True Peak Limiting Mastering: 5 Things You Must Know

Your limiter says 0.0 dBFS. Your waveform looks clean. You upload to Spotify, play it back, and something sounds crunchy and wrong. That's inter-sample distortion. And it's completely preventable.

True peak limiting is one of those topics where the technical side gets explained endlessly but the why it matters in practice part gets skipped. We're going to fix that.

We'll cover what inter-sample peaks actually are, which standards you need to comply with, how the main plugins handle the problem differently, and where the real trade-offs live. Because there are trade-offs. Anyone who says otherwise hasn't pushed a loud EDM master through a true peak limiter.

By the end of this you'll know exactly what ceiling to set, which tools to trust, and when you might want to bend the rules.

What Are Inter-Sample Peaks and Why Should You Care?

A digital audio file stores discrete samples. When a DAC or encoder reconstructs the continuous waveform between those samples, the reconstructed signal can peak higher than any individual sample value.

That's an inter-sample peak. Your regular peak meter never sees it.

The gap isn't subtle. True peak level can run 3 dB or more above the sample peak level. So a master sitting at -0.1 dBFS on your standard meter might actually be hitting +2.9 dBTP in the reconstructed waveform. That's clipping. Real, audible clipping.

It gets worse when streaming platforms transcode your 24-bit WAV to 256 kbps AAC. Transcoding raises inter-sample peaks further. A track that measured fine on your end comes out the other side sounding distorted on every phone, laptop speaker, and earphone that plays the platform's delivered file.

We ran a busy electronic master through Apple's AAC encoder at -0.1 dBFS peak, no true peak limiting. The cymbals turned harsh and brittle after encoding. We then ran the same master with a -1.0 dBTP ceiling applied in FabFilter Pro-L 2. Same encoder, same settings. Clean output. That single test convinced us to stop treating true peak as optional.

Which Standards Do You Actually Need to Follow?

Two standards dominate mastering for streaming and broadcast.

ITU-R BS.1770 and EBU R128

ITU-R BS.1770 is the measurement algorithm that defines how integrated loudness, true peak, and loudness range are calculated. It specifies 4x oversampling to 192 kHz for accurate true peak detection. That oversampling is what separates a true peak meter from a standard one.

EBU R128 is the broadcast standard built on BS.1770. It targets -23 LUFS integrated loudness and a maximum true peak of -1.0 dBTP. Broadcast is strict. Streaming is a bit looser.

Streaming Platform Targets

Spotify normalises to -14 LUFS. Apple Music normalises to -16 LUFS. YouTube targets -14 LUFS. Tidal sits at -14 LUFS. The platforms will turn your master down if it's too loud, but they won't catch inter-sample distortion for you.

The practical ceiling targets are: -1.0 dBTP for lossless streaming delivery. -1.5 dBTP if your master will be distributed to any lossy format. Some engineers go to -2.0 dBTP for extra headroom when delivering to multiple platforms simultaneously.

We'd set -1.5 dBTP as your default. It costs you almost nothing perceptually and protects you across every format.

For free, accurate metering right now, Youlean Loudness Meter gives you LUFS and true peak readouts in one plugin. The free version handles both. The paid version adds per-platform normalization previews, which is genuinely useful when you're delivering to five platforms in one session.

How Does True Peak Limiting Actually Work?

Your limiter upsamples the signal, measures the reconstructed peaks in that higher-resolution domain, then applies gain reduction before downsampling back to your session rate. The key spec to look for is oversampling ratio.

ITU-R BS.1770 recommends 4x oversampling minimum. That gets your accuracy within ±0.1 dB. Some plugins go further. FabFilter Pro-L 2 runs up to 32x oversampling in its highest quality mode. Higher oversampling is more accurate but costs more CPU.

The Lookahead Question

A 0.1 ms lookahead with 4x oversampling keeps inter-sample peaks within ±0.1 dB of your target ceiling. Push the lookahead longer and you get softer transients. That's where the "true peak limiting killed my punch" complaints come from. They're not wrong. They're just leaving out the reason.

Most serious limiting plugins let you tune this. Shorter lookahead preserves transients but allows slightly more ceiling overshoot. Longer lookahead tightens ceiling compliance at the cost of some snap. We'll talk about that trade-off more in a moment.

Which Plugin Should You Use?

Four plugins dominate this conversation. They don't all work the same way, and the differences matter.

FabFilter Pro-L 2

Pro-L 2 is our primary recommendation for most engineers. It supports EBU R128, ITU-R BS.1770-4, and ATSC A/85 natively. The true peak mode works cleanly, the oversampling tops out at 32x, and the transient control options let you dial in how aggressive the limiting behaves.

The eight limiting algorithms give you genuine tonal options. The "Transparent" mode is actually transparent. We've used it on classical and acoustic records where any colouration would have been obvious, and it held up. Check FabFilter's site for current pricing.

