Broadcast loudness is governed by two widely used documents: EBU R 128 in Europe and ATSC A/85 in the United States. Both use the ITU-R BS.1770 measurement and their targets are only 1 LU apart, but they differ in what is measured, how tolerances work and how much peak headroom they require. This comparison uses the current versions, EBU R 128 version 5 (2023) and ATSC A/85:2026-07, and points out figures that are still quoted from older editions.
At a glance
| EBU R 128 (V5, 2023) | ATSC A/85 (2026-07) | |
|---|---|---|
| Target | -23.0 LUFS | -24 LKFS |
| Tolerance | ±1.0 LU only where the target is not practically achievable, such as live; ±0.2 LU for measurement error in QC | About ±2 dB of measurement tolerance; do not target the edges |
| True peak | -1 dBTP maximum in production (±0.3 dB measurement tolerance) | Below -2 dBTP (±0.5 dB measurement tolerance) |
| What is measured | The whole programme, with no emphasis on speech | Long-form: dialogue anchor. Short-form: full mix |
| Short-form | Supplement s1: maximum short-term loudness -18.0 LUFS | Full mix, BS.1770-3 or later |
| Streaming | Supplement s2: -23.0 LUFS; interim -20.0 to -16.0 LUFS without metadata | Streaming annexes: -23 to -27 LKFS |
EBU R 128
Target and tolerance
R 128 states that programme loudness "shall be normalised to a Target Level of -23.0 LUFS". Where that is not achievable practically, for example in live programmes, a tolerance of ±1.0 LU is permitted, and the EBU adds that drifting toward the edges of that tolerance should not become standard practice. Separately, quality-control workflows may allow ±0.2 LU to account for measurement errors.
You will still see ±0.5 LU quoted as the R 128 tolerance. That was the wording of version 3, from 2014, and is no longer current.
The whole programme, not just speech
R 128 measures the whole programme "without emphasis on specific foreground elements such as speech, music or sound effects". There is no dialogue gate: what counts is the integrated loudness of the full mix, with the standard BS.1770 gating described in LUFS vs RMS vs peak.
True peak
The maximum true peak is -1 dBTP during production, with a measurement tolerance of ±0.3 dB. R 128 notes that lower maximums may apply for particular distribution systems and data-reduced codecs, and EBU Tech 3344 recommends an end-stage limiter at -2 dBTP before the distribution encoder.
Loudness range
Loudness range (LRA) describes how much loudness varies across a programme: per EBU Tech 3342, it is the spread between the 10th and 95th percentiles of the short-term loudness. R 128 says LRA "may be used to evaluate the loudness variation" but sets no maximum, and EBU Tech 3343 adds that LRA should not be a brick-wall parameter. LRA is not recommended for programmes shorter than one minute. Broadcasters may set their own limits: EBU Tech 3401 reports, for example, that the BBC works to a maximum LRA of 6 LU for speech in its studios. Those are broadcaster guidelines, not EBU requirements.
Short-form content: R 128 s1
Adverts and promos follow supplement s1, which defines short-form content as up to approximately 2 minutes and typically shorter than 30 seconds. It keeps the -23.0 LUFS target and the -1 dBTP ceiling and adds a maximum short-term loudness of -18.0 LUFS, 5 LU above the target, with a +0.2 LU QC tolerance. LRA is not applicable. The short-term cap limits how loud a brief, dense spot can get even when its integrated loudness is on target.
Streaming and radio
Supplement s2 says programmes should be streamed unchanged at -23.0 LUFS, but where loudness metadata is not used it gives an interim distribution level between -20.0 and -16.0 LUFS. Supplement s3 for radio keeps production at -23 LUFS and -1 dBTP and allows distribution within the same -20.0 to -16.0 LUFS range.
ATSC A/85
Target and tolerance
A/85 sets a target loudness of -24 LKFS. It anticipates a measurement tolerance of up to approximately ±2 dB around that value and states that content loudness should not be targeted to the high or low side of it. The ±2 dB allows for differences between measurements; it is not permission to deliver at -22.
