Every DJ mix uploaded to SetMaster Pro is measured before anything is done to it. We have now measured 826 of them, from 258 DJs, adding up to more than 1,100 hours of audio. The headline is simple: nearly half of them arrived too quiet for streaming, and only about one in five arrived anywhere near the level SoundCloud, Spotify and YouTube play back at. This is what the rest of the numbers say, what they mean for the mix you are about to upload, and how much the measurements can and cannot tell you.
The short answer in 30 seconds
Across 826 DJ mixes measured on arrival between June and November 2026, the typical mix came in at -14.8 LUFS integrated, just under a decibel below the -14 LUFS streaming target. That middle figure hides a wide spread. Nearly half the mixes were quieter than -15 LUFS and more than a quarter were quieter than -18. Three in ten were louder than -13. About one in six had peaks over full scale. After mastering to the Streaming / Radio profile, 99.7% landed within a decibel of -14 LUFS, and none of the masters had peaks over full scale.
How loud DJ mixes are when they arrive
Median loudness: -14.8 LUFS integrated.
Quieter than -15 LUFS: 47.8% of mixes.
Quieter than -18 LUFS: 28.2%, at least 4 dB short of the streaming target.
Within a decibel of -14 (between -15 and -13 LUFS): 21.4%.
Louder than -13 LUFS: 30.8%.
So nearly half the mixes were too quiet, three in ten were too loud, and one in five was close to right. A DJ uploading straight from the recorder is more likely to be off target than on it, in one direction or the other.
Why "too quiet" is the bigger problem
It is tempting to think loudness normalisation sorts this out. Partly it does. Streaming platforms are built to turn loud uploads down, and they do that reliably: a mix at -9 LUFS is played about 5 dB quieter, at the same level as everything else.
Quiet uploads are a different story. Platforms are much more cautious about turning a mix up, because raising the level also raises the peaks, and they will not push those past a safe ceiling. Spotify, which documents this in detail, gives the example of a track at -20 LUFS that it can only lift to -16. So the more than one in four mixes arriving below -18 LUFS are the ones most likely to play noticeably quieter than the mix before them in a listener's feed. That is the gap mastering closes: of the mixes mastered to the Streaming / Radio profile, 48% were brought up by a decibel or more, by a typical 4.5 dB.
Clipping: one mix in six
16.0% of mixes had peaks over full scale when they arrived. A peak over full scale cannot be played back cleanly: somewhere between the file and the speaker it is cut flat, and that is the harsh fizz on the loudest kicks that makes a recording sound cheap on a good system.
Most of it starts in the booth, with a mixer master or recorder input set too hot. Because a quieter recording is easy to bring up and a clipped one can only be repaired, the fix is to leave more room when recording, not less. We cover how in how to fix a clipped DJ mix recording.
DJ mixes are long
66.9% of the mixes were longer than an hour. That matters more than it sounds. Over an hour or two, a mix moves through records mastered in different decades to very different levels, and a quiet track halfway through is one of the most common faults in a long recording. A single loudness figure for the whole file tells you where the mix sits on average. It does not tell you how evenly it holds that level from the first record to the last, which is why long mixes need mastering that rides the level across the whole length, not just a gain change at the end. See how to master a long DJ mix for more on that.
What mastering changed
701 of the mixes were mastered with the Streaming / Radio profile, which targets -14 LUFS integrated and holds the peaks 1 dB below full scale.
Landed within a decibel of -14 LUFS: 99.7%.
Brought up by a decibel or more: 48%, by a typical 4.5 dB.
Brought down by a decibel or more: 32%, by a typical 3.2 dB.
Masters with peaks over full scale: none, across every profile.
The third of mixes that were brought down matter as much as the nearly half that were brought up. Many DJs record hot on purpose, assuming louder is better. On a streaming platform it is not: a hot mix is turned down anyway, and it keeps the extra limiting that made it hot. A master at -14 arrives at the same level without paying that price.
What this data can and cannot tell you
These numbers are honest, but they have limits, and it is worth being clear about them.
The sample is self-selected. These are mixes from DJs who chose to upload to a mastering service. DJs who already master their own mixes, or never think about it, are not here. The real spread across all DJ mixes online could be wider or narrower.
There is no genre or software breakdown. We do not record what style a mix is, or which DJ software or hardware made it, so we cannot say whether drum and bass mixes arrive louder than deep house ones.
Repeat uploads were removed. Where the same DJ uploaded the same recording more than once, for example to try a second profile, it is counted once. Our own test uploads are excluded.
A few DJs upload a lot. Ten DJs account for about a third of the mixes. Counting each DJ once instead of each mix gives much the same picture: just under half arrive too quiet, more than a third too loud, and fewer than one in five close to the streaming level.
Nothing is identifiable. Every figure here is an aggregate. No individual mix, DJ or file is named or published.
Loudness is measured to the ITU-R BS.1770 standard, the measurement behind LUFS. Peak figures are the highest sample value in the file.
Check your own mix
The quickest way to see which group your mix falls into is the free DJ Mix Analysis. It runs in your browser, reads the whole file without uploading it, and reports the integrated loudness and peak level in a couple of minutes. If it reads well below -14 LUFS, or the peaks sit at 0 dB, the mix needs mastering before it goes up. For the background on what the number means, see what LUFS means for a DJ.
Frequently asked questions
How loud are most DJ mixes?
Across 826 DJ mixes measured on arrival at SetMaster Pro between June and November 2026, the median was -14.8 LUFS integrated, just under a decibel below the -14 LUFS streaming target. Nearly half were quieter than -15 LUFS, more than a quarter quieter than -18, and three in ten louder than -13.
What percentage of DJ mixes are clipped?
In our sample, 16.0% of mixes arrived with peaks over full scale, which means clipping on playback. Most of it comes from a mixer or recorder level set too high during the recording.
Do streaming platforms fix a quiet DJ mix?
Not reliably. Platforms turn loud uploads down, but they are cautious about turning quiet ones up because that would push the peaks too high. A mix far below -14 LUFS can play noticeably quieter than the mastered mixes around it.
Should I record my DJ mix louder, then?
No. Record with headroom, so the loudest moment peaks around -6 dB, and bring the level up in mastering. A quiet, clean recording masters well; a recording pushed too hot clips, and clipping can only be repaired, not undone.
Where does this data come from?
From DJ mixes uploaded to SetMaster Pro for mastering, measured before any processing. Repeat uploads of the same recording and our own test files are excluded, and only aggregate figures are published.
