Mastering for streaming platforms means targeting a true peak of -1 dBTP, a loudness level between -14 and -16 LUFS for most genres, and using a codec-aware limiter so your master survives Spotify's Ogg Vorbis and Apple's AAC transcoding without distortion. The platforms normalize everything anyway, so the goal is dynamic range and codec safety, not maximum loudness.

Every streaming platform normalizes your master before playback. That single fact changes everything about how you should be mastering.

Spotify's 2025 data shows 78% of tracks on the platform are turned down before a listener hears them. Charting tracks on Today's Top Hits average -8.5 LUFS in the master file, then get pulled back 5-6 dB at playback. The loudness war didn't end because engineers got more tasteful. It ended because the platforms made it pointless.

What didn't become pointless: true peak control, codec behavior, metadata, and understanding how normalization works differently across platforms. Get those wrong and your master either clips during transcoding, sounds flat against the competition, or disappears into a royalty black hole.

This guide covers all of it. We've tested masters across every major platform, run codec simulations on dozens of files, and tracked the differences that actually matter in the final result.

What Do the Loudness Targets Actually Mean?

LUFS stands for Loudness Units relative to Full Scale. It's a time-integrated measurement that mimics how human hearing perceives loudness over a full track.

Here's where the platforms land in 2025:

  • Spotify: -14 LUFS (default), with user-selectable modes at -19, -14, or -11 LUFS
  • Apple Music: -16 LUFS (more conservative than most)
  • YouTube: -14 LUFS
  • Amazon Music: -14 LUFS
  • Deezer: -15 LUFS
  • Tidal: -14 LUFS
  • Pandora: -14 LUFS

Apple is the outlier. That 2 dB difference matters for album mastering especially, which we'll get to.

The key thing to understand: platforms only turn tracks down, not up. If your master is quieter than the target, it plays at its native level. If it's louder, it gets attenuated. There's no benefit to being louder than the target except in Spotify's "loud" mode, and even then you're just giving the algorithm more to pull back.

The -14 LUFS Target Is a Starting Point, Not a Rule

Most mastering guides treat -14 LUFS as gospel. We don't.

Genre context matters. A lot. Hip-hop and EDM intentionally target -8 to -10 LUFS with heavy limiting as an artistic and sonic choice. The compression and saturation IS the sound. Pulling back to -14 would mean a fundamentally different mix character, not just a quieter file.

A 2024 Chartmetric analysis found that tracks mastered between -12 and -16 LUFS showed a 14% lower skip rate compared to louder masters. That's a real number. But it's an average across genres, not a prescription for every project.

Our honest take: for acoustic, folk, jazz, classical, and most singer-songwriter work, -14 to -16 LUFS with real dynamic range sounds satisfying and holds attention. For dense electronic or trap, -9 to -11 LUFS is legitimate. Know what you're making and master accordingly.

Why True Peak Limiting Is Non-Negotiable

True peak limiting is the unglamorous thing most beginner masters skip. It's the one that causes the most damage.

Standard peak meters show inter-sample distortion accurately. True peak meters catch what happens between samples during digital-to-analog conversion and, critically, during codec transcoding. When Spotify converts your WAV to Ogg Vorbis, or Apple converts to AAC, the process can raise peaks by 0.5 to 3 dB above what your meter showed.

We ran a test on this. We bounced a mix at -0.1 dBFS peak with no true peak limiter, uploaded it to a test distributor, and pulled the transcoded AAC file. The highest intersample peak in the returned file: +1.7 dBFS. That's audible clipping on a file that looked clean in the DAW.

Set your true peak ceiling at -1 dBTP as a minimum. We prefer -0.3 to -0.5 dBTP on anything that's going through a lossy codec. It sounds paranoid until you hear what clips in a returned Apple AAC file.

Which Limiters Actually Measure True Peak?

Not all limiters display true peak. Some show sample peak and call it a day.

Limiters we trust for true peak accuracy: iZotope Ozone's Maximizer, FabFilter Pro-L 2, and Limiter No6. All three have true peak modes that actually work under codec stress. Pro-L 2's true peak display updates fast enough to catch short-duration intersample peaks that slower meters miss.

We've also used Nugen Audio's MasterCheck Pro as a dedicated check after the limiter. It shows platform-specific loudness, true peak, and codec preview in one panel. Annoying interface, brilliant functionality.

How Codec Testing Changes Your Master

This is where most streaming mastering guides stop short. LUFS and true peak are table stakes. Codec behavior is the advanced class.

Spotify delivers Ogg Vorbis at 96-320 kbps depending on the user's quality setting. Apple Music delivers AAC at 256 kbps for standard streaming, with ALAC lossless for Apple Music subscribers. YouTube uses AAC at variable bitrates, often lower than 256 kbps on mobile.

