--- title: "True Peak Limiting: Reconstructed Peaks and Real Limits | Gawergy Audio" description: "Understand how true-peak limiting differs from a sample ceiling, why reconstructed peaks matter, and what later codecs or conversion can still change." canonical_url: "https://gawergy.com/learn/true-peak-limiting" md_url: "https://gawergy.com/learn/true-peak-limiting.md" last_updated: "2026-09-23" date_published: "2026-09-23" --- # True Peak Limiting: What It Is and What It Isn’t A true-peak limiter aims to keep an estimated reconstructed waveform below a specified dBTP ceiling at its output. An ordinary sample ceiling constrains stored sample values, which can miss peaks between them. True-peak control can be valuable for delivery specifications, but its result depends on the limiter's method and on where measurement occurs. It does not guarantee that every later codec, sample-rate conversion, or playback device will produce an identical maximum. ## Key takeaways - Reconstructed audio can peak between samples above the largest stored sample. - True-peak metering estimates that reconstructed maximum; true-peak limiting changes gain or signal shape to constrain it. - A product's true-peak claim applies to its specified output path and mode, not to arbitrary later processing. - True-peak compliance, loudness, and musical quality are separate questions. ## Why a sample ceiling can miss a higher peak A digital file stores samples of a band-limited waveform. Playback reconstruction creates a continuous-time signal consistent with those samples under the system's filtering assumptions. That curve can reach a maximum between the stored sample positions. A sample-peak meter that looks only at stored values may therefore report a lower number than a meter estimating the reconstructed waveform. ITU-R BS.1770 defines a true-peak measurement approach for this issue. The existing [sample-peak versus true-peak guide](/learn/sample-peak-vs-true-peak) teaches the measurement difference; this article focuses on a processor intended to control the reconstructed maximum. The effect is not limited to one processor type. Fast limiting, hard clipping, filtering, resampling, and later encoding can all alter waveform shape and inter-sample behavior. FabFilter's Pro-L 2 documentation explains that its own ultra-fast limiting can generate true-peak overshoot unless that is handled. A sample-domain clipper may hold each sample below a chosen ceiling while still yielding a higher true-peak estimate. Conversely, not every inter-sample overshoot becomes audible distortion in every playback chain. The relevance depends on destination, magnitude, and downstream behavior. A true peak is expressed in dBTP, distinct from a sample value expressed in dBFS. The difference is a measurement definition, not a claim that audio mysteriously exceeds every technical limit. One can measure both at the same file and obtain different maxima. The [LUFS, dBFS, and dBTP article](/learn/lufs-vs-dbfs-vs-dbtp) defines those units. Neither peak measure describes integrated loudness or how powerful a song feels. ## Meters estimate a reconstructed signal A true-peak meter oversamples or otherwise reconstructs the signal sufficiently to estimate maxima between original samples under a defined method. ITU-R BS.1770 specifies algorithms and tolerances for true-peak measurement. This is more informative than scanning stored samples alone when the question concerns a reconstructed output. It is still an estimate tied to the method, filter, and implementation. Two compliant meters can differ slightly, and a later transformation may create a new waveform with a different peak. FabFilter's metering help explicitly notes that small dBTP differences between meters can occur. The measurement point matters. A plug-in's meter describes the signal at that plug-in's output under its own processing mode. If another plug-in, fader, codec, or sample-rate converter follows it, the final file can differ. Measuring the exported artifact answers a different question from reading an internal live meter. The same is true for a fixed-point export versus a floating-point internal bus. A true-peak indicator inside one limiter should not be used as a blanket certificate for all downstream paths. True-peak estimation is not the same as turning on a true-peak limiting mode. The meter observes; the limiter acts. Some products let users enable metering separately from the limiting algorithm. FabFilter documents distinct controls for True Peak Limiting and True Peak Metering in Pro-L 2. If only the meter is active, it can reveal overshoot without preventing it. If only the limiting path is enabled, an output measurement is still useful to verify the resulting file. ## What a true-peak limiter tries to do A true-peak limiter detects peaks relevant to the reconstructed output and adjusts the audio so its own final true-peak output remains under the selected ceiling according to its method. FabFilter says Pro-L 2 handles peaks present at its input and attenuates overshoot created by its fast limiting stage. Ableton's Live 12 Limiter includes a True Peak ceiling mode distinct from its Standard and Soft Clip modes. These are product-specific implementations; the shared concept is a ceiling concerned with reconstructed peaks rather than only original samples. The action can change more than a number. A limiter may adjust gain before or around a transient, change release into following material, alter stereo linking, or use other fast-control stages. True-peak modes may add latency and CPU cost. FabFilter notes an additional latency cost for its mode and discusses oversampling as a way to reduce the amount of inter-sample overshoot its true-peak stage must handle. These are engineering tradeoffs, not proof that one mode always sounds better. The [limiter lookahead article](/learn/limiter-lookahead) explains one related time tradeoff. True-peak limiting should not be confused with simply lowering the output fader by a fixed amount. Lowering gain reduces both sample and reconstructed peaks together but does not adapt to peaks that arise relative to a tight ceiling. A true-peak limiter can respond dynamically to particular events. Equally, if no peak approaches the ceiling, a limiter may do little. The useful question is whether a destination requires a maximum and what sonic cost