify, TIDAL and iTunes all
have an “album mode” to make sure quieter songs aren’t boosted when we listen to
albums.

Setting your mastering monitoring level

Once you’ve decided how loud to master your music, you can choose your final mastering
monitoring level, as I suggested back in Step 1.

It doesn’t matter whether you’re matching a particular song, album or artist – simply adjust
your monitoring level until the music is comfortably loud at the loudest moments, and stick
to that level.

Step 3: Adjust
So, you’ve chosen your reference tracks, and matched them all to your preferred
mastering loudness. That means either:

• As loud as your chosen reference track(s)

• Following my “no louder than -10 LUFS” recommendation, or
• Your own particular preferred level

You’ve set your mastering monitoring level so that these songs sound good and loud at the
loudest moments, and you’ve invested in some acoustic treatment for your studio to get it
sounding as accurate as possible. (You have done this, right ?! DIY acoustic treatment is
very affordable…)



You’re ready to start making mastering adjustments to your music, to help make them
sound as good as they can possibly be !



First though, ask yourself an important question:



Would it be better to go back and make changes to the mix ?
This is a luxury you’ll only have if you’re the artist or engineer, but even if you’re mastering
things for other people it’s an option.



Often there isn’t time, or money, or energy (!) but it’s always worth considering. Usually the
results you get by improving the mix are even more effective and successful than trying to
achieve the same thing at the mastering stage.



Assuming you’ve done all you can with the mix though (and further tweaks aren’t an
option), it’s time to get started. First of all, let’s talk about the kind of processing I use when
I’m mastering.

My Mastering Chain
I see so many posts online about different mastering processing chains. They’re all
different, and often rely on one or more “magical” pieces of gear or plugins.

The one thing that most of them have in common is that they’re often really complicated –
7, 8 or more processors in a row, often with multiple EQs, compressors, maximisers and
more.



My chain is different, in two important ways.



First, it’s simple. It has only 4 stages.



Here they are:


• EQ
• Compression
• Limiting
• Metering


That’s it !

Of course I sometimes vary the order, and I sometimes add extra elements to achieve
specific tasks – stereo width processing, for example. But probably as much as 80% of
everything I master is done using only these 4 processing stages. And the meters don’t
even do anything to the sound !



And second, I don’t really care what I use at each stage.



Actually that’s not strictly true - I do have favourite plugins and bits of gear, just like anyone
else. And some of them I’d miss if I couldn’t use them.



But I could use almost anything and get great results, if I had to. I often go to colleges and
do a workshop where I use only stock EQ and a limiter to make dramatic improvements to
a collection of songs provided by the students. In fact, I made a video of exactly that
process, you can watch it here:



The Power of Mastering (Using Only TWO Plugins)
And actually back when I was first trained, many albums were mastered in exactly this way
– compression was still quite rare in the mastering studio back then.



So, let’s look at each of those stages in detail. I’m going to start at the end of the chain and
work backwards.

Mastering Metering

Your monitoring is the real final stage in your mastering chain, which is why it’s so crucial
to get right. And of course your ears are the best possible judge of what sounds best.



There are many useful meters that can help us while we’re learning though - LUFS
loudness meters, as we’ve seen, but also frequency analysers for example, that help give
an idea of the overall EQ balance.



When I’m mastering, I tend to use:


• An LUFS loudness meter

• A True Peak meter (to make sure there’s no clipping) - see Step 5
• A frequency analyser - for example Voxengo SPAN is a great free option - I’ll talk
more about it in a minute
• A correlation meter - these help give an idea about the stereo width of the signal.
Again I’ll talk more about this in a minute

My other favourite meter is a plugin emulation of an analogue needle-style VU meter. My

favourite is the Klanghelm VUMT, which has a great name and is extremely affordable.

I was trained using a VU meter to judge loudness, and although they’re a bit quirky in
comparison to LUFS meters, they have the huge advantage of being most sensitive in the
middle of the range. You only have to go a dB or two too loud and the needle “pegs”, and if
the level is too low it will rarely reach zero, which is easy to notice.

So when I’m mastering, I use a VU meter calibrated to -11, and aim to have the needle
hovering around zero in the loud sections, pushing 1 or 2 dB into the red occasionally. This
gives results that completely fit with my “no more than -10 LUFS” recommendation. Try it,
you might like it !



