[00:00]
For some people, getting loud and clean
[00:01]
music seem to be impossible. But there
[00:03]
are five steps that I have never seen
[00:05]
anybody talk about in a single video
[00:07]
that will get your music to sound as
[00:08]
loud as you wanted to while making it
[00:10]
cleaner and crisper. My name is Max.
[00:12]
Welcome to Mix Elite Academy. And how
[00:14]
come this track here we get [music]
[00:19]
about minus 10 LFS. We are even peing 1
[00:22]
dB and I cannot push it harder [music]
[00:26]
because it just start distoring. But the
[00:28]
same track with different processing
[00:30]
applied. I can push to plus 17.
[00:36]
You're getting -5 - 6 LFS. [music] We
[00:39]
are not picking
[00:42]
and it sounds clean and crispy. Louder
[00:45]
equals better. Like literally
[00:46]
psychologically and physiologically.
[00:48]
Docs hate it. We love it. But here is
[00:51]
not just louder. It's crispier,
[00:52]
punchier, and glued together. The very
[00:54]
simple way to achieve this loudness is
[00:56]
by playing our song through a limiter
[00:58]
that doesn't allow the audio to get
[00:59]
above zero dBFS threshold, which is the
[01:02]
limit for your speaker to play music,
[01:04]
and turning on the gain. So, we shave
[01:06]
off the peaks of the sound, making the
[01:08]
[music] louder parts quieter and the
[01:10]
quieter parts sound louder.
[01:13]
[music] And this is exactly what I've
[01:14]
just done to this track. is a cardinal
[01:16]
sin VDM right here
[01:17]
>> that you should never ever do. So,
[01:19]
what's going on and what should you do
[01:21]
to get the crispy masters? First, let's
[01:23]
clear up that the PQC on the meter
[01:25]
inside your door has nothing to do with
[01:27]
how loud your music is actually
[01:28]
perceived. We measure this with LUFS
[01:31]
instead. It's a separate metering plugin
[01:33]
you have to use. If you don't have one,
[01:34]
just click the link in the description
[01:35]
below. As electronic music producers, we
[01:38]
want this number to be as high as
[01:39]
possible, ideally to at least minus 8 to
[01:42]
-6 integrated LUFS. Scril X goes minus
[01:45]
three. We'll get to how in a moment.
[01:47]
Second, why did you just say louder
[01:49]
equals battery max when Spotify tells me
[01:51]
that recommended is minus4 LFS or they
[01:53]
will lower it? Because people are
[01:55]
confused between distribution target and
[01:57]
playback threshold. Spotify uses minus14
[01:59]
LUFS as playback settings. Is their
[02:02]
target for average integrated LUFS of
[02:04]
your entire song? If your drop is
[02:07]
hitting minus1 LUFS and it still sounds
[02:09]
great, that's amazing. No streaming
[02:11]
platform is going to be applying limiter
[02:13]
on your song. People will just turn up
[02:15]
the volume as they want it on their
[02:17]
studio Bluetooth speakers and Spotify
[02:19]
might just change the volume. So, we
[02:20]
always master our tracks as loud as
[02:22]
possible without it starting to sound
[02:24]
worse. Full stop. Next, before we start,
[02:26]
fundamentally, your muscle will suck and
[02:29]
not be crispy. Not because of what you
[02:31]
do on the master mainly, but what you do
[02:33]
while mixing. If you want a clean and
[02:35]
loud mix before we get to mastering,
[02:37]
your friends are right click delete
[02:39]
button. You cannot make 50 misins loud
[02:41]
because they're fighting for the same
[02:43]
pixel of headroom. Selfish low end
[02:45]
because low end is carrying the most
[02:46]
energy but least loudness. That's why if
[02:48]
you listen closely, protras actually
[02:50]
don't have as much sub frequencies as
[02:53]
you might think. Frequency masking, a
[02:55]
lot of side chaining, not only from kick
[02:57]
to bass, but to make space for whatever
[02:59]
is the priority in your mix. Clipping
[03:01]
individually and groups to control the
[03:03]
peaks and everything below 100 htz is
[03:06]
mono. These guys will give you 80% of a
[03:08]
clean mix. that we now can get to making
[03:10]
loud and crisp. So, let's start from the
[03:12]
loudest part of our rendered mix.
