How Loud Should a Rap Master Be? LUFS Targets for Rap

Master rap between −10 and −8 LUFS integrated, with a true-peak ceiling of −1.0 dBTP. Dense, 808-heavy trap can sit as hot as −7 without losing much. Sparse beats and dynamic vocals hold together better at −12 to −10. Streaming normalizes every file to its own target anyway, so loudness buys you density and attitude, not volume.

Disclosure: we build audio plugins, and one of them is a mastering plugin called MONOLITH. It appears in one labeled section near the end, with our own probe numbers so you can check them against a meter you trust. No affiliate links on this page, and every external link points at the source document.

What LUFS target should a rap master hit?

Loudness is not a setting you dial in. It falls out of the arrangement, the mix bus, and how hard you push the last limiter. Treat the numbers below as starting points that tell you when to stop, not a score to chase.

One column in that table matters more than the loudness targets: the true peak. Loudness is a taste call. A ceiling above −1 dBTP is a delivery fault.

What does loudness normalization do to your master?

The major streaming services measure your track, then apply a fixed gain at playback so it arrives at about the same loudness as the track before it. Most listeners never hear your file at the level you printed. That one fact resolves most of the argument.

The measurement they use is defined by ITU-R Recommendation BS.1770, the same algorithm behind the broadcast standard EBU R 128. Broadcast settled on −23 LUFS. Music streaming landed a lot higher.

Only some of those figures are published. Spotify lists its three playback levels on its loudness normalization page, and Apple publishes delivery requirements under Apple Digital Masters. Apple's Sound Check figure is not documented, and neither are YouTube's or Tidal's, so treat every "about" above as a measured estimate rather than a published number, and expect it to shift.

Two consequences follow. Mastering to exactly −14 LUFS does not make you louder on Spotify; it means no gain gets applied, which is not an advantage. And going very quiet hands control to the platform: Spotify's documentation notes that in the Loud setting, tracks below the target get gain plus a limiter. Their limiter, their settings, on your record.

To hear the turn-down before you commit, drop a bounce into Loudness Penalty, which plays your file at each platform's level.

Why do rap masters run hotter than the −14 LUFS folklore?

The −14 number comes from Spotify's own loudness page, where it is both the playback target and a suggested master level. Forums turned the suggestion into law. Rap sits above it for four reasons, none of them about winning a volume contest.

The density is already there. A trap mix arrives at the master with a saturated 808 holding the low end, a compressed and de-essed vocal on top, and hats filling the space between. Very little peak-to-loudness headroom is left to give up, so the limiter that would gut an acoustic trio barely moves a beat like that.

The 808 is a sustained tone, and it is the loudest thing in the song. Sustained low frequency is what pushes an integrated reading up. Get the sub relationship right in the mix and the loudness follows on its own. Get it wrong and every dB you take is paid for with pumping, which is a mix problem to fix upstream in the chain, not at the limiter.

Playback is phones, earbuds and cars. A phone speaker does not reproduce a 40 Hz fundamental and earbuds barely do, so most of what a listener hears of your 808 is its harmonics. Saturation, not level, is what makes it survive. A loud master does not translate better on its own: normalization applies on phones too.

Density survives normalization. Playback gain equalizes loudness. It cannot equalize how packed a record is. A rap record that reads as finished tends to be dense, and dense material measures loud. The loudness is the symptom, not the goal.

Keep loudness and dynamics separate. A −13 LUFS master of a flattened mix is quiet and lifeless; a −8 LUFS master of a balanced mix can still swing. Mastering quieter is not a shortcut to dynamics.

What true-peak ceiling should a rap master use?

Set the ceiling to −1.0 dBTP on an oversampled true-peak meter. Drop to −2.0 dBTP if you are mastering hot, or if you can hear the codec on an encoded bounce. That mirrors Spotify's published guidance: −1 dBTP normally, −2 dBTP if your master is louder than −14 LUFS.