You can see more about VU meters and how they relate to LUFS loudness in this video:



LUFS, dBFS, RMS… WTF ?!? How to read the new loudness meters




Limiting and Compression

These are two separate stages in the processing chain really, but I use them together to
get the best results. In fact, strictly speaking a limiter is just a compressor with a very high
ratio, and very fast attack and release times.
My goal with compression and limiting in mastering is to be invisible. In mixing,
compression can be an important creative tool, changing the sounds dramatically, but in
mastering I want people to just think the music sounds amazing without knowing why, not
think “wow, that’s some great mastering compression”.

Limiters are best at handling transients cleanly, and compressors are better at handling the
“body” of the sound. If a limiter is working so hard it’s cutting into the vocal, or bass guitar
for example, it’s almost certainly going to sound poor.

So my recommendation for mastering is:


• Always have a high-quality mastering limiter at the end of the chain

• Start by adjusting the input gain before the EQ so your song is at about the right
loudness - more on this later
• Remember the EQ comes before the compressor so it should already balanced - I’ll
talk more about this in a minute, too
• If it sounds great already, you’re done ! No compression needed
• If it needs more “glue” or dynamic control, try adding some compression
• Start with moderate settings, for example 2:1 ratio, and 100ms attack/release times
• Avoid compressing too heavily, and adjust the threshold so that the compressor
releases back to zero gain reduction several times a bar

Using these guidelines will enable you to get great results on a really wide range of
material.

By adjusting the loudness using a simple gain change first, you minimise the amount of
tweaking of threshold and make-up gain, and hear less level change when you bypass the
compressor to see if you like how it’s working, reducing the risk of being fooled by the
loudness deception.

Avoiding very fast attack and release times in the compressor will allow the transients to
come through much more cleanly - and the limiter can deal with those, which is what it’s
good at. But the compressor will control the body of the sound more gently, and help ease
the load on the limiter, which also helps it do it’s job better.

Making sure the compressor releases back to zero gain reduction regularly helps avoid
over-doing things. I’ve written more about this topic, and why you might sometimes want to
break the rule, here:



When over-compression sounds great

Finally, I often like to use multi-band compression when I’m mastering, to get even more
invisibility. There isn’t space to go into it fully here, but you can find out more about how I
use it in this post:



Multiband compression – the mastering engineer’s secret weapon?

Remember I mentioned above that it’s important to have balanced EQ going into the
compressor, which is why I usually apply EQ first. In the next section we’ll look into this in
more detail.

EQ in Mastering

EQ is probably the most important element of mastering. So often I see people throwing
more and more processing at their audio, without getting the EQ right first.

A tiny change of only 1 dB or more can make a snare seem to open out, or tame a harsh
vocal, or really help a kick thump. Whereas other times very large changes are worthwhile
– for example if a song was mixed on very small speakers and has way too much bass as
a result.

Of course there’s a limit to what EQ in mastering can influence, but I’m still amazed at how
effective it can be, even after all this time. We can’t really choose to only EQ once element
of a mix, but quite often we can make it seem as if we can !

As I mentioned above, it’s important to get the EQ balanced before compression – here’s
why:



Imagine a mix with a very thumpy kick drum – too much 50 Hz energy, say. If we compress
that without EQ first, the chances are it’ll cause pumping – the compressor will react the
the kick and pull the overall gain back, affecting the whole mix. If we then add EQ after the
compression to take the excess 50 Hz out, the EQ will sound balanced but the pumping
will still be there. In fact, it could be more noticeable because now the heavy 50 Hz that
caused it has been removed.



But now imagine the same situation where we apply the EQ to reduce the 50 Hz boom
before the compressor instead. Now the overall EQ is still balanced, but with less (or no)
compressor pumping. The compression is less noticeable, more invisible – mission
accomplished !



You can hear a more complicated example of the same effect in this video:



How EQ can make your music sound louder – using LESS compression and limiting



So the goal of mastering EQ is balance - but what is balanced EQ ?



In a nutshell, you want to avoid large lumps or dips at any point in the frequency spectrum
– IF you have a full arrangement. So a full band should probably have fairly even
frequency response across the whole range, whereas an acoustic guitar & vocal ballad is
unlikely to have pounding bass or sizzling high frequencies.