[03:19]
This applies to every genre, ADM, trap,
[03:22]
hip-hop, etc. First, on your master,
[03:25]
since low frequencies are the heaviest,
[03:27]
we always apply a low cut to cut
[03:29]
everything that's unnecessary. So, I'll
[03:31]
apply Fruity Parametric EQ2. And
[03:33]
unnecessary starts just under the lowest
[03:36]
note that your song plays. that
[03:38]
shouldn't be lower than 30 Hz. For
[03:40]
example, D1 or D2 in FL
[03:43]
place at 36 Hz. In FL, there's this
[03:47]
preset you can apply. And I will take it
[03:49]
a little bit higher because my track is
[03:51]
in E. It hits at around 41 Hz. So, I'll
[03:54]
just start rolling down after it. You
[03:56]
should not turn on linear phase or
[03:58]
oversampling. Linear phase and
[04:00]
oversampling will leave for the final
[04:02]
limiter.
[04:13]
You can hear it's getting worse within
[04:15]
our face and over something applied.
[04:17]
Some people like to also use shelf
[04:19]
instead. That would look something like
[04:21]
this. You use this shelf if you feel
[04:23]
like after the low cut your track became
[04:26]
boommy, right? There are some phasing
[04:28]
issues. Then we don't need any lows from
[04:30]
the side channels because side channel
[04:32]
is everything that's different between
[04:34]
left and right ear. Side is what creates
[04:37]
stereo width effect. And serial width in
[04:40]
low frequencies is bad. Very bad. Want
[04:42]
to know why? Right? Because they
[04:44]
oscillate in lower frequency. And if two
[04:46]
play the exact opposite signal, you'll
[04:48]
hear silence. Not in your headphones,
[04:51]
but when you listen back to it on mono
[04:53]
speakers or there will be blind spots in
[04:56]
clubs with no bass audible at all. I
[04:58]
will use a free midside EQ preset for
[05:01]
Patcher that you can download for free
[05:02]
below and just start increasing the side
[05:04]
low cut.
[05:18]
So this is how it looks like from two to
[05:20]
300 Hz. We're starting to roll off
[05:23]
everything so that at 120 Hz everything
[05:26]
down is only mono. There's no side
[05:28]
signal and this is what we are cutting.
[05:30]
Take a listen and notice how the phase
[05:32]
shifts. So the bass is inaudible every
[05:35]
couple of seconds.
[05:43]
So it's also really inconsistent and it
[05:45]
will screw up our face a lot. That's why
[05:47]
we do it like this. And if you worry
[05:48]
that we are losing a little bit of this
[05:50]
punch and bass, we'll be adding the good
[05:52]
one back in a moment after we would get
[05:55]
into dynamics. take a look at our
[05:57]
waveform, how many weird little peaks we
[06:00]
have. That's why we would usually apply
[06:03]
a clipper just after the EQ. If we
[06:06]
wanted to really push the loudness, I
[06:08]
really like this small plugin, literally
[06:10]
very small, called the G-Clip, you push
[06:13]
it until you just barely hear it change
[06:15]
the knock of your drums, then back it
[06:18]
off 10%.
[06:33]
And now we also turn on the over sample.
[06:40]
We do not turn on softness. We keep it
[06:42]
at 0%.
[06:48]
So I had it started to distort and then
[06:50]
I backed it off. So I don't hear it no
[06:52]
more. Take a look at what waveform looks
[06:54]
like with and without it.
