How loud shoulda club master be?

Ask a forum how loud your master should be and you’ll get the same answer every time: -14 LUFS, because that’s where Spotify normalises. It sounds authoritative and it keeps getting repeated, so producers dutifully master their club tracks to -14 and then wonder why they sound polite next to everything else in a set.

I got tired of the argument, so I measured it. We keep a calibration corpus at Echo Forte: commercial club releases paired with their actual premasters, the real before-and-after from records that came through our studio work. Six pairs, measured with the same BS.1770 tooling we use to verify every master the engine ships.

The median commercial club master in our corpus sits at -7.7 LUFS integrated. Loudness range around 4 LU. Crest factor around 11 dB. And here’s the one that surprises people: true peaks at +1.2 dBTP on the median master. Over full scale. Professional club masters routinely let intersample peaks run past zero, because the engineers mastering them decided level mattered more than a meter nobody in the room can hear.

The delta between premaster and master is the more useful number. Across the corpus, mastering added about 5.2 LU of loudness, spent about 3.6 dB of crest to get it, and moved LRA by less than a tenth of an LU. Read that again: the macro dynamics, the travel from verse to drop that makes a track feel alive, barely moved. Good mastering buys its level from transient tops and density, and leaves the arrangement’s dynamics alone.

True, and mostly beside the point for club music. Normalisation applies on streaming platforms with it switched on. It does not apply in a DJ booth. CDJs don’t normalise. Rekordbox doesn’t normalise your gain for you in any way that saves a quiet master. When your track sits in a set between two records mastered at -7.5, a -14 LUFS master needs the DJ to ride the trim six or seven dB for you, and mostly they won’t, they’ll just skip the track that sounds small.

Streaming turndown is also cheap insurance compared to the alternative. A loud master turned down by Spotify still keeps its density and its tone. A quiet master turned up by nobody stays quiet everywhere it was supposed to hit.

Here’s the honest part. When we calibrated our Club preset against those reference masters, the uncomfortable finding was that the real studios push harder than sounds “correct” in a quiet room. Solo the loud master on studio monitors next to a gentler one and you can argue the gentler one is the better record. In a set, on a big rig, at 2am, the loud one wins, and the engineers mastering chart club records know it. They push right to the limit even when it costs the track something, because that’s what the job is.

That’s why the engine ships two club settings instead of one diplomatic compromise. Club targets -8.5 LUFS with a -0.5 dBTP ceiling and keeps more of the punch. Club Hot targets -7.5 with a -0.1 ceiling and trades dynamics for density where the track can carry it, the way a real club master does. You preview both level-matched, so you’re choosing on character instead of being tricked by volume, then you pick the trade you actually want.

Give whoever masters it, us or anyone, room to work. Peaks around -6 dBFS, no limiter on your mix bus, and resist the urge to pre-loudness the mix; the 5 LU comes from the mastering chain, and it comes out cleaner from an open mix than a slammed one.

Then pick your target by destination, with real numbers instead of forum folklore: Balanced at -10 LUFS for general streaming releases, Club at -8.5 for dance floors with punch intact, Club Hot at -7.5 when the reference is whatever the biggest record in the set is doing, Warm at -10 when the track wants softness more than level. Upload a mix and the engine hands you all four to compare, free, level-matched, before you pay anything.