The reason is inter-sample peaks. Your DAW's meter shows sample peaks, the values stored in the file, but the waveform reconstructed between those samples can be higher. A file reading −0.1 dBFS can exceed 0 dBTP once a converter or a lossy encoder gets to it, and the distortion lands on the most fragile material: hats, esses, snare tops. BS.1770-4 Annex 2 handles this by oversampling before it measures. A meter that does not is telling you a comfortable half-truth.

So bounce at your session sample rate, let the distributor convert, and check an encoded copy instead of assuming. Distortion you can hear on an AAC bounce is distortion the listener gets.

How do you measure loudness on your own master?

Put a loudness meter last on the master bus, after the limiter, and play the entire song from silence to silence. Not a loop, not the hook. The integrated reading is a whole-program measurement, and metering 30 seconds of the loudest section is the most common way people end up quoting a number that has nothing to do with their file.

What the algorithm does: BS.1770 applies K-weighting, a two-stage filter that approximates how we perceive energy across the spectrum, averages power over the program, then gates. An absolute gate at −70 LUFS discards silence, and a relative gate 10 LU below the ungated level discards quiet material, so a long quiet intro or a spoken outro barely moves your average. Meter behavior is specified in EBU Tech 3341, which is why two compliant meters agree.

LUFS and LKFS are the same unit under two names, and LU is the relative version, so "3 LU louder" behaves like "3 dB louder". Logic Pro ships a Loudness Meter, Reaper measures loudness in its render dialog, and Youlean Loudness Meter has a free version that covers everything here. Bob Katz's writing at Digital Domain is the clearest background on how level practices got this way.

From Our Own Line: What We Measured Building MONOLITH

MONOLITH is our whole-song mastering plugin, and the measurement side had to hold up before anything else in it mattered. The numbers below come from its own probe suite. They are ours, not third-party certification.

The working ceiling sits 0.3 dB under whatever you dial, as an inter-sample pad, which is why the worst case above landed under the ceiling instead of over it. The limiter runs 1.5 ms of lookahead with a sliding-minimum design and a min-clamp, so it cannot leak overs, and its release moves between 0.6× and 3× the base time.

The lesson from building it transfers to your own sessions. Our first version measured the chain at unity gain and then applied the trim, and on hot material it overshot the target by roughly 1.5 dB. Glue compression and limiting are level-dependent, so a measurement taken at the wrong operating level predicts the wrong outcome. It now trims coarsely partway through the capture and measures near the final level. If you are matching loudness by hand, do the same: measure with your limiter working, not before it.

Two of its defaults are genre calls that affect level directly: a 24 Hz subsonic filter is on, and everything below 120 Hz is summed to mono, because low content spread wide across the stereo field loses level the moment a system sums it.

It is not a neural network. It measures, then applies one static gain trim plus one conservative tonal move, offline, with nothing uploaded. We broke that distinction down against the services in AI mastering vs human mastering.

How MONOLITH hits the target

The workflow follows from the measurement problem above. Integrated loudness is a whole-song number, so the plugin will not solve from a few bars.

Press CAPTURE, then play the track end to end. An overlay takes over the window while it listens, with an elapsed timer running. Stop before four seconds and it returns CAPTURE TOO SHORT and changes nothing. That guard matters: a chorus-only pass reads hotter than the full song, so a master solved from it lands too quiet.

Press STOP and it solves once. The readout switches to MASTER LOCKED and reports the input LUFS, the predicted output, the single static gain trim it applied, and confirmation that the gain is frozen. One trim across the whole song is why the dynamics survive, because nothing is riding the level bar by bar.

Pick the target that matches the release. The TARGET LUFS row in the plugin window at the top of this section is this article's argument rendered as six buttons, running from -14 to -7 and defaulting to -10. CEILING sets the true-peak limit alongside it, defaulting to -1 dBTP.