[07:05]
>> We're cutting the unnecessary peaks that
[07:07]
don't really make it sound different,
[07:09]
but we'll screw up our limiter in a
[07:11]
minute. Technically, even better way
[07:13]
would be if we applied multiband
[07:14]
clipping. So you would not be limited by
[07:16]
the whole waveform and you'll find a
[07:18]
whole niche for multiband clipping on
[07:21]
YouTube. But ideally instead you don't
[07:23]
do multiband clipping on your master
[07:25]
when mastering but do it on individual
[07:27]
groups of instruments when mixing. This
[07:29]
is what for example Skrillex does to
[07:31]
achieve this minus 3 LFS. That's why
[07:33]
some mastering engineers will want you
[07:35]
to send stems of your track for
[07:37]
mastering so that they can do stuff like
[07:39]
this. But this is basically mixing
[07:40]
already. We are not doing that today.
[07:42]
Now, why did we use hard clipping before
[07:44]
compressors or a limiter? Because our
[07:46]
limiters are trying to be smart. Their
[07:49]
goal is to be as transparent as possible
[07:51]
without adding distortion. If you have a
[07:53]
huge snare hit like that upcoming, this
[07:55]
smart brain of this thing will soften
[07:58]
the hit to fit the snare under the
[08:00]
ceiling we set over here. On the other
[08:03]
hand, clipper is damp and that's why we
[08:04]
love it. It just chops all the waveform.
[08:07]
That's why the next step after we got
[08:08]
our signal clear from this weird harsh
[08:11]
peaks is multiband compression. You can
[08:13]
do this in Maximus or something like Fap
[08:16]
Filter Promb MB. Inside Maximus, I think
[08:18]
it's pretty easy to see. I like to set
[08:19]
the preggain of each band by like 7 to
[08:22]
10 dB so we have more [music] room to
[08:25]
work with.
[08:35]
Our goal here is just to lower the
[08:37]
dynamic range of every single band
[08:39]
individually instead of at once. And at
[08:41]
the same time, very important, make up
[08:43]
for that. So if we are going down from
[08:45]
plus 8 dB to around five, we would add
[08:49]
about 3 dB of gain to the lows like
[08:52]
that. Then do the same for mids
[08:59]
[music] and highs.
[09:08]
And now we can get to limiting and
[09:10]
maximizing. This is the most important
[09:12]
part. If you were to watch just one
[09:13]
minute of this whole video, this is the
[09:15]
one. I really like to use this one from
[09:16]
Ozone because it has all the features we
[09:18]
really need. First, before we even tweak
[09:20]
anything, we turn on always true peak
[09:23]
and set the output level to start at
[09:26]
minus0.3
[09:27]
dB ceiling. Why this is important?
[09:30]
Because when audio leaves your computer,
[09:32]
it goes through digital to audio
[09:34]
converter DAC from PC to speakers. It
[09:37]
basically takes the dots and draws a
[09:39]
smooth curved line between them to send
[09:41]
to your speakers. If two dots are too
[09:43]
high, the curve can get about zero dB
[09:46]
causing distortions. So speakers try to
[09:49]
play signal that's technically
[09:50]
impossible for them to play and it's
[09:52]
called inter sample peing because the
[09:54]
signal is peing between the samples.
[09:57]
When you turn on true peak, it over
[10:00]
samples the audio to create more dots to
[10:02]
make sure they won't go above the
[10:04]
thresholds you've set after converting
[10:06]
to the signal that speakers can read.
[10:08]
You will not hear these issues inside
[10:10]
your DA because your D works in 32-bit
[10:13]
float at least. But this all only
[10:15]
applies to lossless file, meaning a wave
[10:18]
file that nobody will be listening to.
[10:20]
Actually, everybody will be listening to
[10:22]
something like MP3 versions of your
[10:24]
file. Like maybe DJs on festivals play
[10:26]
wave files, but usually this would be
[10:28]
Spotify converting your file to MP3. And
[10:31]
this settings don't fix this. Your
[10:33]
limiter even in true peak mode doesn't
[10:35]
care about MP3 or any other codec.
