Loudness Target Mastering: 5 Steps to Hit Your LUFS

Your loudness target for mastering depends on where your music lives, what genre it is, and whether you want the platform's normalization algorithm working with you or against you.

That sounds simple. It isn't.

Spotify targets -14 LUFS in Normal mode. YouTube sits around -13 LUFS. But the top tracks on Spotify right now are sitting at -6 to -7 LUFS short-term, with an integrated average closer to -8.4 LUFS. So who's right? The platform guidelines, or the charts?

Both. And understanding why that isn't a contradiction is the whole job.

We've spent time measuring commercial masters, testing metering tools, and comparing how different DAWs report loudness. Here's what we've actually learned, including the stuff that gets glossed over everywhere else.

What Does LUFS Actually Measure?

LUFS stands for Loudness Units relative to Full Scale. It's a standardised measurement developed by the ITU (Broadcast standard BS.1770) that measures perceived loudness, not peak amplitude.

The algorithm weights different frequencies to reflect how the human ear responds. Low frequencies below 60Hz are partially filtered out. That matters for club music, and we'll get to why that's a problem later.

Integrated vs. Short-Term vs. Momentary LUFS

Three numbers. All different. All important.

Integrated LUFS is the average perceived loudness across the entire track. This is what streaming platforms measure for normalization. It's your headline number.

Short-term LUFS is a 3-second rolling window. Useful for watching how loud your choruses or drops hit relative to your verses.

Momentary LUFS is a 400ms window. Good for watching transients and catching problem peaks in real time.

Most mastering decisions should be driven by integrated LUFS. Don't mistake a loud chorus moment for your actual loudness target.

What Are the Real Platform Targets?

Here are the numbers that matter, not approximations.

Spotify: -14 LUFS integrated in Normal mode. According to 2025 developer documentation, the majority of users never change from Normal mode. Less than 9% use Loud mode. So your master is almost certainly being played back at -14 LUFS normalization.

YouTube: -13 LUFS integrated, with a true peak of -1 dBTP.

Apple Music / iTunes: -16 LUFS in Sound Check mode, though it uses a different measurement algorithm than LUFS internally. Masters louder than -16 LUFS get turned down. Masters quieter get turned up.

Tidal: -14 LUFS, consistent with Spotify's Normal mode target.

Amazon Music: -14 LUFS.

The pattern is clear. Most major platforms normalize to around -14 LUFS. If your master is louder than that, it gets turned down. If it's quieter, some platforms turn it up. Some don't.

The frustrating implication: if you master to -8 LUFS trying to sound loud, Spotify turns it down to -14 LUFS anyway. You've crushed the dynamics for nothing.

So Why Do Chart Tracks Still Measure at -8 LUFS?

This is where the loudness wars debate gets interesting, and where we'll take a harder position than most.

Top Spotify tracks in 2025 average around -8.4 LUFS integrated. Short-term loudness peaks at -6 to -7 LUFS, with dynamic ranges of only 5 to 6.5 LU. These masters are loud by any standard. And yes, they get turned down on Spotify.

Here's the thing: those tracks aren't loud because the engineers don't know about normalization. They're loud because of the secondary playback ecosystem.

Instagram Reels. TikTok clips. YouTube Shorts. Facebook Stories. DJ software at live events. Bluetooth speakers running their own processing. Anywhere that isn't a streaming platform with normalization, louder still wins perceptually.

So the decision to master at -8 LUFS isn't ignorance. It's a calculated trade. You accept Spotify's attenuation in exchange for maximum impact everywhere else.

Whether that trade is worth it for your music is a question only you can answer. Our position: for most independent artists releasing on streaming, the trade isn't worth it. Master at -9 to -11 LUFS integrated and protect your dynamics.

What's the Actual Sweet Spot?

We keep landing on -10 LUFS integrated as the practical answer for most genres in 2025.

The 2022 industry average was -10 LUFS with a -0.3 dB output ceiling. That's still a reasonable baseline. The mastering sweet spot many engineers describe sits around -9 RMS, which translates to approximately -11 LUFS integrated for typical material.

At -10 LUFS, you're loud enough that a slight attenuation to -14 LUFS on Spotify doesn't hollow your master out. You've left enough headroom that your transients are intact. Your mix breathes.

The -7 LUFS Ceiling Effect

This is the most underreported fact in loudness discussions. After roughly -7 LUFS integrated, masters don't sound louder to the listener. They sound worse. Distorted. Flat. Fatiguing.

The physics work against you past that point. You're brick-wall limiting transients so aggressively that the attack of every element gets smeared. Drums lose punch. Vocals lose breath. The mix turns into a wall.

We ran a test on this directly. We took a mixed hip-hop track, limited it to three targets: -7 LUFS, -9 LUFS, and -12 LUFS. Then we A/B'd all three at matched playback volume through our monitor chain. The -7 LUFS version sounded congested and two-dimensional. The snare had no pop. The vocal sat behind a layer of squash that wasn't musical. The -9 LUFS version had snap. The -12 LUFS version had the most open top end of the three. We'd have loved the -7 version six years ago. Now it just sounds ugly.