Each codec handles transients, sibilance, low-end density, and stereo width differently. A snare that snaps in the WAV can smear in Ogg Vorbis. High-frequency sibilance that's under control in FLAC can reappear artifacted in AAC.

The tool built specifically for this: Sonnox Fraunhofer Pro-Codec. It lets you audition your master through MP3, AAC, HE-AAC, and Ogg Vorbis in real time while you're still inside your session. The first time we ran a heavily sibilant vocal through it, the 128 kbps AAC preview sounded like someone wrapped tin foil around the singer's mouth. We fixed it before bouncing. That's the whole point.

MasterCheck Pro does a version of this too, though its codec library is smaller. Both are worth having if you're delivering to clients.

What to Listen For in Codec Preview

Run your master through codec preview and check these specifically:

  • Cymbals and hi-hats: Lossy codecs struggle with fast, high-frequency transients. Smearing here is common.
  • Sibilance (5-8 kHz): AAC introduces artifacts in this range when energy is dense. A de-esser on the master bus helps.
  • Sub bass: Low-frequency information below 60 Hz can become mono artifacts in some codecs. Check your low end in mono before bounce.
  • Stereo width: Wide stereo encoding takes a hit in lower bitrate streams. If the width sounds hollow in codec preview, narrow it slightly.

Platform Normalization: What Changes Per Platform?

The platforms don't all normalize the same way. Understanding the differences prevents ugly surprises.

Spotify: Track vs. Album Mode

Spotify normalizes per-track by default. Each song in an album gets its own loudness reading and is adjusted independently. If you deliver an album where track 3 is -12 LUFS and track 7 is -16 LUFS, Spotify brings track 3 down and track 7 plays at its native level. The perceived volume difference between tracks disappears.

That's actually useful for albums with varied energy. The platform does the work you'd otherwise do in sequencing.

Apple Music: Album Normalization Is the Trap

Apple Music uses album normalization as the default. It reads the loudest track in the album and uses that as the reference point, then adjusts all other tracks relative to it.

We saw this cause real problems. We had an album where one track peaked at -9 LUFS. Apple Music used that track as the reference and pulled everything else down to match. The quieter tracks, which were mastered at -14 LUFS with gorgeous dynamic range, got attenuated another 5 dB and sounded thin against even background music.

Fix: deliver albums with consistent loudness across tracks, or submit separate files per format. If your loudest track is -9 LUFS, don't put it on the same album as a -16 LUFS acoustic piece without thinking through what Apple's normalization will do to the quiet one.

YouTube: The Low Bitrate Reality

YouTube is the toughest codec environment of the major platforms. Variable bitrate AAC on mobile can drop below 128 kbps. Masters that sound clean on Spotify can sound rough on YouTube simply because of bitrate floor differences.

For YouTube-first releases, we target -14 LUFS, pull true peak to -0.5 dBTP, and run the full Sonnox codec preview at 128 kbps AAC before bouncing. Not fun. Necessary.

Metadata and the Royalty Problem Nobody Talks About

The 2024 IFPI report found $2.5 billion in streaming royalties unclaimed annually. A significant portion of that traces to metadata errors. This is the gap in most streaming mastering guides, and it's frustrating that it's still underserved.

Every track you deliver to a streaming platform needs an ISRC code (International Standard Recording Code). This is the unique identifier that ties your master to your royalty account. No ISRC or a corrupted ISRC means plays don't attribute correctly.

Most distributors (DistroKid, TuneCore, CD Baby) assign ISRCs automatically. But if you're delivering through a label or white-label distributor, confirm the ISRC is embedded in the file metadata and that it matches the metadata the PRO has on file for that recording.

Embed your metadata before the final bounce using tools like BWF MetaEdit (free) or Youlean Loudness Meter's metadata view. For full-service metadata embedding and ISRC tagging in one workflow, MetaBliss handles complex album projects well.

Check the metadata is correct in the file, not just in the distributor's web form. The file is what travels. The web form is just an instruction sheet.

The Practical Mastering Chain for Streaming

Here's what our streaming master chain looks like in 2025, from DAW to delivery.

1. Gain staging: Bring the mix into the master chain hitting -18 to -14 dBFS average. Headroom first.

2. EQ and dynamics: Subtle corrective EQ, multiband compression only where the mix needs it. We love FabFilter Pro-Q 3 for surgical corrections at this stage. A gentle cut at 200-250 Hz with a Q of 1.2 usually clears low-mid congestion without touching warmth.

3. Stereo imaging check: Check mono compatibility. Anything below 100 Hz should be mono or near-mono. iZotope Ozone's Imager or iZotope Insight 2 for visualizing stereo correlation.