[10:37]
That's why inside of ozone you have this
[10:39]
codec so that we can preview how it will
[10:42]
sound and what will be the true peak
[10:44]
after converting to MP3. And this is
[10:47]
key. Okay, we don't even start turning
[10:49]
on the gain before [music] we do that.
[10:51]
So we have true peak minus 0.3 is a good
[10:54]
starting point. And now I'll start with
[10:55]
the chaotic off so we can hear the
[10:57]
difference. And we'll start turning on
[10:59]
the gain up until I hear audible
[11:02]
distortion. No LFS [music] matters for
[11:04]
now.
[11:13]
All right.
[11:17]
So [music] you can see we are under
[11:20]
minus 6 LUFS. This is crazy. And we are
[11:23]
picking at minus0.1
[11:25]
even if you have output level set to
[11:27]
minus0.3 with crooic on. Okay. And now
[11:30]
let's turn on codec. go to for example
[11:33]
AAC because that's what Spotify for
[11:35]
example uses for their web player and
[11:37]
set it to 256 and we'll be looking at
[11:39]
this peak to see if we are clipping
[11:44]
[music] and immediately we're clipping
[11:52]
even though with no [music] codec light
[11:55]
we are pretty much at the true peak Go.
[12:02]
Now this plus one shouldn't be outable
[12:04]
at all. But if we go lower in codec,
[12:12]
it's getting much worse. That's why so
[12:14]
many people will tell you to go lower in
[12:17]
the true peak ceiling. Spotify tells you
[12:20]
to go minus 2 dB true peak ceiling just
[12:23]
because of this codec. If we go minus 2
[12:26]
dB,
[12:30]
we only have half a dB of headroom after
[12:33]
converting it to AAC 128 kilob. I would
[12:37]
personally before limiting apply some
[12:39]
kind of imager to make the stereo width
[12:43]
of our track a bit more rich. I would do
[12:46]
something sort of like that where we
[12:48]
have lows fully in mono and mids to
[12:51]
highs a bit more stereo. This will also
[12:54]
increase a little bit of our headroom.
[13:02]
The problem I see here is that our
[13:05]
[music] peaks are still a little bit too
[13:08]
high even though we've applied clipper
[13:10]
already. If you seen something like
[13:11]
this, what I would do after our
[13:13]
multiband dynamics, I would apply
[13:15]
another G clip just to make sure we are
[13:18]
controlling those tiny little peaks that
[13:21]
don't really bring anything. But
[13:23]
unnecessary volume peaks.
[13:36]
As you can see, now [music] we are much
[13:37]
more in control environment. Now, if I
[13:40]
just apply a limiter and apply the same
[13:42]
amount of gain we've added, this is how
[13:43]
it would sound.
[13:49]
And this is after everything we've done
[13:51]
today.
[14:02]
>> [music]
[14:02]
>> Even though I'm trying to give you my
[14:04]
best here on YouTube, there's just so
[14:05]
much that I can cover. That's why inside
[14:07]
our Mix Elite Academy, you get lifetime
[14:09]
access to all of our premium courses,
[14:11]
all of our samples, presets, song
[14:13]
templates, including this one, and
[14:15]
weekly live group coaching, unlimited
[14:17]
one-on-one feedback on your music,
[14:19]
discounts on thousands of plugins. So,
[14:21]
if you want a foolproof way to start
[14:23]
making professional sounding music,
[14:24]
check the first link in the description.
[14:26]
And remember that the best way to get a
[14:28]
loud master is to make music that people
[14:30]
want to turn up. Quality is the best
[14:32]
volume booster. That's why in this video
[14:34]
you can see here, I'll take you through
[14:36]
the start to finish process of mixing
[14:38]
your track. So click over and I see you