/truː piːk/
True Peak is a measurement of the highest amplitude an audio signal reaches after digital-to-analog conversion, capturing inter-sample peaks that occur between sample points. Standard peak meters can miss these peaks, causing distortion on playback devices. True Peak meters use oversampling to reveal the actual reconstructed waveform maximum.
Your limiter says −0.1 dBFS. Your master clips on every phone on earth. True Peak is the measurement that finally tells you the truth.
True Peak (often abbreviated TP) is a measurement standard that quantifies the maximum instantaneous amplitude a digital audio signal will reach once it is reconstructed into a continuous analog waveform. Unlike a conventional sample-peak meter — which measures only the amplitude of discrete sample points — a True Peak meter employs oversampling, typically at 4× or higher, to model the continuous signal that a digital-to-analog converter (DAC) will produce between those sample points. The difference between what a standard meter displays and what actually emerges from a DAC is not theoretical; it is audible distortion that manifests on consumer playback hardware, streaming delivery pipelines, and any downstream codec that performs its own level normalization.
The conceptual core of True Peak is the inter-sample peak (ISP). When a digital audio file is played back, the DAC does not simply output a staircase of individual sample values — it reconstructs a smooth, continuous waveform via a reconstruction filter (sinc interpolation). This reconstruction can produce amplitude values that exceed any individual sample in the digital file, sometimes by as much as 3–4 dB. A track that measures −0.3 dBFS on a conventional peak meter may reconstruct to +1.2 dBTP or higher. Every streaming platform that applies loudness normalization — Spotify, Apple Music, YouTube, TIDAL, Amazon Music — measures True Peak at the delivery stage, and many will clip, reject, or re-process masters that exceed their stated True Peak ceiling before normalization is applied.
True Peak measurement is codified in the ITU-R BS.1770 standard, the same international specification that defines integrated loudness in LUFS. The standard specifies that True Peak meters must use a minimum of 4× oversampling and must conform to defined filter characteristics to ensure consistent readings across metering implementations. This standardization means that a compliant True Peak reading in your mastering DAW should correspond to the measurement Spotify's ingestion pipeline produces — provided your metering plugin is genuinely BS.1770-compliant and not simply marketing itself as such. Producers and mastering engineers should verify compliance by checking their meter against reference tones rather than trusting plugin branding alone.
In practical session work, True Peak is most critical at two stages: the final limiter ceiling on a master, and the export or bounce settings of any file destined for streaming, broadcast, or further processing. The EBU R128 broadcast standard recommends a True Peak ceiling of −1.0 dBTP. Most streaming platform delivery specifications — including those of Spotify, Apple Music, and Amazon Music — recommend or require a True Peak maximum of −1.0 dBTP, with some pipelines accepting −0.5 dBTP. The extra headroom below 0 dBFS absorbs the inter-sample reconstruction peaks and protects against codec-induced gain changes during MP3, AAC, or Opus encoding, which can themselves introduce additional amplitude overshoot of 0.5–2 dB.
Understanding True Peak shifts how a producer thinks about the entire gain structure of a mix. It is not enough to leave −0.3 dB of headroom before the digital ceiling. A mix with dense, transient-rich content — tightly limited drums, saturated 808s, heavily compressed vocals — will exhibit more severe inter-sample peaking than a sparse arrangement, because the reconstruction filter interacts with closely spaced high-amplitude samples in ways that amplify overshoot. Monitoring True Peak from the mix stage, not just at mastering, gives producers earlier warning of content that will cause problems downstream, and informs decisions about saturation, limiting aggressiveness, and stereo bus processing that are far easier to revisit in a mix session than in a mastering chain.
At the signal level, True Peak measurement works by upsampling the digital audio stream before peak detection. A BS.1770-compliant meter takes the incoming PCM signal — typically 44.1 kHz or 48 kHz — and applies a polyphase interpolation filter to produce a version at 4× the original sample rate (176.4 kHz or 192 kHz). This upsampled signal more closely approximates the continuous waveform a DAC will reconstruct, because it fills in amplitude values between the original sample points. The meter then tracks the absolute peak of this oversampled signal and reports it in dBTP (decibels True Peak). Some high-quality implementations use 8× oversampling for additional accuracy, particularly relevant for material that has already undergone format conversion or codec processing.
