Skip to content

Loudness range (LRA): what it measures and how to use it

Send feedback

Last updated 5 min readTags:MasteringDynamicsLoudness

Understand loudness range (LRA), how it differs from LUFS and crest factor, and when changes in your mix's loudness need attention.

Loudness range (LRA) describes how much loudness varies over the course of a track. It is measured in loudness units (LU). A lower reading means more consistent loudness; a higher reading means greater variation.

LRA is useful when a mix feels too uniform or its changes in level feel excessive. It gives you a reason to investigate, but the number alone cannot tell you whether the dynamics suit the music.

What does LRA measure?

LRA summarises the spread of short-term loudness readings across the audio you measure. It can reflect differences between verses, choruses and breakdowns, as well as variation within those sections.

It does not identify song sections or tell you exactly how much louder the chorus is than the verse.

For example, 6 LU LRA does not mean a 6 dB verse-to-chorus lift. That result could come from several different patterns of loudness throughout the track. To inspect a particular transition, listen to it and compare the short-term loudness readings on either side.

How is LRA calculated?

The method is defined in EBU Tech 3342. It uses overlapping three-second loudness measurements, then excludes readings below two thresholds: −70 LUFS, and 20 LU below the loudness calculated from the readings that pass the first threshold.

LRA is the difference between the 10th and 95th percentiles of the remaining readings. This reduces the influence of silence, very quiet passages and isolated loud events. [1]

If those percentile values are −18 LUFS and −12 LUFS, the LRA is 6 LU. One LU represents a 1 dB difference on the loudness scale.

What is a good LRA for music?

There is no universal LRA target for a good mix. The EBU's production guidelines explicitly reject a single limit for all material and stress that LRA cannot guarantee mix quality. [2]

A track built around sustained intensity may need little change in loudness. Another may depend on a quiet opening and a much louder final section. Neither approach needs correction merely because of its LRA.

The useful question is whether the changes in level serve the arrangement:

What you hear What to check
The chorus feels no stronger than the verse Arrangement, section levels and processing that may reduce contrast
Quiet passages become difficult to follow Vocal or lead-instrument levels, and whether the drop is intentional
Loud sections arrive too abruptly The transition, section automation and the balance of newly added parts
The track holds your attention and its transitions work Keep the dynamics unless you have a specific reason to change them

Use reference tracks with a similar arrangement and intended sound. Match their playback loudness before comparison. A reference with a sparse intro and a full-band ending may have a different LRA from a track with the same instrumentation throughout.

LRA, LUFS and crest factor: what is the difference?

These measurements answer different questions:

Measurement What it describes Useful question
Integrated loudness, in LUFS Overall loudness across the measured audio How loud is the track overall?
Short-term loudness, in LUFS Loudness over a moving three-second window How does loudness change through this passage?
LRA, in LU The spread of short-term loudness readings How much does loudness vary across the track?
Crest factor, in dB The difference between peak and RMS levels over the measured interval How far do the peaks sit above the average signal level?

Two tracks can have similar integrated loudness and different LRA. Two mixes can also have similar LRA while sounding quite different in their drum attack and transient detail.

Low LRA and low crest factor may justify a closer look at compression or limiting. They do not establish which processor caused a problem, or whether there is a problem at all. The arrangement and source recordings also affect the results.

How to use LRA when you revise a mix

1. Measure the complete track

Reset your meter before each pass and measure the same start and end points when comparing revisions.

Avoid judging a whole song from a short loop. The EBU cautions against LRA for short-form material because there may be too little data for a meaningful result. [3] EBU Mode meters must also flag the reading as unstable during the first 60 seconds. [4]

2. Find the passage that sounds wrong

Listen through at a comfortable, fixed playback level. Note where you want more contrast or where a level change distracts you.

Use the short-term loudness history to locate changes. LRA gives you one result for the measured audio; it does not show where those changes occur.

3. Test a change that addresses what you hear

If a chorus lacks impact, first check what changes in the arrangement. More level may not resolve a lack of musical contrast. Try adjusting the verse level, the chorus balance or the transition into it.

If you suspect bus compression or limiting, bypass it briefly and match the listening level. Compare the same transition. If the contrast improves, try less gain reduction or different settings.

If a quiet passage loses the vocal, adjust the vocal before compressing the whole mix. A local balance problem calls for a local correction.

4. Compare the result at matched loudness

Re-measure the full track and listen again. Keep the change if the passage works better. An increase or decrease in LRA is useful evidence of a change, but it is not the objective.

Sources

  1. EBU Tech 3342: Loudness Range, algorithm definition.
  2. EBU Tech 3343: Guidelines for Production of Programmes, section 3.3.
  3. EBU R 128 s1: Loudness Parameters for Short-form Content.
  4. EBU Tech 3341: Loudness Metering, EBU Mode.

    Type to search titles, headings and summaries.