Normalization - Why It Works And Why It Doesn’t Make Everything “The Same Loudness”

KEY POINTS

Streaming platforms don’t change your master - they only adjust playback level: Spotify and Apple Music measure integrated LUFS and turn playback up or down to hit a target. They do not remaster, compress, or alter your file.

-14 LUFS is a playback target, not a mastering target: Spotify normalizes to about -14 LUFS and Apple Music closer to -16 LUFS - but mastering to those numbers often sacrifices the density and energy modern productions rely on.

Two songs can measure the same LUFS and feel completely different: Perceived loudness is shaped by arrangement, density, transients, vocal level, and contrast - not just a number on a meter.

Dynamic songs can feel bigger after normalization than constantly loud ones: Because LUFS is an average, tracks with contrast may have lower integrated loudness - which means their loud sections can hit harder once normalized.

Production choices often impact perceived loudness more than mastering does: Controlled low end, focused midrange, vocal placement, and arrangement density determine whether a song feels powerful - even at the same LUFS.

We subconsciously judge loudness by the vocal: A forward, confident vocal makes a track feel louder and more present - even without aggressive limiting.

You can chase apparent loudness - but it often costs musicality: Overly aggressive upper mids, reduced dynamics, and hyper-forward intros may grab attention, but they can create fatigue and shrink emotional impact.

Impact comes from production and contrast - not from chasing LUFS: Normalization is automatic. Your arrangement, balance, and mastering decisions are where the real leverage lives.

In Part 1, we covered what loudness is and how we measure it.

Now let’s talk about normalization.

Because this is where most of the confusion starts.

What Audio Normalization Actually Is

Audio normalization is the process of measuring a file and adjusting its level to hit a target.

That target might be:

• A peak level (peak normalization)

• An average loudness value like LUFS (loudness normalization)

  
  

On streaming platforms, we’re almost always dealing with loudness normalization.

Platforms like Spotify analyze your track when it’s delivered. They measure its integrated LUFS, then adjust playback gain up or down to meet their target.

Importantly, they do not change your file. They don’t remaster it. They don’t compress it. They just adjust playback gain.

If your master is louder than the platform target, it gets turned down. If it’s quieter, it may get turned up (within limits).

When an album plays front to back, Spotify normalizes the entire album as one unit. Your internal dynamics stay intact. Quiet songs stay quiet. Loud songs stay loud.

When tracks are played in playlists or shuffled, normalization happens per track. That’s when your song is being level-matched against everything else.

And that’s when artists usually ask:

“Why doesn’t my song sound as loud as XYZ?”

Streaming Loudness Targets - And Why They’re Misunderstood

Spotify normalizes to roughly -14 LUFS integrated.

Apple Music is closer to -16 LUFS integrated.

It’s easy to assume:

“Okay. I should master to -14 LUFS.”

Even Spotify suggests targeting -14 LUFS integrated with a -1 dBTP ceiling.

But the reality is we do not master to the normalization target.

Why? Because normalization happens after delivery.

If your master feels right at -10 LUFS, Spotify turns it down. If it feels right at -12 LUFS, it turns it down less. If it feels right at -8 LUFS, it turns it down more.

The internal balance of your master doesn’t change.

And there’s another layer people miss:

The sound of density and compression is part of modern music.

A pop song that genuinely measures -14 LUFS before delivery often won’t have the saturation, glue, and controlled density we expect from contemporary productions.

Loudness isn’t just playback level. It’s texture. It’s excitement.

Chasing -14 LUFS can produce technically compliant masters that feel underpowered next to commercial releases.

Meters are tools. They are not creative targets.

The goal of mastering is impact, not compliance.

LUFS Is a Measurement. Loudness Is a Perception.

Two songs can measure the same integrated LUFS and feel wildly different.

Why?

Because perceived loudness is shaped by:

• Arrangement

• Density

• Spectral balance

• Transients

• Vocal level

• Internal contrast

That’s production.

Normalization equalizes averages. It does not equalize energy.

Integrated LUFS Is an Average - And Averages Hide Dynamics

Integrated LUFS is a single number representing the average loudness of an entire song.

But songs aren’t averages. They’re moments.

Imagine:

Track A

• Constant energy throughout

• Measures -7 LUFS integrated

Track B

• Open verses

• Huge choruses that hit just as hard as Track A

• Measures -9.5 LUFS integrated because of quieter sections

Both get normalized to -14 LUFS.

Track A gets turned down 7 dB. Track B gets turned down 4.5 dB.

Now Track B’s choruses are effectively louder.

Internal contrast lowered its average.

This is why dynamic songs can feel more explosive after normalization than tracks that were loud the entire time. Integrated LUFS does not account for internal movement.

Contrast creates impact.

Arrangement and Density Matter

When someone asks why their track doesn’t feel as loud as a modern pop release, it’s rarely just a mastering issue.

High-impact productions are built for energy:

• Controlled low end

• Focused midrange

• Shaped transients

• Strategic arrangement

• Precise vocal placement

Nothing is accidental. A dense, intentional production often feels louder than a loosely arranged mix, even at the same LUFS.

At the same time, a sparse track with strong contrast can feel bigger in its loud sections once normalized because its average level is lower.

Again - normalization equalizes averages, not intensity.

Your Brain Tracks the Vocal

Our ears prioritize the human voice.

We subconsciously use the vocal as a loudness reference.

If the lead vocal is forward and confident, the track often feels louder - even without extreme limiting.

If the vocal is tucked or masked, the song can feel softer at the same LUFS.

Two tracks at -14 LUFS can feel completely different simply because of vocal placement.

Meters measure signal. They don’t measure perception.

Loudness Tools Are Helpful - But Limited

Spotify’s recommendations and loudness penalty tools are useful.

They are also incomplete.

They estimate how much your track will be turned down. They don’t predict how it will feel once normalized.

A dense dance track at -7 LUFS and an acoustic song at -10 LUFS can feel equally loud after normalization. In some cases, the acoustic track even feels more forward because of midrange focus and dynamic contrast.

Integrated loudness is a simple average. It ignores internal dynamics.

Normalization solves consistency.

It does not solve perception.

Can You Game Loudness Normalization?

Some try.

Common tactics include:

• Over-emphasizing upper mids for presence

• Mixing with minimal internal dynamics

• Making intros aggressively forward

These approaches can create strong initial impact.

But they often lead to:

• Listener fatigue

• Harshness

• Reduced depth

• A smaller emotional arc

Attention isn’t the same as impact. Loud isn’t the same as powerful.

Musicality wins.

The Real Question

Instead of asking:

“Why isn’t my song as loud as XYZ?”

Ask:

“Is my arrangement, balance, and energy delivering the emotion I want?”

Once your song hits streaming platforms, normalization is automatic.

But production and mastering choices are not.

That’s where the real leverage is. That's where musicality and artistry live.

Post written by Ruxandra Dedu, edited by Kristian Montano.