is acceptable to meet it, not whether every song must always be processed through the strongest available mode. ## Oversampling helps, but it is not the definition Oversampling runs internal processing at a higher sample rate and uses filtering around the conversion. In nonlinear or ultra-fast stages, it can reduce aliasing and affect inter-sample behavior. FabFilter documents both benefits and costs for Pro-L 2, including more CPU use and potential filter effects. True-peak limiting, however, is a specific output-control promise under a reconstruction-aware method. An oversampled clipper or limiter does not automatically become a true-peak limiter simply because it processes at a higher internal rate. The [oversampling guide](/learn/clipper-oversampling) covers that separate concept. Similarly, a true-peak limiter need not rely on only one oversampling setting. Implementations can combine prediction, high-rate processing, gain control, and post-stage correction differently. FabFilter's mode description identifies two stages in its own design. The user-facing word *true peak* should be checked against the documented output behavior rather than inferred from a switch labeled oversampling. Higher oversampling is not automatically better for every task because it consumes resources and may alter latency or filtering; verify the actual result. If a very short lookahead forces a limiter to act sharply, it can create more overshoot or distortion. More lookahead may ease some control demands but can change transient impact and add latency. The interaction is implementation-specific. The safe public explanation is causal: the processor needs to prevent reconstructed output peaks by some combination of detection and control, and each design makes tradeoffs. It is not a recommendation for a particular factor, time, or chain. ## Why later stages can change the peak again A lossless file measured immediately after a limiter has a particular sample sequence and true-peak estimate. Lossy encoding creates a different decoded waveform; sample-rate conversion resamples the signal; downstream EQ or gain changes it again. Those operations can produce new maxima. FabFilter explicitly notes possible changes during MP3 conversion in its true-peak discussion. A limiter can validly hold its own output to a documented ceiling while a later file or playback path yields a different measurement. This is a boundary on the claim, not evidence that true-peak limiting is pointless. A specific distribution service may publish true-peak guidance for uploads or encoding headroom. Spotify's current artist documentation includes true-peak tips for lossy quality. Other destinations can use different formats and policies. A technical requirement should be verified from the current destination's official documentation, then checked on the actual final artifact when feasible. Do not turn one platform's number into a universal clipping target. The [streaming loudness targets article](/learn/streaming-loudness-minus-14-lufs) addresses the related playback-reference myth. A true-peak estimate also does not predict every converter perfectly. Real hardware reconstruction filters, headroom, codec paths, and gain settings differ. The ITU method supplies a standardized estimate suitable for specified use, not omniscience about all future devices. State the file format and measurement stage when reporting compliance. A statement about a limiter plug-in's output, a rendered PCM file, and a transcoded stream are three different claims that require their own evidence. ## Peak compliance is not a sonic quality grade A master can meet a strict true-peak ceiling and still sound harsh, flat, or distorted from other causes. Another master can have excellent musical balance but need a small technical adjustment for a particular delivery specification. The ceiling is a constraint, not an artistic endpoint. More limiting to produce a lower maximum can alter attack and sustained energy without improving the song. Conversely, a transparent adjustment that meets a genuine delivery requirement can be entirely appropriate. The [peak-control versus loudness guide](/learn/peak-control-vs-loudness) explains why neither a peak value nor its reduction predicts perceived loudness on its own. A useful validation separates observation from inference. Inspect the final file with a true-peak meter configured for the relevant standard, compare sample peaks, and listen to the passage where the limiter acts at a matched output level. If the output changed after encoding, measure that encoded or decoded version rather than claiming the original limiter failed. If the result sounds worse, consider whether the constraint, processor behavior, or prior mix balance needs attention. There is no universal true-peak mode that makes every music decision for you. The bounded conclusion is clear: true-peak limiting controls a reconstruction-aware maximum at a defined output point. It is stronger than a mere sample ceiling for that question, yet it is implementation dependent and downstream stages can create new peaks. Use it when a specified destination or technical check calls for it, and judge the sonic change separately. ## About G-Clipper Pro G-Clipper Pro shapes samples through a soft-clipping curve. Its output peak should be measured separately when a true-peak delivery ceiling matters. ## Sources & References - [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 - True peak limiting](https://www.fabfilter.com/help/pro-l/using/truepeaklimiting) - [FabFilter Pro-L 2 Help - Oversampling](https://www.fabfilter.com/help/pro-l/using/oversampling) - [FabFilter Pro-L 2 Help - Metering](https://www.fabfilter.com/help/pro-l/using/metering) - [Live Audio Effect Reference — Ableton Reference Manual Version 12](https://www.ableton.com/en/manual/live-audio-effect-reference/) - [Loudness normalization on Spotify](https://support.spotify.com/au/artists/article/loudness-normalization/) ## Continue Reading - [Sample Peak vs True Peak](https://gawergy.com/learn/sample-peak-vs-true-peak) - [What Does Lookahead Do in a Limiter?](https://gawergy.com/learn/limiter-lookahead) - [Streaming Loudness Targets: Do You Actually Need to Master to -14 LUFS?](https://gawergy.com/learn/streaming-loudness-minus-14-lufs) ## Sitemap See the full [Gawergy.com sitemap](https://gawergy.com/sitemap.md).