The dialogue anchor
For long-form content, A/85 measures the dialogue anchor: loudness taken with a dialogue-gated BS.1770-1 meter or, if that is unavailable, a BS.1770-2 or later meter with dialogue gating switched on. The programme level is therefore set by its dialogue rather than by music or effects. Short-form content, such as commercials, is measured on the full mix with BS.1770-3 or later.
For a programme that is almost entirely speech, dialogue loudness and full-mix loudness tend to be close. For content with loud music or effects they can differ substantially, which is the main practical difference between delivering to R 128 and to A/85.
True peak
A/85 asks for true peaks to be kept below -2 dBTP, with a measurement tolerance of ±0.5 dB. That is 1 dB more headroom than R 128 requires in production.
Versions and regulation
ATSC approved A/85:2026-07 on 8 July 2026. It keeps the -24 LKFS target, the approximate ±2 dB tolerance and the -2 dBTP limit of A/85:2013, and adds annexes for streaming with codecs that do not carry loudness metadata, which recommend a value between -23 and -27 LKFS. Regulation lags behind: the CALM Act refers to A/85 "and any successor thereto", but the FCC rules in 47 CFR 73.8000 still incorporate A/85:2013, and the FCC applies it to commercials.
What the differences mean in practice
- The 1 LU gap. A file mastered to -23 LUFS for Europe sits 1 LU above the US target. Re-master to -24 LKFS rather than relying on the tolerance.
- Different ceilings. A master that peaks at -1 dBTP meets R 128 but not A/85. Limiting a little below -2 dBTP satisfies both.
- Different measurements. R 128 measures the full programme; A/85 measures dialogue for long-form. For speech-led programmes the difference is usually small; for drama or effects-heavy content it can be large.
- House specifications. Broadcasters publish their own delivery specifications built on these standards, and they can be stricter, for example on loudness range, true peak or file format. Always check the specification you are delivering to.
LUFS and LKFS are the same unit. R 128 says so in a footnote. The EBU uses LUFS; ITU-R BS.1770 and ATSC use LKFS.
Other regional specifications build on the same measurement. Free TV Australia's OP-59, for example, uses -24 LKFS, with deviation that should generally not exceed ±1 LU, and a maximum true peak of -2 dBTP. For more targets in one place, see the loudness targets reference; for why ceilings sit at -1 or -2 dBTP, see true peak vs sample peak.
LufsReady includes broadcast presets for both standards: EBU R 128 at -23.0 LUFS with a -1 dBTP ceiling, and ATSC A/85 at -24 LKFS with a -2 dBTP ceiling, each delivered as 24-bit, 48 kHz WAV with a measurement report. Both measure the integrated loudness of the full mix rather than a dialogue anchor, so for drama or effects-heavy long-form content, confirm the result with your broadcaster's own measurement.
Sources
- EBU R 128: Loudness normalisation and permitted maximum level of audio signals
- EBU R 128 s1: Loudness parameters for short-form content
- EBU R 128 s2: Loudness in streaming
- EBU R 128 s3: Loudness in radio
- EBU Tech 3342: Loudness range
- EBU Tech 3343: Guidelines for production of programmes in accordance with EBU R 128
- EBU Tech 3344: Practical guidelines for distribution systems in accordance with EBU R 128
- EBU Tech 3401: Loudness in radio, case studies
- ATSC A/85:2026-07: Techniques for establishing and maintaining audio loudness for digital television
- ATSC A/85:2013 with Corrigendum No. 1
- ITU-R BS.1770: Algorithms to measure audio programme loudness and true-peak audio level
- Free TV Australia OP-59: Measurement and management of loudness for TV broadcasting
Check your file in seconds
Measure loudness, true peak and noise floor for free in your browser, or master a file to the exact target with a report.