The physics that create inter-sample peaks are rooted in the Nyquist-Shannon sampling theorem. When a signal is sampled at 44.1 kHz, frequencies up to 22.05 kHz can be represented accurately. However, the reconstruction filter required to recover those frequencies from discrete samples is a brick-wall low-pass filter that, by its nature, introduces Gibbs phenomenon ringing around sharp transients. This ringing can constructively add to adjacent samples and push reconstructed amplitude above any individual sample value. The effect is most pronounced when the signal contains energy close to the Nyquist frequency — high-frequency transients, overdriven saturation harmonics, or heavily compressed broadband content all exhibit elevated inter-sample overshoot. This is why limiting a master hard to 0 dBFS sample peak is insufficient: the reconstruction filter has not yet had its say.
Codec encoding compounds the problem. AAC and MP3 encoders operate on frequency-domain representations of audio (modified discrete cosine transform, MDCT) and, during quantization and reconstruction, can alter the time-domain peak amplitude of the signal. Studies by researchers at the Audio Engineering Society (AES) and by streaming platforms' own engineering teams have demonstrated that AAC encoding at 256 kbps can introduce up to 2 dB of additional peak overshoot relative to the input PCM file. If the input PCM already has a True Peak of −0.1 dBTP, the AAC output may reconstruct to +1.9 dBTP — audible clipping on any device that plays back without digital headroom to absorb the gain increase. This is the precise reason streaming specifications mandate −1.0 dBTP or lower on submitted masters.
The relationship between True Peak and integrated LUFS is not fixed. A signal can have a low LUFS value (quiet, sparse content) while still exhibiting high True Peak values (a single loud transient), or it can be dense and loud in LUFS while every individual sample remains well below 0 dBFS. True Peak and LUFS are orthogonal measurements that serve complementary purposes: LUFS governs perceived loudness over time; True Peak governs instantaneous amplitude safety. Both must be managed simultaneously in a compliant master. Setting your limiter ceiling to −1.0 dBTP, then measuring integrated loudness to match your target platform, is the correct sequence — attempting to hit a LUFS target first and then capping peaks as an afterthought often results in limiting artifacts because the loudness and peak constraints are being resolved in the wrong order.
In summary, True Peak measurement is the bridge between the integer-quantized world of PCM audio and the continuous analog waveform that listeners actually hear. Every digital audio workflow that terminates in consumer playback — streaming, broadcast, DVD, Blu-ray, podcast distribution — operates in an environment where the reconstruction filter is not optional and the codec is not transparent. True Peak metering makes the behavior of both visible before a file leaves your session, giving you actionable data rather than a false confidence that your peak meter has already provided.
Diagram — True Peak: Diagram comparing sample-peak metering versus True Peak metering, showing inter-sample overshoot above 0 dBFS after DAC reconstruction.
Every true peak — hardware or plugin — operates on the same core parameters. Know these and you can work with any implementation.
Set in your limiter or export settings, this value defines the highest True Peak level the processed signal may reach after oversampled reconstruction. For streaming delivery, −1.0 dBTP is the near-universal standard; for broadcast (EBU R128), −1.0 dBTP is mandated. For intermediate mix bounces that will undergo further processing, −3.0 dBTP or lower is safer.
Determines how many interpolated points between samples are evaluated when calculating True Peak. BS.1770 mandates a minimum of 4× (e.g., 176.4 kHz when the source is 44.1 kHz). Higher factors (8×, 16×) yield marginally more accurate readings, particularly on material that has already been through codec encoding or sample-rate conversion. Most professional metering plugins default to 4× for CPU efficiency while remaining spec-compliant.
True Peak meters typically display both a live reading and a held maximum. Hold time determines how long the highest detected True Peak value remains visible before the display resets or begins to fall. In mastering contexts, infinite hold is preferred so that the session's worst-case True Peak is always visible without requiring real-time monitoring. Short hold times (1–3 seconds) suit mixing contexts where you need ongoing feedback rather than a session maximum.
Many metering tools include a dedicated ISP indicator — typically a colored LED or text flag — that lights when the signal's True Peak exceeds the sample peak, confirming that inter-sample overshoot is occurring. This is distinct from a clipping indicator, which fires at 0 dBFS sample level. An ISP warning at −0.5 dBTP on a mix bus is actionable information: it means the reconstruction filter is adding ≥0.5 dB of amplitude that your standard meters are not reporting.
