--- title: "Loudness Range (LRA): Meaning, Gating, and Limits | Gawergy Audio" description: "Learn how EBU Loudness Range summarizes longer-term loudness variation through gating and percentiles, and why it is neither a peak difference nor a quality score." canonical_url: "https://gawergy.com/learn/loudness-range-lra" md_url: "https://gawergy.com/learn/loudness-range-lra.md" last_updated: "2026-09-23" date_published: "2026-09-23" --- # What Is Loudness Range (LRA)? Loudness Range, or LRA, is an EBU-defined descriptor of how much a programme's loudness varies over a longer timescale. It is calculated from a distribution of short-term loudness readings after specified gating, then expressed in Loudness Units. It is not the difference between the highest sample and the quietest sample, and it is not a grade for how dynamic or well mastered a song is. ## Key takeaways - LRA describes the spread of time-varying loudness over a macroscopic timescale. - The EBU method uses three-second analysis windows, gates out very quiet material, and compares selected distribution percentiles. - A brief loud event or a fade-out need not dominate the result. - LRA differs from crest factor, generic dynamic range, integrated loudness, and maximum true peak. ## What problem LRA was designed to describe EBU Tech 3342 introduces Loudness Range as a supplement to programme loudness under EBU R 128. Programme loudness describes an overall loudness level; LRA describes the variation of loudness over a longer period. Those are different dimensions. A piece can have a given integrated LUFS value but move substantially between quiet and loud sections, or remain comparatively consistent throughout. A single integrated number cannot communicate that movement. LRA was created to add a robust descriptor of the variation, especially for decisions about programme handling and distribution. The standard uses the term *macroscopic timescale*. That directs attention away from one waveform sample or a single transient tip and toward changes across meaningful portions of a programme. A snare hit can be much taller than surrounding samples without making the song's LRA large. A quiet verse followed by a loud chorus can create broad loudness variation even if both sections have controlled sample peaks. The [dynamic-range-in-music article](/learn/dynamic-range-music-production) separates these timescales conceptually; this page explains the specific LRA algorithm and its limits. The EBU explicitly says LRA should not be confused with other measures such as generic dynamic range or crest factor. That caution matters because all of them can be casually called 'dynamics.' A technical description should name the measure and the programme interval, rather than treating a meter's LRA value as a substitute for listening to the arrangement. LRA is a descriptor, not a universal production target. ## LRA is a distribution statistic, not maximum minus minimum EBU Tech 3342 bases LRA on a statistical distribution of short-term loudness readings. It does not simply subtract the quietest measured instant from the loudest. The method estimates a spread using the 10th and 95th percentiles of readings that remain after gating. The lower percentile reduces the influence of a fade-out or unusual low-level section; the upper percentile reduces the influence of an isolated extremely loud event. This makes the result more robust than a raw extreme-to-extreme difference for the purpose the EBU defined. Percentiles have an intuitive meaning here: they describe positions within the retained distribution, not the amplitude of individual samples. The method discards the very lowest and highest tails of that distribution when forming its reported range. Therefore a brief gunshot-like event in a long programme, or a short ending fade, may not dictate the LRA. The EBU document gives these examples to explain why the statistic is useful. A listener can still find the event musically or emotionally important. Statistical robustness is not a statement that outliers do not matter aesthetically. The programme interval still changes the result. Measuring a single chorus and measuring an entire song are different questions. An LRA value for a short isolated fragment may be unstable or misleading because there are not enough representative loudness windows. The EBU itself warns about some very short programmes with leading or trailing silence. Report the analyzed segment and do not compare numbers from differently scoped excerpts as if they describe the same property. ## Three-second windows emphasize larger movement The LRA calculation begins with a sequence of loudness values obtained with a sliding three-second analysis window under the prescribed method. Consecutive windows overlap heavily so changes can be tracked without skipping shorter passages. This time base is much longer than a waveform cycle or a narrow transient tip. It is one reason LRA describes broad variation rather than immediate punch. A brief peak may still influence the loudness of nearby windows, but it cannot be interpreted as a simple peak-height input to the final LRA number. The sequence is then analyzed as a distribution. This is distinct from integrated loudness, which yields a programme-level average under its own method, and from momentary loudness, which follows a shorter window in EBU Mode. The EBU's metering specification distinguishes momentary, short-term, and integrated views. A meter that places LRA beside integrated LUFS is showing two complementary properties, not redundant readouts. The [RMS versus LUFS article](/learn/rms-vs-lufs) explains why these standardized loudness views are not simply raw amplitude averages. A producer who wants to understand a rhythmic transient should not expect LRA to diagnose its shape. A peak meter, waveform, short-window measure, and listening may be more relevant. A producer assessing whether a long programme varies strongly from section to section may find LRA informative. The measure is strongest when the question matches its timescale and weakest when used as a catch-all adjective for dynamics. ## Gating keeps silence