6 Steps To Releasing Your Music 


with Complete Confidence


Part 3
by Ian Shepherd

Welcome to Part 3 of the Home Mastering Guide. So your songs have been mastered,
they sound 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)