Useful Monitoring Metrics References - LUFS, Crest Factor, LRA and more

Music production is a deeply technical subject that comes with a tonne of terminology - much of it borrowed from signal processing and electrical engineering - and it's not always easy to keep track of what everything means. 

This guide covers the core monitoring metrics you'll encounter when mixing and mastering: what they measure, why they matter, and what to look for. 

LUFS

LUFS (Loudness Units relative to Full Scale) is the measurement for perceived loudness across an entire track.

It’s the metric that you’re probably most familiar with, and is what streaming platforms use to normalize playback.

There are three LUFS values you'll typically see:

• Integrated LUFS is the average perceived loudness across the whole track. 
• Short-term LUFS is a rolling measurement over the last 3 seconds. 
• Momentary LUFS is a 400ms window. The most reactive of the three - can be useful for catching loud transients.

Streaming platforms publish loudness targets (Spotify is -14 LUFS, Apple Music is -16 LUFS), but as the 2026 Grammy nominees demonstrate, commercial releases regularly come in at -8 or -9 LUFS. The platforms turn loud tracks down, so going over the target isn't catastrophic - it's just a stylistic and genre choice.

Dynamic Range (LRA)

Loudness range or LRA is the difference between the quietest and the loudest sessions of a track.

Instinctively, it’s easy to imagine dynamic range as the result of the quietest peak subtracted from the loudest peak, but technically LRA is calculated using LUFS, and perceived averages - not True Peak (TP)

A high LRA means the track has significant dynamic contrast - a quiet intro, a big drop, a breakdown with genuine space in it. A low LRA means the track is compressed and consistent throughout.

Neither is inherently better. Billie Eilish's "WILDFLOWER" had an LRA of 17.4 LU at the 2026 Grammys. Lady Gaga's "Abracadabra" had 3.9 LU. Both are intentional, appropriate choices for those tracks.

Crest Factor

Crest factor is the difference between the peak level and the RMS (root mean square) level of a signal. 

RMS is the literal (mathematical) average of a signal's power - it's purely arithmetic, calculating the root mean square of the amplitude values over a window of time. 

A high crest factor means there's a lot of distance between the loudest peaks (transients) and the average level - plenty of punch and impact, but low perceived loudness. A low crest factor means the transients and the body of the sound are closer together - more compressed, louder-sounding, but with less punch.

Aggressive limiting reduces crest factor which can make a track feel louder and more competitive, but it also flattens kicks and snares and removes the sense of energy and movement.

True Peak (TP)

True peak measures the actual peak level of your audio, including inter-sample peaks - the distortion that can occur between samples when audio is decoded by a playback device or converted between formats.

Standard peak meters don't catch these - a waveform can show -0.5 dBFS on a regular meter and still clip on playback after encoding.

True peak is measured in dBTP, and the technical recommendation is to keep it under 0 dBTP before uploading to streaming platforms. In practice, plenty of major releases exceed this - the 2026 Grammy nominees averaged +0.13 dBTP, and "APT." by ROSÉ and Bruno Mars came in at over +2 dBTP.

Whether you follow the guideline depends on context. For most tracks, staying at or below -0.3 dBTP gives you a safe margin. For heavily limited, maximized club or pop masters, it's more of a suggestion.

Spectral Balance

Spectral balance describes the distribution of energy across the frequency spectrum - how much of your master lives in the low end vs. the mids vs. the highs.

There's no universal "correct" spectral balance. Different genres have very different signatures - a deep house record is supposed to be bass-heavy, whereas a bright vocal country track is supposed to have presence in the upper mids and highs. Trying to flatten everything to a neutral curve is not the goal.

Comparing your master's spectral curve to reference tracks in similar genres is one of the most practical ways to catch issues before you export - it doesn’t have to be an exact science, and you don’t have to try to measure and match specific values - just visually referencing another song’s frequency curve over time is all you need to do. 

Some plugins, like iZotope’s tonal balance control, are built specifically for this:

Stereo Width

Stereo width measurements tell you how much of your signal is in the mid channel (mono/center) vs. the side channel (stereo-only information).

Most metering plugins show this as a correlation meter or an M/S (mid-side) balance readout. A correlation meter reads +1 when the signal is perfectly mono, 0 when mid and side are equal, and goes negative when there's phase cancellation or the sides are dominant.

Red flags to watch for:

• Low-end energy in the sides. Sub and bass need to be mono. If kick and bass are spread in the stereo field, the track will lose energy in mono and cause problems on club sound systems.
• Phase correlation going negative. This usually indicates phase issues that will cause elements to cancel or disappear in mono playback.
• Extreme stereo width - it might sound impressive on headphones but it will collapse on speakers and sound systems.

Most DAW’s have a mono button - it’s useful to listen to your track in mono to get an idea for what’s going on.

And additionally, a good EQ that lets you control both mid and sides is essential - something like the FabFilter ProQ will make your mixing and mastering journey much easier.

Putting It Together

The role of these metrics is to tell you what's happening, not to tell you what to do. A good master isn't one that hits specific numbers across all six - it's one where the numbers make sense relative to each other and relative to the genre.

One of the most impactful things that you can do to improve your mixing and mastering is to ensure that you understand these metrics, have access to reliable metering plugins, and know how to make sense of them.

Even the best engineers don’t just rely on their ears - if you want your work to be precise, you need to be precise with your metering.