RESTRAINT is the control worth knowing about. It scales how much the plugin does against how much the track needs. Feed it a master that is already finished and the needs assessment reads 1.00, roughly 5 percent of the processing gets applied, and corrective EQ sits at 0.00 dB. It is built to do nothing when nothing is required.

MONOLITH is $29, VST3 on macOS and Windows plus AU on macOS, keyless, so it installs on every computer you own with no license server. No fake sales. No intro-price games. The price is the price. The rest of the line is on the plugins page.

Why does my master sound loud but not hit hard?

Because loudness and impact are different measurements. The one that tells you whether a master still hits is not LUFS, it is the gap between your peaks and your loudness. Ian Shepherd, who writes at Production Advice, has spent years arguing that the number to watch is PSR: peak minus short-term loudness, moment to moment.

No threshold to memorize. Just watch the number while you push. Two signs you have gone too far:

1. PSR stops moving. When the verse and the hook return the same value, the limiter is setting your dynamics instead of your arrangement.
2. The 808 loses its front edge. The first audible casualty on a rap master. The note is still there, the thump that started it is gone, and the beat feels like it sits behind the vocal.

Other symptoms, in the order I tend to notice them: sibilance climbing as the limiter pushes the top end forward, the snare or clap shrinking while everything around it grows, the vocal sliding backward the louder the master gets. Hear any of those and the fix is 1 dB off the limiter and more room in the mix.

The five-minute loudness check before you upload

1. Integrated LUFS, whole song. Silence to silence, meter after the limiter, checked against the first table above.
2. True peak under −1.0 dBTP. On an oversampled meter, not your DAW's sample-peak meter.
3. Mono check. Sum to mono; the low end should not vanish and the vocal should not drop.
4. Level-matched reference. Match a track from the same lane to your loudness before comparing. Comparing at different levels tells you nothing.
5. Phone and earbuds. Not for tone. For whether the 808 reads as a note and the words stay intelligible.
6. Project loudness. On an EP, master the tracks relative to each other so one does not jump 4 LU front to back.

Anything that check turns up is usually a mix problem wearing a mastering costume. The full walkthrough, in order, is in how to master a rap song at home.

Frequently Asked Questions

Is −14 LUFS the right target for a rap master?

Not as a rule. −14 LUFS is Spotify's playback target, and Spotify suggests it as a master level too, but mastering to −14 does not make your track louder there; it only means no gain gets applied. Most rap sits between −10 and −8 LUFS integrated because the arrangement is already dense. Let the mix decide.

Should I master at −8 LUFS or −14 LUFS?

For rap, −8 is the more useful print. Spotify turns a −8 LUFS master down about 6 dB and leaves the density, low-end weight and saturation you mastered in. A −14 master gets no gain applied, which is not an advantage. Keep the ceiling at −1 dBTP either way.

What LUFS is too loud for rap?

Past roughly −6 LUFS integrated you are usually trading the groove for a number. Transients flatten, the 808 loses its attack, and lossy encoding starts adding distortion you did not choose. Streaming turns the file down anyway, so the extra limiting buys nothing a listener can hear.

Should my true peak be −1 dBTP or −0.1 dBFS?

Use −1 dBTP on an oversampled true-peak meter. Sample-peak dBFS misses inter-sample peaks, so a file reading −0.1 dBFS can still clip a converter or a lossy encoder. Spotify's guidance is −1 dBTP, and −2 dBTP if you master louder than −14 LUFS.

Will a louder master sound louder on Spotify?

Not louder, only different. Normalization applies gain so every track arrives at the same loudness, which removes the volume advantage. What survives the turn-down is tone, density and how your low end is managed. Two masters at the same LUFS can feel completely different.

How do I measure LUFS for free?

Put a loudness meter last on the master bus, after the limiter, and play the whole song from silence to silence. Logic Pro ships a Loudness Meter, Reaper measures loudness in its render dialog, and Youlean Loudness Meter has a free version that works in any DAW.