The best way to train your ears to this is to start watching a frequency analyser. Many
DAWs and EQ plugins have them built in, or you can get something like Voxengo SPAN
for free. If you decide to try it, I recommend switching it to Master Mode - this will give you
a much less detailed, slower overall display, which makes it easier to pick out the overall
EQ shape without being distracted by too much detail. Look for big gaps or build-ups, and
experiment with tweaking them.

And finally, remember to adjust the loudness of your music before you start to add EQ. If
you EQ when the level is low and then turn it up later, you’ll hear more bass and treble (the
loudness deception) and have to re-adjust it. Many EQs allow you to change the gain
internally, but I recommend you do it as a separate process beforehand, to minimise the
change in loudness when you bypass the EQ to hear it’s effect clearly.

Stereo Processing

I mentioned that I do add extra stages to this chain sometimes, and one of the most
common goals is to tweak the overall stereo image. There isn’t enough room to go into this
in detail here, but if you’re interested you can see some of the options I use in this video:



Stereo width in mastering – how to measure, monitor and modify it

Summary
So that’s it ! You should now understand the entire mastering processing chain. Let’s
recap, from start to end:


• Adjust the gain to achieve your desired loudness

• Balance the EQ
• Add compression if necessary
• Check the loudness again (in case the EQ and compression have affected it)
• Compare to your reference track
• Repeat if necessary

Here’s another video where I walk through the whole process from beginning to end:



The simple secret of real mastering – subtlety and simplicity

And remember – you need to do this separately for each and every song. Mastering is
NOT just adding a bit of global limiting and EQ. Every song needs to be considered and
optimised individually.

This may sound really obvious, but it’s a point that lots of people miss. In fact (guess
what !) I wrote a blog post about it:



The heart and soul of mastering



I gave it quite a ‘big’ title, because I think it deserves it !
Step 4: Repair

In an ideal world, your mixes will sound perfect, with pristine quality.

In reality, that’s unlikely.




Here’s a list of just some of the faults that may crop up in the recording and mixing
process.

• Hum – Amp noise, earthing problems, a nearby road, air conditioning…


• Crackle – A dodgy cable, poor connection, dirty EQ pots, rustling clothes…

• Buzz – Guitar pickups, bright stage lights, amp noise (again !)…

• Clicks – Mouth noises, a creaky chair, dodgy edits, faulty gear…

• Distortion – Overloading inputs, impedance mismatch, too many guitar pedals (!)…
You name it, you’ll hear it, one way or another. And of course ideally we’d go back to the
mix to fix it, but that’s not always possible.



Either way, mastering is the final stage in the quality control chain, and if you notice a fault,
it’s up to you to make the client or artist aware and possibly try and fix it.



You may wonder why I didn’t mention this earlier on, and it’s certainly true that problems
like this are best fixed as early as possible. But often the mastering problem will enhance
or reveal a fault in all it’s glory, that hadn’t really bothered you before. Or, it can just be the
process of putting the music under the microscope that makes it bother you. So ideally, if
you notice it during mastering, go back to the source and try and fix it there, if you can. But
it’s almost always at the mastering stage that you’ll notice.

I’ll talk about some of the ways we can help improve problems like these in a minute, but
before I do you may have noticed there’s one fault I didn’t mention - hiss. And it’s actually
one of the most common “faults” we come across, and something that many people
obsess about.



It’s always been here – some of our favourite classic recordings are bathed in hiss ! Tape
noise, amp noise (again !), poor gain-staging, air conditioning units… you name it.


But here’s the thing - hiss is our friend.



OK, not always - not if it’s seriously distracting. But a bit of white noise never hurt anyone,
and our ears quickly tune it out. Added to which, even the most sophisticated modern
processing adds annoying artefacts when removing hiss. It’s so easy to over-do it, and
end up with something worse than the original problem.



So de-hiss if you must, but maybe just try rolling out a little top end in the quiet sections
instead - it can often sound more natural.

Removing faults

A detailed discussion of repairing all the problems we’ve been talking about (and more)
could make a small book in itself.


Instead, I’m going to give you some (hopefully) helpful pointers to help you get better
results.

1. Get IZotope RX




Many plugins and DAWs offer tools to report and remove faults, and some of them are
pretty good. By all means give them a try, but the best value I’ve discovered so far is
the amazing iZotope RX. 



Clicks, thumps, pops, buzz, hum, crackle, mouth noises, lost sheep – yes, even hiss.
There’s almost nothing RX can’t at least help with.



