LUFS Loudness for Video Post-Production: Targets, Limits, and Workflows That Do Not Get Your Audio Rejected

LUFS (Loudness Units Full Scale) measures perceived loudness — not peak level, not RMS, but how loud your audio actually sounds to a human ear weighted across frequency and time. This distinction matters because a mix with a sharp transient hitting -0.3 dBFS can measure at -20 LUFS integrated, while a heavily compressed master peaking at -3 dBFS can measure -10 LUFS. They hit the same peak meter. They sound completely different. If you are still mixing to peak meters, you are delivering inconsistent loudness across your projects.

The loudness measurement you care about depends on the time window. Momentary LUFS uses a 400ms window — it captures transient spikes and is useful for catching individual events like a door slam or a sudden music hit. Short-term LUFS uses a 3-second window — it tracks the loudness of phrases and scenes, useful for monitoring dialogue consistency. Integrated LUFS measures the entire program from start to finish with a gating mechanism that ignores silence and very quiet passages. Integrated LUFS is what platforms check and what delivery specs reference.

Opinionated take: targeting -14 LUFS for YouTube is technically correct but practically unnecessary for most content. YouTube's normalization will adjust your audio regardless. If you deliver at -16 to -18 LUFS with clean dynamics, YouTube will not touch it, and your audio will have more punch and headroom than the competition who slammed everything to -14. The only time -14 matters is when your content is played outside YouTube's player — embedded players that bypass normalization, or downloaded copies.

This is where most people get confused, and for good reason — the terminology is unhelpful. Here is a practical breakdown by use case.

Dialogue consistency in post-production: watch Short-term LUFS (3-second window). Your dialogue should hover within a 4-6 LUFS range on the short-term meter. If you see readings jumping between -24 and -14 LUFS short-term, your dialogue levels are inconsistent and the viewer will reach for the volume knob. Use the short-term meter during your initial mix to establish a dialogue baseline, then check integrated at the end.

Music mixing for streaming: Integrated LUFS is the only measurement that matters for compliance. But use Momentary LUFS during the mix to catch problematic peaks — if you see momentary readings above -5 LUFS, you probably have a transient that will trigger your true peak limiter aggressively and cause audible pumping.

Broadcast delivery: Integrated LUFS for compliance, plus you need to monitor the Loudness Range (LRA). EBU R128 (EBU Tech 3341) recommends an LRA below 15 LU for general content and below 20 LU for cinematic content. LRA above 20 LU means your quietest and loudest moments are too far apart — viewers will complain about volume jumps between dialogue and action scenes.

True peak (dBTP) measures inter-sample peaks — signal levels that occur between digital samples when the audio is reconstructed by a DAC, as defined in ITU-R BS.1770-4. A waveform can show -0.5 dBFS on a standard peak meter but produce peaks above 0 dBFS on playback due to inter-sample reconstruction. This causes clipping in the DAC, downstream encoding artifacts in lossy codecs (AAC, Opus), and distortion on consumer playback systems.

Broadcast and most streaming platforms require -1 dBTP. Netflix specifies -2 dBTP. The extra headroom accounts for lossy encoding that happens after you deliver your master. When YouTube encodes your -14 LUFS audio to Opus or AAC at 128-256 kbps, the codec can introduce new peaks up to 1 dB above your original. If you delivered at -0.5 dBTP, the encoded version clips at +0.5 dB, and the listener hears distortion on their phone speaker or headphones.

Use a true peak limiter, not a standard peak limiter. In Pro Tools, the iZotope Ozone Maximizer with True Peak mode enabled catches inter-sample peaks. In DaVinci Resolve Fairlight, the built-in limiter has a True Peak option — enable it and set the ceiling to -1.0 dB for broadcast or -2.0 dB for Netflix. In FabFilter Pro-L 2, select the True Peak mode and set your ceiling. These oversample the signal to detect inter-sample peaks that a standard limiter misses entirely.

iZotope Insight 2 is the industry standard for loudness metering in post-production. It displays integrated, short-term, and momentary LUFS simultaneously, shows a loudness history graph so you can see the full program trajectory, and includes a true peak meter. It runs as a plugin in Pro Tools, Nuendo, Logic, and DaVinci Resolve Fairlight. Price is around $199 standalone or included in iZotope's Music Production Suite.

Youlean Loudness Meter is a free plugin (with a paid upgrade for additional features) that displays integrated LUFS, loudness range, and true peak. It logs measurements over time and can export PDF reports for delivery documentation. For a free tool, it covers 90% of what you need. The paid version adds history recall and batch analysis, which is worth it if you deliver to broadcast regularly.

DaVinci Resolve Fairlight has a built-in loudness meter accessible from the Fairlight page: Fairlight menu, then Loudness. It measures integrated LUFS with EBU R128 gating, shows short-term and momentary readings, and displays the loudness history. It is adequate for basic compliance checking. For detailed analysis and delivery reports, pair it with Youlean or Insight.

In Pro Tools, the workflow is similar but the tools differ. Use a loudness meter plugin like Youlean or Insight on the master bus. Use a true peak limiter as the last insert on the master — iZotope Ozone Maximizer, FabFilter Pro-L 2, or the Avid Channel Strip limiter with true peak mode. Pro Tools does not have a built-in loudness meter, so a third-party plugin is mandatory for compliance work.

One limitation worth disclosing: loudness metering in real-time during mixing can mislead you if your project has long silent sections or extended very quiet passages. The integrated LUFS measurement gates out silence below -70 LUFS, but very quiet ambient content that sits at -50 to -60 LUFS still pulls the integrated reading down. A documentary with 30 seconds of near-silence between interview segments will measure 1-2 LUFS quieter than the actual perceived loudness of the content. Always check the loudness history graph to identify these sections.

YouTube normalizes audio to -14 LUFS integrated, but only in one direction. If your video measures -10 LUFS (too loud), YouTube applies gain reduction to bring it to -14 LUFS. If your video measures -20 LUFS (too quiet), YouTube does nothing. Your video plays at -20 LUFS, which will sound noticeably quieter than other content on the platform.

This asymmetry exists because applying gain to quiet content amplifies noise. A video recorded on a phone with a noisy mic at -20 LUFS would sound terrible if YouTube boosted it 6 dB to reach -14 LUFS — the noise floor would come up by the same amount. So YouTube leaves quiet content alone and only pulls loud content down.

The practical implication: if you want your YouTube content to sound competitive, you need to mix to approximately -14 LUFS yourself. YouTube will not help you if you deliver quiet. But you also should not slam your mix to -12 or -10 LUFS expecting it to sound louder — YouTube will turn it down and you lose dynamic range for no benefit. The sweet spot is -14 to -16 LUFS integrated with -1 dBTP true peak.