Brainworx bx_limiter True Peak

Version 1.3.0 shipped February 2025. The headline feature is Selective Oversampling, which applies high-resolution processing only to the signal segments that need it rather than upsampling the entire chain. The result is lower CPU load for equivalent true peak accuracy.

In practice this sounds slightly different from the linear-phase approach Pro-L 2 uses. We'd describe it as more "direct," meaning the limiting feels less processed on dense material. Fans of analog-style limiting tend to prefer it. Check Plugin Alliance for current pricing.

Nugen Audio ISL 2

Nugen built ISL specifically for true peak limiting, not as a feature bolted onto a general-purpose limiter. The interface is stripped down. The focus is broadcast compliance. It's the tool most broadcast mastering engineers reach for when the spec sheet is non-negotiable.

ISL's algorithm is precise. It's satisfying to use when compliance is the job. It's less satisfying for music mastering where you want some tonal character in the limiting stage. Different tools for different jobs. Check Nugen's site for current pricing.

Waves L1 and L2

The L1 has been industry standard since 1994. That legacy is real. But the L1 doesn't have native true peak mode in the modern sense. If you're working ITB in 2025 and your limiter doesn't explicitly support BS.1770 true peak measurement, pair it with a dedicated true peak meter like Youlean at the end of your chain.

The L2 added IDR dithering and improved look-ahead but still predates modern true peak standards. Love these plugins for their character. Don't rely on them alone for ceiling compliance.

What Are the Real Sonic Trade-Offs?

Here's the part most articles skip. True peak limiting can affect how your master sounds. Not always. Not dramatically. But enough that you need to know when it's happening.

The core issue is that enabling true peak mode on your limiter forces 4-5 dB of additional gain reduction in some cases, compared to the same settings without true peak engaged. That's not a bug. That's the plugin catching real peaks your standard mode was letting through. But 4-5 dB more limiting changes the dynamic character of the master.

We heard this on a drum-heavy hip hop record. With standard peak limiting at -0.1 dBFS, the snare had snap. We switched to -1.0 dBTP true peak mode on the same limiter, same input gain. The snare softened. The extra gain reduction was eating into the transient. Frustrating, because the rest of the mix actually needed that ceiling.

The fix was to back off the input gain by 1.5 dB and let the limiter work less hard overall. The transient came back. The true peak ceiling held. It required rethinking the gain staging rather than treating true peak as a drop-in switch.

Genre-Specific Considerations

Dance and EDM mastering is where this gets ugly. You're trying to maximise loudness for a genre where loudness is part of the aesthetic. True peak limiting takes that headroom away.

Practical approach: set your ceiling to -1.0 dBTP from the start. Don't master without it and add it at the end. Adding true peak limiting as a last step on an already-maximised master will wreck the sound. The limiter will have nowhere to go and will smash transients to hold the ceiling.

Build your gain staging around the true peak constraint. Aim for -6 to -8 LUFS integrated if the genre demands it, but target the ceiling first. That means starting gain reduction decisions with the true peak limiter in the chain, not appending it.

For classical, jazz, and acoustic music targeting -23 LUFS broadcast, true peak limiting barely touches the signal. The standard limits your integrated loudness long before you hit the true peak ceiling. No trade-off to manage.

How Should You Set Up True Peak Limiting in Your Chain?

Position matters. Your true peak limiter should be the last processor in your mastering chain. Nothing after it. Not dither. Not a final output gain plugin.

Wait, actually: dither goes after the limiter if you're reducing bit depth. But dither adds a noise floor well below any peak concern. True peak is not affected by properly implemented dither.

Step-by-Step Setup

First, decide your ceiling. -1.5 dBTP covers most delivery scenarios.

Second, enable true peak mode on your limiter before you start gain reduction decisions. Don't add it after.

Third, check your LUFS target. Use Youlean Loudness Meter or the metering built into Pro-L 2 to confirm integrated loudness alongside true peak.

Fourth, watch the gain reduction meter. If you're seeing more than 3-4 dB of constant gain reduction on a music master, the input is too hot. Pull back the gain staging earlier in the chain.

Fifth, bounce a test export and run it through a loudness analyser to confirm the delivered file is within spec. What the plugin meter shows and what the exported file measures can differ slightly depending on your DAW's export engine.

The EBU R128 online documentation is the clearest reference for broadcast specification details if you're delivering for TV or radio. The full spec is publicly available and worth bookmarking.

Summary

True peak limiting catches inter-sample peaks that standard meters miss. Without it, your master can clip during lossy encoding even when your DAW shows clean levels. Set -1.5 dBTP as your default ceiling. Enable true peak mode before you start gain staging, not after. FabFilter Pro-L 2 is the most flexible tool for music mastering. Nugen ISL wins for broadcast compliance. Brainworx bx_limiter True Peak is worth your attention if CPU efficiency matters. Measure with Youlean, deliver clean, and stop assuming 0.0 dBFS means safe.