On the downside it’s a deep, complex application, and takes a lot of work to master.
But if you’re serious about trying to repair faults in mastering, you need RX. And
especially, learn how to use the Spectral Editor.


2. Be creative – experiment



Even with a heavyweight tool like RX, it can take time to figure out the best way to
repair certain faults. 



For example, I mentioned the buzz caused by powerful stage lights that you
sometimes get into a live recording. The interesting thing about this particular type of
fault is that although it sounds like a buzz, it’s actually often made up of tiny clicks, at a
very high rate. So the best way to remove it isn’t necessarily a de-buzz process, but
actually a de-click pass. And so on.

3. Don’t give up




This follows on directly – keep trying. It can be discouraging if a processor that you
think should be perfect for the task completely fails to get a result you’re happy with,
but that doesn’t mean you’ll never succeed. Keep trying ! Especially if you’re trying to
repair That Perfect Take. 


4. Always go back to the source, if possible




I know I’ve already said this, but working on the master can be the worst place to repair
faults. 



I shudder to think of the number of times I’ve spent an hour or more removing
annoying mouth noises in a song, only to have the artists decide they want a mix
tweak and send me a new version, so I have to do it all over again !



And if you really must de-hiss something, it’s far better to work only on the problem
 channel, rather than applying the processing to the whole mix. Wherever possible,
repair it as soon in the processing as you possibly can.

Time For Action (Again)

So there you go ! By this point you’ve learned to optimise your mastering monitoring,
chosen the loudness you want to master at, compared you mixes to suitable reference
material (at matched loudness), adjusted the EQ, compression and limiting to get the best
possible result, and repaired all the faults.



All you need to do is give all these ideas a try, and see how you get on.


So once you’ve finished adjusting your songs and they’re sounding great - what’s left ?



Quite a bit, actually ! First, you have to make sure you’re happy with the result of your hard
work. And next, you have to get it safely onto it’s destination format or platform.



Neither of these steps are very difficult, but neither of them are entirely simple, either.



Let’s dig in !

Step 5: Preview


Surely we already “previewed” our work by listening to it during Step 3 ?




Yes, but now we need to test it out more thoroughly, in three different ways:

1. Out in the real world

2. After encoding
3. Online

Let’s look at each of these in turn.

Previewing masters in the real world

“In the real world” could mean almost anywhere, but ideally it should be places you often
listen to music, so you have an instinct for how it should sound.

So for example, this could include 2-3 of any of these options:


• On earbuds or a great pair of headphones

• In the car
• On a smart speaker
• In your living room stereo
• Through your TV
 • On a laptop
• At a friend’s house
…and so on.



Don’t bother listening in places or on systems you don’t know well, though. One of my
favourite sayings is “A man with a watch may know what the time is, but a man with two
can never be sure”. This applies absolutely to audio, but with knobs on !



In other words, there’s a real risk of getting utterly confused.



To avoid this, I suggest you use this process:

• Listen in the Real World

• If you notice something, make a note
• Listen again back in the studio
• If you hear something – boomy bass or harsh mids, say – adjust it
• Repeat as necessary

If you don’t hear the same thing in the studio, listen again somewhere else. If you never
hear that problem again, it was probably just a quirk of the particular playback system you
were listening on.

But if you notice it in other places too, there’s a chance you’ve noticed something that
you’re not hearing in your studio, and you need to be careful.

A boomy bass note would be a classic example - the room acoustics in your studio might
prevent you from hearing a particular frequency accurately, so you end up adding too
much of it. You’ll then hear this frequency booming out in plenty of other places. Time to
experiment with speaker placement or more acoustic treatment, perhaps !

In the same way, if you keep hearing something on your favourite real world listening
system, but no-where else, you’ve probably identified a quirk of that system, and can
safely ignore it if you hear it in future.



As you can probably tell, this is a continual learning process ! The good news though is
that every time you go through the process, you learn more about mastering, more about
what makes your music sound great, and more about your mastering monitoring.

Previewing masters after encoding

The reality is, your music will be ripped, meaning: be encoded to some form of lossy
codec. MP3, AAC, Ogg Vorbis, whatever.



And all these formats change the sound. Not always massively, but in important ways.
There’s not a huge amount we can do about this, but we can follow a few simple
guidelines to minimise the damage.



