Audio Normalization: Making All Your Tracks the Same Volume

You know that moment when you're listening to a playlist and one song is whisper-quiet so you turn the volume up, then the next track absolutely destroys your eardrums? That's an audio normalization problem.

If you've ever made a playlist, edited a podcast, or compiled audio from different sources, you've felt this pain. And it's not just annoying — it's genuinely bad UX. Your listeners shouldn't have to ride the volume knob like a DJ.

Audio normalization is the fix. But here's the thing: most people get it wrong because they don't understand what kind of normalization to use. Let's fix that.

What Is Audio Normalization, Actually?

Simply put, normalization adjusts the volume of an audio file so it hits a specific target level. The entire waveform gets scaled up or down proportionally — loud parts stay louder than quiet parts, but the overall volume becomes predictable.

Think of it like this: imagine you have a photo that's too dark. You increase the brightness. The shadows are still darker than the highlights, but now the whole image is visible. Same concept.

But there are two completely different ways to measure "loudness," and that's where things get messy.

Peak Normalization vs Loudness Normalization

Peak normalization looks at the single loudest moment in your audio file (the highest waveform peak) and adjusts everything so that peak hits a specific level — usually 0 dB or -1 dB.

Sounds logical, right? Here's the problem: a track can have one brief spike (like a drum hit) but be mostly quiet. Peak normalization will make that spike touch 0 dB, but the rest of the track stays whisper-quiet. You haven't actually made the song sound louder.

Loudness normalization (measured in LUFS — Loudness Units relative to Full Scale) measures the perceived loudness of the entire track. It's based on how human ears actually hear audio, not just the waveform peaks.

This is what Spotify, YouTube, Apple Music, and basically every modern streaming platform use. And it's what you should use too.

Why Spotify Doesn't Sound Louder When You Crank It

Streaming platforms normalize everything to a target LUFS value. For Spotify, it's -14 LUFS. For YouTube, it's -14 LUFS. Apple Music uses -16 LUFS.

What happens if you upload a track that's louder than the target? The platform turns it down. So all that aggressive loudness mastering you did? Wasted. Worse — if you squashed your audio to make it louder, you've just sacrificed dynamic range for nothing.

Conversely, if your track is quieter than -14 LUFS, the platform will turn it up (but only to a point — they won't let it distort).

This is why the "loudness wars" are basically over. Mastering engineers used to crush audio to make tracks sound louder on the radio. Now? Pointless. Streaming platforms just turn everything down to match.

When You Actually Need Normalization

Here are the real-world scenarios where normalization matters:

• Making a playlist from different sources. Downloaded songs, ripped CDs, recorded voice memos — they're all over the map volume-wise. Normalize to -14 LUFS and they'll play nicely together.
• Podcast editing. Multiple guests on different mics? Some people talk quietly, others are loud? Normalization saves you from manually riding faders. Most podcasters target -16 to -19 LUFS.
• Audiobook production. Audiobooks need consistent volume across hours of content. ACX (Amazon's audiobook platform) requires -18 to -23 LUFS with peaks no higher than -3 dB.
• Video soundtracks. Background music, dialogue, sound effects — all need to sit at compatible levels. YouTube normalizes to -14 LUFS, so aim for that.
• DJ mixes and radio shows. Tracks from different decades and genres? Normalization keeps the energy consistent.

If you're working with audio from multiple sources or preparing files for distribution, you probably need normalization.

How to Actually Normalize Audio

Most audio editors have normalization built in. Audacity has both peak and loudness normalization (use the "Loudness Normalization" effect, not the "Normalize" effect — confusing naming, I know).

If you're working with MP3, WAV, or other common formats and need a quick solution, you can use KokoConvert's audio converter to adjust levels before exporting. It won't give you surgical LUFS control, but it'll get you in the ballpark.

For batch processing (normalizing a whole folder of files at once), ffmpeg is your friend. This command normalizes to -16 LUFS:

ffmpeg -i input.mp3 -af loudnorm=I=-16:TP=-1.5:LRA=11 output.mp3

You can loop that over a folder and process everything in one go. (If you don't want to mess with command-line tools, apps like Levelator and Auphonic do this automatically.)

Common Mistakes People Make

Using peak normalization for playlists. This is the #1 mistake. Peak normalization doesn't account for how loud something actually sounds. Use LUFS-based normalization instead.

Normalizing already-compressed audio. If you've already exported an MP3 at 128 kbps, normalizing it won't magically improve quality. Normalize during editing, before your final export. Better yet, work in a lossless format (WAV, FLAC) until the very end.

Setting the target too high. If you normalize to -8 LUFS for a Spotify upload, Spotify will just turn it down to -14 LUFS anyway. You're not gaining anything — and you might be clipping or distorting in the process.

Forgetting about true peak. LUFS measures average loudness, but you still need to make sure your peaks don't clip. Most normalization tools have a "true peak" setting (usually -1 dB to -2 dB). Use it.

Does Normalization Affect Audio Quality?

Not in the way you think. Normalization is just volume adjustment — it's math, not magic. Your waveform gets scaled up or down, but no audio data is removed or compressed.

However, if you normalize a quiet file way up, you might amplify background noise or hiss. And if you normalize too aggressively without watching your peaks, you could introduce clipping (distortion).

But done correctly, normalization is completely transparent. Your file size stays the same. The frequency content stays the same. It's just louder or quieter.

If you need to compress audio files for size reasons, that's a separate thing — you'd use lossy encoding (like MP3 or AAC), which does reduce quality to save space.

The Right LUFS Target for Your Use Case

Here's a quick reference guide:

• Spotify, YouTube, Apple Music: -14 LUFS
• Podcasts: -16 to -19 LUFS (some people go as high as -16, but -19 is safer for dialogue-heavy content)
• Audiobooks (ACX standard): -18 to -23 LUFS
• Broadcast TV (US): -24 LKFS (essentially the same as LUFS)
• Film and cinema: -27 LKFS for dialogue (but this varies by region)
• Instagram, TikTok: -14 LUFS (same as YouTube)

When in doubt, -14 LUFS is a safe bet for almost anything destined for the internet.

Tools That Make This Easy

If you don't want to dive into DAWs or command-line tools, here are some beginner-friendly options:

• Auphonic: Automatic normalization, noise reduction, and loudness balancing. Great for podcasts. Free tier available.
• Levelator: Free, dead-simple app. Drag in an audio file, it spits out a normalized version. Works surprisingly well for speech.
• Audacity: Free and open-source. The "Loudness Normalization" effect does LUFS-based normalization. Works on Windows, Mac, Linux.
• Adobe Audition: Professional-grade, includes loudness metering and batch processing. Paid subscription.

And if you just need to quickly merge audio files or convert formats before normalizing, browser-based tools like KokoConvert work well for one-off tasks.

Look, normalization isn't glamorous. But it's one of those foundational things that separates amateur audio work from professional audio work. If your podcast sounds like a volume roller coaster, or your playlist makes people reach for the volume knob every 3 minutes, you're losing listeners.

Take 10 minutes to normalize your audio properly. Your ears (and your audience's ears) will thank you.