Technique
From the engineer's desk

Mastering for Streaming: LUFS Targets, True Peak, and What Actually Matters

Every streaming platform processes your audio differently. Here's what the specs actually mean, why they matter, and how to master once and sound great everywhere.

7 min read Spotify, Apple Music, YouTube, Tidal & more Updated 2026

One of the most common questions I get from artists is some version of: "Do I need a different master for Spotify vs. Apple Music?" The short answer is usually no — but understanding why requires knowing what streaming platforms actually do to your audio when someone hits play.

Every major platform uses loudness normalization. That means if your master is louder than their target, they'll turn it down automatically. If it's quieter, they'll bring it up. This happens before the listener ever hears a single note. The goal of mastering for streaming isn't to get as loud as possible — it's to deliver a file that sounds exactly how you intend it to sound after the platform does its thing.

The key insight

Loudness normalization means a brickwall-limited, over-compressed master doesn't give you an advantage on streaming — it just sounds worse when the platform turns it down. Dynamic, well-mastered tracks often translate better at normalized volumes than crushed ones.

Specs by platform — click to explore

Here are the current loudness targets and true peak limits for every major streaming service. Click each platform to see the details and understand how they compare.

Integrated loudness
–14 LUFS
Target for normalization
True peak max
–1.0 dBTP
Hard ceiling before encoding
Quieter–14 LUFSLouder

Spotify is the platform most artists master toward by default — it's the most widely used and –14 LUFS is a comfortable target that preserves dynamics well. Tracks louder than –14 LUFS get turned down; quieter tracks get normalized up. They have loudness normalization on by default but users can disable it.

Integrated loudness
–16 LUFS
Sound Check normalization target
True peak max
–1.0 dBTP
Apple Digital Masters requirement
Quieter–16 LUFSLouder

Apple Music normalizes 2 LUFS quieter than Spotify via Sound Check. This is intentionally more conservative — Apple prioritizes dynamic range and offers the Apple Digital Masters program for high-res delivery. If your track is mastered to –14 LUFS, Apple will bring it down 2dB, which is generally fine. If you want absolute precision for Apple, a –16 LUFS alternate master is worth doing for key releases.

Integrated loudness
–13 LUFS
YouTube's normalization target
True peak max
–1.0 dBTP
Before AAC encoding
Quieter–13 LUFSLouder

YouTube Music targets –13 LUFS — slightly louder than Spotify. A master at –14 LUFS will be brought up 1 LUFS, which is barely perceptible. Worth noting that YouTube applies AAC encoding, which can add up to 1 dB of intersample peaks — another reason to keep true peaks at –1.0 dBTP or lower, not 0.

Integrated loudness
–14 LUFS
Normalization target
True peak max
–1.0 dBTP
Applies to all tiers including HiFi
Quieter–14 LUFSLouder

Tidal matches Spotify's –14 LUFS target. Notably, Tidal offers lossless HiFi and HiFi Plus tiers — so delivering a high-quality 24-bit master is particularly worthwhile here. The same spec applies regardless of tier.

Integrated loudness
–14 LUFS
Standard normalization target
True peak max
–1.0 dBTP
Consistent across Music Unlimited
Quieter–14 LUFSLouder

Amazon Music Unlimited uses the same –14 LUFS target as Spotify. Amazon also offers HD and Ultra HD streaming tiers, making a high-resolution master worthwhile. A Spotify-targeted master translates perfectly here.

Integrated loudness
–14 LUFS
Normalization target
True peak max
–1.0 dBTP
Before MP3/AAC encoding
Quieter–14 LUFSLouder

Deezer aligns with the –14 LUFS standard. They offer HiFi lossless streaming (FLAC) alongside standard tiers. Normalization behavior is consistent with other major platforms.

The practical takeaway

Four of the six major platforms target –14 LUFS. Master to –14 LUFS and –1.0 dBTP and you've covered Spotify, Amazon, Tidal, and Deezer in one pass. For Apple Music, your track gets turned down 2dB — which is fine for most releases. For YouTube, it gets nudged up 1dB — also fine. One well-made master covers the whole landscape.

File format — what to deliver to your distributor

The format you deliver matters even after mastering. Your distributor will encode your file into the compressed formats each platform uses (AAC, MP3, FLAC etc.), so starting from the highest quality source gives the encoding process the best material to work with.