The first of these is to test, by listening to encoded versions of our files. Either by encoding
them ourselves, or more conveniently by using one of the pieces of “codec preview”
software that are available.

These allow you to listen to encoded versions of our audio in real-time, in a variety of
formats. Some even allow you to listen to the “null” signal, or the information that has been
thrown away during the encoding process – mulchy distortion, most of the time.

Whatever the result, there’s not usually much you can do about it, but with one exception.

PAY ATTENTION TO THE PEAKS

Almost all encoded files have higher peak levels than the original. This is because the
encoding process involves very aggressive processing. Very loosely speaking, the audio
gets filtered into 30 or more separate bands, up to 90% of the information is thrown away
and the bands are then rebuilt into something (that sounds surprisingly like the original) in
the decoder.



When you think of it like this, it’s not that surprising that the peak levels change !



The problem is, often these peaks are above 0 dBFS, and many of the mainstream
decoders slice them straight back off again immediately after decoding, which can cause
extra distortion.

The solution is to reduce the peak levels slightly before encoding, so that the decoded
peaks don’t go above zero, which is one of the reasons I suggest keeping peak levels to -1
when you’re mastering.



In fact, I recommend -1 dBTP. The TP stands for “True Peak”, which is basically an
oversampled version of the normal peak level, to try and predict when problems like this
might occur.

The other thing that helps is to not master too loud. The higher the LUFS loudness, the
more likely the decoded peaks are to go above zero. I’ve seen examples as high as +4 !

If you’re following my suggestion of “no higher than -10 LUFS”, with True Peaks no higher
than -1, you should be pretty safe. If you’re mastering louder than this though, you need to
pay attention to this. Spotify even recommend you keep peaks below -2 dBTP if you’re
mastering louder than -14 LUFS.

Of course in theory you could supply masters for encoding at slightly lower levels, but in
my experience even people who still buy CDs also rip them, so I think it’s better to simply
keep to peaks no higher than -1 dBTP, on any format.

To hear an example of this in action, take a look at this video:

Is Billie Eilish too loud ? (Here come the Loudness Police)

Previewing masters online

You’ve probably realised by now that I pay a lot of attention to what happens to the music I
master online. That’s because it’s where most people listen, these days !



And the two factors that affect the quality online are lossy encoding, as we’ve just seen,
and loudness normalisation.



Overall I think normalisation is a Good Thing - it gives us a better user experience, and it
means people don’t have to “compete” with loudness in the way that they used to on CD.
But it’s still quite a new technology. It’s doesn’t always work the way we expect it to, or
would like, and all the online platforms do things a little differently. Which is why I worked
with MeterPlugs to develop the Loudness Penalty website.


In a nutshell, the site tells you how the loudness of your music will be changed, on all the
most popular streaming sites. For free, and without uploading !

More importantly though, it allows you to listen to a realtime Preview of the loudness
change, so you can directly compare the results for your music on YouTube (say) against a
suitable reference track playing directly from the platform in another browser tab.

This is a powerful and useful technique, especially for Spotify, who DON’T yet use LUFS.
They use something called ReplayGain instead, which often gives similar results but can
sometimes be different – by as much as 3 dB. That’s a big difference, and it can be very
helpful to hear the effect of that, and adjust for it if necessary.


You’ll notice I keep stressing the importance of using Preview to hear the effect. One
downside of the site is that people are getting very hung up on the numbers. Remember,
they aren’t targets.

Personally if I find something I’ve mastered is being pulled down more than 3 dB on
YouTube, I can’t help wondering if I mastered it a bit hot, and experimenting with a more
dynamic master. But that’s just me – if you love the sound of your master compared to
other reference tracks even though it’s been turned down by 6 dB or more, great ! The
important thing is to Preview and check.



Knowledge is power.
Step 6: Export
OK, we’re there ! You mastered your music, you checked it out in the real world, listened
through some lossy encoders and on the Loudness Penalty site, and you love it. You’re
good to go!

All you need now is to export it in the correct file format to make sure all your hard work
actually translates as you’ve intended. And luckily, this is simple. Here’s the formula:



16-bit, 44.1 kHz stereo lossless (WAV, AIFF, FLAC) 


Use this format for everywhere except CD, iTunes and vinyl. In fact you can even use to
for them too, if you like.