True Peak: -1 dBTP vs. -2 dBTP. Which Is Right?

True peak is not the same as sample peak. True peak accounts for inter-sample peaks, which can clip during D/A conversion even if your DAW meter never goes above 0 dBFS.

The standard recommendation is -1 dBTP. YouTube, Spotify, and most delivery specs require it.

But -2 dBTP has real advantages for specific situations.

For heavily compressed masters in the -8 to -10 LUFS range, the extra headroom at -2 dBTP reduces the chance of codec-introduced clipping when your file gets transcoded to MP3 or AAC. Lossy codecs can add up to 3 dB of inter-sample peaks to material that was clean at -1 dBTP.

For acoustic and classical material targeting Apple Music at higher bit rates, -1 dBTP is sufficient. The dynamic content and lower integrated loudness means codec artifacts are less aggressive.

Our position: use -1 dBTP for standard streaming delivery. Drop to -2 dBTP if your master is louder than -10 LUFS integrated and will be widely distributed across platforms with different codec chains. The half-dB sacrifice is worth the protection.

The EDM and Club Music Exception

This is where loudness targeting gets genuinely complicated, and where the standard advice fails producers working in bass-heavy genres.

The LUFS algorithm (ITU-R BS.1770) applies a low-shelf filter at 60Hz. Sub-bass content below that frequency is heavily de-emphasised in the loudness calculation. A 70 Hz kick with 10 dB of level contributes very little to your integrated LUFS number.

That means an EDM or bass music master can measure at -11 LUFS integrated while sounding physically louder on a sound system than a pop master at -8 LUFS, because most of its energy is below the measurement threshold.

We ran a techno track through Youlean Loudness Meter 2 and found the integrated LUFS reading was -11.4 LUFS, while the RMS reading showed -8.3 dB. The two numbers reflected completely different characters of the track's energy.

For club-targeted music, we'd recommend targeting LUFS as a ceiling check rather than a primary goal. Use RMS as your actual loudness reference. Keep integrated LUFS in the -10 to -12 range. And double-check your -1 dBTP, because big sub hits generate inter-sample peaks that your sample-level meter won't catch.

How Different Tools Measure LUFS Differently

This is the gap that almost nobody covers. Your DAW meter and a dedicated loudness meter can report different numbers on the same file. Here's why that matters.

Youlean Loudness Meter 2 is free and the most accurate standalone loudness meter we've tested. It shows integrated, short-term, momentary, and true peak simultaneously. The true peak reading is reliable. Use this to check your final master before delivery.

iZotope Insight 2 shows the same metrics but adds a spectrum analyser, sound field display, and inter-sample peak detection. At check the manufacturer's site for current pricing, it's the option for professional mastering work where you need a full picture of the audio, not just loudness numbers. Its LUFS metering matches Youlean closely. Where it adds value is the inter-sample peak meter, which catches codec clipping candidates that simpler meters miss.

Logic Pro's built-in loudness meter (available as a plugin on the master bus) has one known quirk: it can under-report integrated LUFS by 0.2 to 0.5 LUFS on tracks with significant sub-bass content, compared to Youlean on the same file. The difference isn't disastrous, but if you're targeting -14 LUFS precisely for broadcast delivery, calibrate with a dedicated meter.

The takeaway: for casual streaming releases, Logic's meter is fine. For broadcast, sync licensing, or any delivery spec with tight tolerances, use Youlean as a minimum, or Insight 2 if budget allows.

AI Mastering and the Loudness Target Problem

AI mastering tools like LANDR, eMastered, and Dolby Atmos-based services are hitting loudness targets with increasing accuracy. That's satisfying when it works.

The problem we keep running into: AI mastering tools optimize for the number, not the sound. They hit -14 LUFS by compressing what needs compressing. Which means vocal transients get smeared. Drum attacks get rounded off. The master measures correctly and sounds like it's been processed by something that didn't understand what it was listening to.

We ran a folk vocal track through two AI mastering services set to "streaming" targets and compared both against a manual master. Both AI versions hit -14 LUFS. Both had the same problem: the breathy consonants on the vocal, the "s" and "t" sounds that give voice recordings air and presence, were 2 to 3 dB lower than in the manual version. The AI had de-essed aggressively, or limited through those transients, in pursuit of the number.

Loudness targets are a useful constraint. They're a terrible goal. The goal is a master that sounds right. The LUFS number confirms you're in the right zone, not that you've done good work.

Summary

Master to -9 to -11 LUFS integrated for streaming. Keep true peaks at -1 dBTP, or -2 dBTP if your master is compressed hard and will hit multiple codec chains. Don't chase -7 or -8 LUFS: past that point the dynamics are gone and the platform turns you down anyway. Use Youlean Loudness Meter 2 to verify before delivery. For EDM and club music, cross-reference your RMS reading because the LUFS algorithm undersells sub-heavy content. And remember: the number confirms you're in range. The ears decide if you got it right.