Format
WAV or AIFF
Lossless. Never deliver an MP3 to your distributor — encoding MP3 to MP3 doubles the artifacts.
Bit depth
24-bit
Don't dither to 16-bit before delivering to a distributor. Most accept 24-bit and many prefer it.
Sample rate
44.1 or 48kHz
Match your session. Don't upsample — 96kHz doesn't help and some platforms downsample anyway.
True peak ceiling
–1.0 dBTP
Not –0.1 dBTP, not 0 dBFS. The –1.0 margin accounts for encoding adding intersample peaks.

How to actually measure this

LUFS isn't the same as dBFS (the level shown on most DAW meters). A standard peak meter won't tell you your integrated loudness — you need a dedicated LUFS meter. Here's what to use:

Metering tools worth knowing

The free Mix Analyzer shows you your track's LUFS, True Peak, dynamic range, and tonal balance — and lets you hear exactly how each streaming platform will play it.

Try the free Mix Analyzer to check your own track →

iZotope Insight 2 — my go-to in the studio. Loudness measurement, spectrum analysis, true peak metering, and platform-specific targets all in one window. Shows you exactly where you stand relative to each platform's spec.

Youlean Loudness Meter — free version is genuinely excellent for basic integrated LUFS and true peak measurement. A solid starting point if you're new to loudness metering.

NUGEN Audio VisLM — professional broadcast-grade metering. Worth it if you're delivering to film and TV as well as streaming.

Whatever tool you use, check two numbers: integrated LUFS (measured over the whole track, not a section) and true peak (the highest intersample peak in the file). Both matter, and both need to be right.

Integrated vs. short-term vs. momentary LUFS

Streaming platforms normalize based on integrated LUFS — the average loudness across the entire track. Short-term and momentary readings are useful for mixing but irrelevant for streaming compliance. Make sure your meter is set to display integrated loudness when you're checking your master against platform targets.

What goes wrong and why

Over-limiting to get louder before delivery
The platform turns it down anyway. All you've done is sacrifice dynamics and punch for a number that gets normalized away the moment someone presses play.
Setting the limiter ceiling to 0 dBFS instead of –1.0 dBTP
AAC and MP3 encoding adds intersample peaks of up to 1 dB. A master sitting at 0 dBFS true peak becomes a clipped master after encoding. The –1.0 dBTP ceiling exists for this exact reason.
Measuring peak LUFS instead of integrated LUFS
A loud intro can make your momentary or short-term reading look fine while your integrated measurement is actually several LUFS below target. Always measure integrated loudness over the full track.
Delivering MP3 to your distributor
Distributors encode your file into compressed formats for each platform. Starting from a lossy MP3 and encoding it again compounds the artifacts. Always deliver lossless WAV or AIFF.
Assuming specs never change
They do. Platforms adjust their normalization targets occasionally. For releases that matter, verify current specs via your distributor or directly from the platform before delivery.

Pre-release checklist

Run through this before sending your master to your distributor. Tick each one as you confirm it.

Streaming delivery checklist
0 of 6 confirmed
Integrated LUFS measured and on target
Measured over the full track using integrated mode, not momentary or short-term. –14 LUFS covers most platforms.
True peak at or below –1.0 dBTP
Not dBFS — dBTP (true peak). Confirmed with a true peak meter, not a standard peak meter.
Exported as 24-bit WAV or AIFF
Lossless format. Not MP3, not 16-bit unless your distributor specifically requires it.
Sample rate matches your session
44.1kHz or 48kHz — whichever your project ran at. No unnecessary conversion or upsampling.
Listened to the master in full — not just a section
Check the intro, the loud sections, the quiet sections, and the fade-out. Especially on earbuds and in a car, not just studio monitors.
Metadata is correct before upload
Song title, artist name, ISRC code if you have one. Errors in metadata are harder to fix after distribution.

Want a master that's already optimized for every platform?

This is what I do every day. Send me your mix and I'll deliver a streaming-ready master to Spotify, Apple Music, Tidal, and beyond — with the LUFS, true peak, and dynamics dialed in correctly. First-time clients get a free sample before committing.