Understanding EBU R128: More Than Just a Number

The European Broadcasting Union's R128 recommendation, first published in 2010 and revised several times since, was born out of the need to eliminate the chaotic loudness levels that plagued television and radio. Before R128, broadcasters used traditional peak metering (PPM) that didn't correlate with perceived loudness. A quiet dialogue scene could be followed by an ear-splitting commercial, driving audiences to reach for the remote. R128 introduced a standardized method for measuring and normalising loudness, using the unit LUFS (Loudness Units relative to Full Scale) – also known as LKFS in the ITU-R BS.1770 standard. The core principle is simple: measure the average loudness of a programme over its entire duration (Integrated Loudness) and target a specific level – -23 LUFS for broadcast. This ensures that every piece of content, from a live sports event to a late-night documentary, sits at a consistent perceived volume.

But R128 is not just about a single number. It addresses three key parameters:

  • Integrated Loudness (I): The average loudness over the entire programme, measured in LUFS. The target is -23 LUFS with a tolerance of ±0.5 LU for absolute compliance. This measurement is gated to ignore silence, preventing quiet passages from artificially lowering the average.
  • True Peak (TP): The absolute maximum level of the analogue waveform after digital-to-analogue conversion. High true peaks can cause distortion on consumer devices. R128 recommends a maximum of -1 dBTP. Unlike sample peak, true peak detection must be oversampled to catch inter-sample peaks.
  • Loudness Range (LRA): A measure of how loudness varies throughout a programme, expressed in LU (Loudness Units). LRA helps preserve the dynamic contrast (e.g., a quiet whisper vs. an explosion) while keeping the overall loudness consistent. A typical LRA for drama might be 12–18 LU, while for news it might be 4–6 LU. The EBU specifies that LRA is measured from the lowest 10% to the highest 5% of the loudness distribution, excluding silence.

Understanding these three pillars is the foundation of practical compliance. For deeper reading, refer to the official EBU R128 specification and the associated measurement standards EBU Tech 3341 (loudness metering) and Tech 3342 (loudness range).

Practical Implementation Tips for Broadcast Engineers

Choose a Compliant Loudness Meter

Not all loudness meters are created equal. You need a tool that implements the ITU-R BS.1770 algorithm correctly, with true peak detection that is anti-aliasing and oversampled. Free options include Youlean Loudness Meter and TBProAudio's dpMeter. Professional solutions like iZotope Insight, Nugen LM-Correct, or Waves WLM Plus offer additional features such as real-time correction and streaming loudness monitoring. Ensure your meter displays:

  • Integrated Loudness (I) – the long-term average over the entire programme.
  • Short-Term Loudness (Momentary) – a 3-second sliding window for real‑time adjustments.
  • True Peak level in dBTP.
  • Loudness Range (LRA) – to check your dynamic consistency.
  • Gating (speech-gated loudness) – for programmes with significant silence, as per EBU Tech 3341.
  • Histogram or loudness distribution view – helps visualise LRA and spot outliers.

Calibrate your monitoring chain so that your reference speakers reproduce -23 LUFS at a comfortable 79 dB SPL. Use a pink noise burst at -18 dBFS to set this level. This calibration ensures that what sounds right in the control room translates to the home listener.

Set Your Levels During Mixing, Not After

The worst time to fix loudness is in the final master. Integrate R128 targets into your mixing workflow from the start. Set your monitoring chain so that -23 LUFS sounds comfortable in your control room. Use a loudness meter on your master bus, but also regularly check individual stems. For example, a quiet dialogue stem mixed at -33 LUFS will pull down the entire programme’s integrated loudness unless you compensate with other elements. Keep a consistent balance theme throughout. When mixing for broadcast, avoid heavy compression on the master bus; instead, manage dynamic range at the source using clip gain, automation, and careful compression on individual tracks. For dialogue, use a dedicated compressor with a fast attack and release to keep speech intelligible without pumping. For music beds, side‑chain compress them slightly under dialogue to maintain clarity.

Consider employing a loudness browser plugin that lets you see multiple loudness parameters at once. Many modern DAWs offer built‑in loudness monitoring, but always verify with an external meter for compliance.

Mastering True Peak Compliance

True peaks are the bane of many broadcast mixes. Limiting to -1 dBTP is not just about slapping a brickwall limiter at -1 dB. Use a loudness meter with accurate true peak detection, then apply an oversampled limiter (at least 4x oversampling) with a ceiling of -1.1 dBTP for safety. Some engineers advocate for -2 dBTP to avoid any inter-sample peaks in consumer DACs. However, note that extremely low true peaks can reduce perceived loudness. The standard says -1 dBTP is the absolute maximum, so aim a little lower. Always check your mix on different playback systems – true peak distortion often becomes audible on poor-quality converters. If you find clicks or distortion after limiting, increase the limiting threshold or adjust the attack time. For super‑clean limiting, use a clipper on individual tracks before the master limiter.