from defining the range The EBU method uses a cascaded gate before estimating the distribution's spread. It excludes extremely low-level periods and applies a relative threshold tied to the programme. The purpose is to avoid allowing silence, background noise, or a low-level ambience to set an artificially wide range when the foreground programme remains fairly consistent. The current EBU Tech 3342 document specifies an absolute threshold of minus 70 LUFS and a relative threshold 20 LU below the absolute-gated loudness level. These are public standard parameters, not music-production settings or targets. Gating does not claim that quiet details lack artistic value. It is a statistical rule for a particular descriptor. A whispered opening or a fading tail may be central to the experience while having limited influence on the reported LRA. Conversely, sections above the gate contribute to the distribution according to the algorithm, not according to a listener's narrative sense of importance. This is why LRA should accompany listening, not replace it. The method has a defined purpose, and every purpose leaves some questions unanswered. Meter implementations are expected to follow EBU Mode requirements and test signals, but small differences can occur within permitted tolerance. EBU Tech 3342 notes that passing minimum test signals does not prove complete accuracy for every possible input. If two meters differ slightly, check their mode, reset state, analysed interval, and compliance rather than assuming the music changed. If they differ greatly, configuration or implementation may be at issue. The published algorithm is the reference, not the visual style of a plug-in panel. ## What a high or low LRA can and cannot mean A higher LRA generally indicates more spread in retained short-term loudness readings over the analyzed programme. A lower LRA indicates a narrower spread under the same method. Those statements are descriptive. They do not say whether the song is exciting, fatiguing, too quiet in a car, or artistically correct. A piece can deliberately maintain intensity for its full length; another can deliberately move from sparse to dense sections. The relevant aesthetic choice belongs to the material and intended playback context. The [louder-master article](/learn/louder-master-not-always-better) discusses such tradeoffs without prescribing a universal range. LRA also cannot tell you how sharp transients are. A mix could have consistent broad loudness with tall attacks, or broad loudness movement with soft attacks. Crest factor concerns peak-to-average relationships under its own window; LRA concerns a distribution of longer-term loudness. The [crest-factor guide](/learn/crest-factor-music-production) describes the former. Nor can LRA tell whether the file meets a true-peak ceiling. Maximum true peak is a separate EBU descriptor, and a programme can have any LRA while violating or meeting a peak specification. A common misuse is to treat LRA as a score to maximize or minimize. Doing so can encourage changes to the arrangement or dynamics that do not serve the music. Instead, use it to verify a narrower observation: did a change actually narrow the broad loudness movement, or is the sense of compression coming mainly from altered transients or tone? If the meter and ears disagree, inspect the timescale and listen to the specific sections. A single statistic cannot settle every perceptual question. ## How LRA fits beside other loudness measures Integrated LUFS summarizes programme loudness; LRA describes spread across retained short-term readings; a short-term LUFS trace shows where level changes occur; a true-peak meter addresses the reconstructed maximum. Together they describe more of a programme than any one alone. ITU-R BS.1770 supplies the programme-loudness and true-peak algorithms underlying common metering systems, while EBU Tech 3342 supplies this specific range descriptor. A meter's display can combine the values, but each is still a distinct calculation and should be cited by its proper name. A normalization system may adjust playback gain based on a loudness measure without compressing the programme's internal dynamic shape. That can shift the whole signal up or down while LRA remains broadly similar, absent additional constraints or processing. If a playback chain applies actual dynamic-range control, LRA may change. The [loudness-normalization guide](/learn/loudness-normalization) separates simple gain adjustment from compression. This distinction prevents a platform's volume match from being mistaken for an automatic rewrite of every transient or section relationship. The useful conclusion is bounded: LRA is a standardized, statistically robust measure of macroscopic loudness variation. It intentionally resists being dominated by isolated extremes and very quiet background. It is neither a complete definition of musical dynamics nor a target to chase. Use it when the question is about broad loudness movement, and state the analyzed programme and meter method so the number has a clear meaning. ## About G-Clipper Pro G-Clipper Pro affects waveform peaks and tone where its curve is active. EBU LRA describes broader variation in programme loudness and must be measured separately. ## Sources & References - [Loudness Range: A Measure to Supplement EBU R 128 Loudness Normalization](https://tech.ebu.ch/docs/tech/tech3342.pdf) - ['EBU Mode' metering to supplement EBU R 128 loudness normalisation](https://tech.ebu.ch/publications/tech3341) - [BS.1770: Algorithms to measure audio programme loudness and true-peak audio level](https://www.itu.int/rec/R-REC-BS.1770-5-202311-I) - [FabFilter Pro-L 2 Help - Loudness metering](https://www.fabfilter.com/help/pro-l/using/loudnessmetering) ## Continue Reading - [What Is Dynamic Range in Music Production?](https://gawergy.com/learn/dynamic-range-music-production) - [Crest Factor in Music Production](https://gawergy.com/learn/crest-factor-music-production) - [What Is Loudness Normalization?](https://gawergy.com/learn/loudness-normalization) ## Sitemap See the full [Gawergy.com sitemap](https://gawergy.com/sitemap.md).