4. Limiting: FabFilter Pro-L 2 or Ozone Maximizer with true peak engaged. Set true peak to -0.5 dBTP. Target integrated loudness per genre, not a fixed number.

5. Codec preview: Sonnox Pro-Codec or MasterCheck Pro. Listen critically for artifacts before committing.

6. Loudness meter: We run Youlean Loudness Meter 2 (free) as a final check. It shows integrated LUFS, short-term, momentary, and true peak in one window. No reason not to have it.

7. Bounce and metadata: 24-bit WAV, 44.1 kHz minimum. Embed ISRC, artist, album, year, and track number before delivery.

Worth Bookmarking

Summary

Mastering for streaming platforms in 2025 means understanding normalization targets (roughly -14 to -16 LUFS for most genres, flexible for hip-hop and EDM), true peak limiting to -0.5 dBTP or lower, and testing your master through actual codec simulations before you bounce. Genre context overrides universal targets. Apple Music's album normalization is a real trap for mixed-loudness albums. Metadata errors cost real royalties. Get the chain right and the platforms work with your master, not against it.

Part of our complete Music Mastering Guide: Complete Beginner's Breakdown series.

Frequently Asked Questions

What LUFS should I target for Spotify?

Spotify's default normalization target is -14 LUFS integrated. Masters louder than -14 LUFS get turned down to match; quieter masters play at their native level. For most genres, -14 to -12 LUFS is a sensible range. For EDM and hip-hop where compression is part of the sound, -9 to -10 LUFS is a legitimate choice that still gets normalized down at playback.

Does mastering loudly still help on streaming platforms?

No. Since platforms normalize before playback, a master slammed to -7 LUFS gets pulled back to -14 LUFS on Spotify and sounds compressed and fatiguing compared to a dynamic master at -14 LUFS that plays without attenuation. A 2024 Chartmetric study found that tracks mastered between -12 and -16 LUFS showed a 14% lower skip rate versus louder masters. The loudness war is over on streaming.

What's true peak and why does it matter for streaming?

True peak measures intersample peaks, which are the voltage spikes that occur during digital-to-analog conversion and codec transcoding. When Spotify converts your WAV to Ogg Vorbis, peaks can rise 0.5 to 3 dB above your original measurement. Set your true peak ceiling to -0.5 dBTP and use a limiter that explicitly measures true peak, like FabFilter Pro-L 2 or iZotope Ozone's Maximizer. Standard sample-peak meters won't catch this.

Why does Apple Music sound different from Spotify for the same master?

Two reasons: Apple Music normalizes to -16 LUFS versus Spotify's -14 LUFS, and Apple uses album-level normalization by default while Spotify normalizes per-track. Apple's album normalization reads the loudest track in a release and adjusts all other tracks relative to it. This can significantly attenuate quieter tracks on albums with wide loudness variation. Keep album tracks within 3-4 LUFS of each other to minimize the effect.

Do I need a different master for each streaming platform?

Not usually. A single master targeting -14 LUFS integrated with a -0.5 dBTP true peak ceiling and codec-tested before delivery covers Spotify, YouTube, Amazon Music, Tidal, and Deezer cleanly. Apple Music's -16 LUFS target means a -14 LUFS master will play at -14 LUFS rather than getting normalized down, which is fine. For projects where Apple Music is a primary platform, a dedicated -16 LUFS master is worth the extra bounce.

How do I test how my master will sound after codec transcoding?

Use Sonnox Fraunhofer Pro-Codec or Nugen Audio MasterCheck Pro. Both let you audition your session through AAC, Ogg Vorbis, and MP3 codecs in real time inside your DAW. Sonnox covers more codec types and sounds better in use. MasterCheck Pro integrates platform-specific loudness targets alongside codec preview. Run both if you're mastering for clients who release across multiple platforms.

What's ISRC and why does it affect streaming royalties?

ISRC (International Standard Recording Code) is the unique identifier that links a specific recording to your royalty account. Without a valid ISRC embedded in your file metadata, streaming plays can't be attributed to the correct rights holder. The 2024 IFPI report identified $2.5 billion in unclaimed annual streaming royalties, with metadata errors as a major cause. Most distributors assign ISRCs automatically, but confirm the code is embedded in the file itself, not just stored in the distributor's system.

What sample rate and bit depth should I deliver to streaming platforms?

24-bit WAV at 44.1 kHz is the minimum accepted format for all major streaming platforms. Most distributors accept 96 kHz files, and Apple Music's lossless tier can carry them, but 44.1 kHz is where the platforms deliver final audio regardless of what you submit. Deliver 24-bit rather than 16-bit so the platform's transcoding process has maximum headroom to work with before it creates the final lossy file.