Why ? Actually, that’s a good question, and it’s worth knowing the reasons. Plus we need
to look at those exceptions…

CD Masters

CD is a stereo 16-bit, 44.1 kHz format. If you supply individual files though, all kinds of
things can go wrong. For CD masters you’re far better off supplying a DDP Image master.



DDP images include all the audio in a single package, built-in error-checking to help
ensure the file doesn’t get corrupted, and can contain some limited meta-data, like song
titles and ISRC codes.



Several DAWs support DDP export - if not there are several standalone applications, for
example made by HOFA and Sonoris.

If you’re using a company like Distrokid, CD Baby or similar to distribute your music,
including making CDs, you can send files - but personally I always use DDP.




Online Masters

I’m including online distributors like DistroKid in this category, since they’ll upload your
music to all the major streaming platforms for you. It also applies if you upload the files
yourself, though.



First of all, never supply encoded files like MP3. If you’re lucky the online services will use
the files as-is, but there’s always the risk of double encoding (which can sound ugly) and
you would never want a CD made from MP3. Ugh.



So, as I said, my recommendation is for 16-bit, 44.1 kHz stereo lossless files. This is
actually a little controversial though, because lossy codecs like MP3 and AAC can actually
encode signals at higher effective bit-depth than 16-bit.



So why not supply higher bit-depth files ?



The answer (again) is that you can’t guarantee reliable results from the decoders. Some
will correctly reduce the files to 16-bit when necessary using dither, but others won’t.

(What’s dither ? You don’t want to know. You do ? OK, I’ll come back to it later – but only
briefly.)



There’s also some doubt that all streaming services will make a good encode from higher
bit-depth files, if you supply them. They should, but I’m a control freak and I don’t trust
them, which is why I recommend higher bit-depth !




ITUNES

The one exception to this is iTunes. The Mastered for iTunes system can accept higher
bit-depths and sample rates, and in fact 24-bit audio is the minimum requirement to get the
Mastered for iTunes badge. It needs to be mastered by an approved engineer or facility
though, and often involves an extra charge.

Vinyl Masters

The good news is that if you follow the guidelines in this guide, your masters will be in
great shape for vinyl. You probably want to submit 24-bit files though, and at a higher
sample rate, because why not ?

I also quite often remove the final limiter. It’s very hard for the cutting head to reproduce
heavily limited material accurately, and in fact they can burn out if the loudness is too high,
so if you submit a super-loud master for vinyl the cutting engineer will just turn it down
anyway.

Finally, you may be asked to supply two files – Side A and Side B – with timings, by some
manufacturers.



I’ve also written more about mastering for vinyl here:

Why custom mastering for vinyl isn’t necessary

Don’t dither about dither

One last thing – when you’re exporting your final file, always use dither.



Dither is very low-level noise, added when you reduce the bit-depth to prevent truncation
distortion. (See ? I told you you didn’t want to know.)



Seriously though, if you’re saving to 16 or 24-bit, make sure you apply dither. And if you’re
really a glutton for punishment, here are a couple of blog posts on the subject:

Dither or distort ? You decide

When in doubt, dither

Time For Action !

That’s it.

You made it.



You now understand all 6 steps to releasing your music with complete confidence – so go
for it !



Don’t be tempted to spend all your time on Step 3, though – messing around with
mastering processing is fun, but without the other steps you’ll never get the results you’d
like.

Put the time into improving your monitoring, learning your room (and how to loudness-
match) and pick some great reference tunes.

Compare them carefully with the songs you want to master.

Adjust the sound to make it sound as good as it can be, repairing any faults you might
hear.



Preview it, both in the real world and online.



And finally, export it with care and confidence.

This is the “home” mastering guide, but actually you’ve probably guessed, this is how I
master everything, not just at home. Because the ideas, methods and strategies in this
guide work, and now you can use them, too.



I want to say thanks again for downloading this guide, and for reading it ! If you ever get
stuck with mastering your music, I might be able to help. There are hundreds of articles
and videos about mastering on my website and channel, and you might like to try listening
to my Mastering Show podcast as well.

Come and say “Hello” on social media, too ! I can’t guarantee to answer every question,
but I’d love to hear from you, and if enough people ask me the same thing there’s a good
chance I’ll write a post or record a video about it sometime soon. 


Now, stop reading this and build yourself some acoustic treatment panels !

Make your music sound the best it can be.




Cheers,



Ian (Production Advice)