Managing Loudness Range (LRA) for Different Genres

LRA is your friend when you want to preserve artistic dynamics without causing listener fatigue. For news, talk shows, and live sports, keep LRA below 6–8 LU. The programme is mostly speech and needs to remain intelligible. For drama, music, or documentaries, you can allow higher LRA (12–18 LU) but be aware that the EBU R128 requirement still applies to the integrated loudness. If your LRA exceeds 20 LU, parts of your programme may sound too quiet or too loud on-air. Use loudness range as a diagnostic tool: if the LRA is unexpectedly high, look for sections that are too quiet (e.g., long pauses) or too loud (e.g., a shouted word). You can apply makeup gain to quiet sections or use volume automation to bring extremes into a tighter dynamic range without losing the emotional punch. BBC R&D has published practical studies on how LRA affects listener experience.

For music content, LRA should ideally stay below 12 LU to maintain consistency in a playlist. Classical music may naturally have higher LRA, but broadcasters often require a gentler hand with dynamic compression to fit within the norm.

Advanced Workflows: Gating and Short-Term Loudness

EBU Tech 3341 specifies a gate (a threshold of -70 LUFS absolute plus -10 LU relative to the ungated loudness) to ignore silence when calculating integrated loudness. This is crucial for programmes with gaps (e.g., between songs or scenes). Without gating, a quiet stretch could artificially lower the integrated loudness measurement, leading to the whole mix being pushed too loud in an attempt to meet -23 LUFS. Ensure your loudness meter applies the appropriate gating as defined by the EBU. Some meters allow you to see both gated and ungated values – always use the gated reading for compliance verification. Additionally, use Short-Term Loudness (3-second window) to spot problematic transitions. Sudden spikes in short-term loudness (above -18 LUFS, for example) can cause the programme to feel inconsistent even if the integrated loudness is correct. Smooth these transitions with automation or careful leveling.

Mixing for Different Delivery Platforms

Broadcast is not the only game anymore. Many engineers now deliver to streaming, podcasts, and digital TV with different loudness targets. While EBU R128 applies primarily to European broadcast, streaming platforms like Netflix and Spotify use similar LUFS targets (often -24 LUFS for surround, -27 LUFS for stereo, or the platform-specific norm). You can use the same meter and workflow, but adjust the target. Create mix presets for each platform. A common mistake is mixing for streaming and then delivering the same mix for broadcast without re-measurement. Always re-check with the correct target. Also note that loudness normalisation in streaming often applies a gain offset, so true peak limitations are even more critical (many streaming services reject files with true peaks above -1 dBTP). For podcast delivery, target -16 LUFS integrated with a true peak of -1 dBTP (as per ITU‑R BS.1770‑4), but check your specific platform’s guidelines.

Common Pitfalls and How to Avoid Them

  • Over-compressing to hit -23 LUFS: If your mix sounds too dynamic and the integrated loudness is low, don't just compress everything. Instead, adjust balance and automate level. Over-compression destroys mix quality and increases listener fatigue. Use loudness range as a guide; if LRA is too high, use clip gain or volume automation on individual elements before resorting to compression.
  • Forgetting about loudness of individual clips: A loud explosion sound effect might push true peaks above -1 dBTP even if the overall mix is quiet. Trim sound effects and foley to prevent spikes. Use a clip gain tool to reduce levels by 3–6 dB before they hit the master bus. Check peak levels on individual tracks with a true peak meter.
  • Ignoring the headphone mix: Some engineers monitor through headphones with a different calibration. Ensure your monitoring environment is calibrated to a reference level (e.g., 79 dB SPL from your loudspeakers when a -18 dBFS sine wave is played). For headphones, use a measurement mic or a reference track. Consider using headphones with a neutral frequency response and a correction curve if needed.
  • Relying on a single meter reading: Different meters can show slight variations due to algorithm differences. Stick to one meter that you trust and that is fully compliant with EBU Tech 3341. Periodically calibrate it against a known reference file. For critical delivery, run your mix through a second meter for cross‑verification.
  • Not documenting your loudness measurement: For quality assurance, maintain a log of integrated loudness, true peak, and LRA for each programme. This helps prove compliance and troubleshoot problems if a complaint arises. Many DAWs and metering plugins allow you to export a loudness report.
  • Misinterpreting LRA: LRA is not a target to hit but a diagnostic tool. A news programme with LRA of 10 LU may still sound good; a drama with LRA of 8 LU may feel squashed. Use LRA to inform your compression and automation decisions, not as a strict limit.

Conclusion: R128 as a Creative Tool

Far from being a restrictive constraint, EBU R128 gives sound mixers and broadcast engineers a reliable framework to ensure their work translates well across thousands of receivers. By mastering the concepts of integrated loudness, true peak, and loudness range, you can preserve the emotional impact of your mix while meeting technical specifications. Invest in a good metering setup, integrate loudness monitoring into your mixing workflow, and treat R128 as a standard that empowers you to deliver consistent, high-quality audio. The result is a better experience for the listener – and fewer complaints about volume jumps. For further official guidance, consult the EBU R128 implementation guidelines (Tech 3343) which includes practical examples and a sample workflow.

Remember: a compliant mix is not necessarily a good mix, but a good mix is always compliant. Use the tools, trust your ears, and let R128 be your compass, not your cage. With practice, loudness compliance becomes second nature, freeing you to focus on the art of audio storytelling.