Some meters offer a choice between instantaneous True Peak (updates every few milliseconds) and short-term True Peak averaged over a defined window. Instantaneous mode is essential for mastering compliance checks; windowed averaging is more useful in mixing for tracking trends in dynamic content without reacting to single-sample spikes that may not be perceptually significant. For export and delivery purposes, always switch to instantaneous mode to confirm absolute ceiling compliance.
Advanced metering plugins such as the Nugen Audio MasterCheck and iZotope RX's codec preview module offer a codec simulation mode that models the True Peak overshoot introduced by AAC 256 kbps or MP3 320 kbps encoding. This gives the mastering engineer a delivery-stage True Peak prediction before the file leaves the session. A common finding is that a master at −1.0 dBTP pre-codec may reach −0.3 to +0.5 dBTP post-AAC, reinforcing the case for a −1.5 dBTP or −2.0 dBTP safety margin on very transient-heavy material.
Session-ready starting points. All dBTP values reference ITU-R BS.1770 compliant True Peak measurement at 4× oversampling minimum; apply post-limiting before final export.
| Parameter | General | Drums | Vocals | Bass / Keys | Bus / Master |
|---|---|---|---|---|---|
| Streaming delivery ceiling | −1.0 dBTP | −1.0 dBTP | −1.0 dBTP | −1.0 dBTP | −1.0 dBTP |
| Broadcast (EBU R128) ceiling | −1.0 dBTP | −1.0 dBTP | −1.0 dBTP | −1.0 dBTP | −1.0 dBTP |
| Pre-mastering mix bounce | −3.0 dBTP | −3.0 dBTP | −3.0 dBTP | −3.0 dBTP | −3.0 dBTP |
| Typical ISP overshoot (limited pop master) | 0.5–1.5 dB | 1.0–3.0 dB | 0.3–1.0 dB | 0.5–2.0 dB | 0.5–2.0 dB |
| Post-AAC 256 kbps additional overshoot | 0.5–1.0 dB | 0.5–2.0 dB | 0.3–0.8 dB | 0.5–1.5 dB | 0.5–2.0 dB |
| Safe limiter ceiling (allows for codec headroom) | −1.5 dBTP | −2.0 dBTP | −1.5 dBTP | −1.5 dBTP | −1.5 dBTP |
| Minimum metering oversample rate | 4× | 4× | 4× | 4× | 4× |
All dBTP values reference ITU-R BS.1770 compliant True Peak measurement at 4× oversampling minimum; apply post-limiting before final export.
The concept of inter-sample clipping was documented in the engineering literature well before it became a practical concern for commercial music production. In the early CD era — 1982 onward — mastering engineers working with 16-bit 44.1 kHz audio occasionally noted that discs which metered cleanly at 0 dBFS sample peak nonetheless caused audible distortion on certain consumer players. Bob Ludwig, one of the defining mastering engineers of the compact disc era, has described in interviews observing early pressing plant feedback about discs that exhibited harshness on budget players not present on higher-end hardware. The root cause — that consumer DAC reconstruction filters were producing amplitude overshoot from high-sample-density material — was understood at a theoretical level by engineers familiar with sampling theory but had no standardized measurement framework.
The formal codification of True Peak measurement arrived with the International Telecommunication Union's publication of ITU-R BS.1770 in 2006. The standard was developed to provide broadcasters with a unified, consistent method for measuring audio loudness and peak levels across digital broadcast chains — a pressing need as digital television and radio rollouts accelerated in Europe, North America, and Asia. BS.1770 defined both an integrated loudness measurement (which would later be calibrated in LUFS by the EBU in R128, published 2011) and a True Peak measurement methodology specifying oversampled peak detection. The European Broadcasting Union's adoption and refinement of the standard in EBU R128 (2011) and its subsequent updates (R128 S1 for streaming, 2014) transformed True Peak from a broadcast engineering specification into the de facto standard for all professional audio delivery.
The widespread adoption of True Peak in commercial music mastering accelerated between 2013 and 2017 as streaming platforms formalized their loudness normalization and delivery specifications. Spotify published loudness normalization details (targeting approximately −14 LUFS integrated, −1.0 dBTP maximum) and Apple introduced Sound Check normalization targeting −16 LUFS for iTunes-delivered content. iZotope's Ozone 6 (released 2014) was among the first widely adopted mastering suites to integrate a True Peak-compliant limiter ceiling control directly into its workflow, rather than treating True Peak metering as a separate analytical tool. FabFilter released the Pro-L 2 limiter in 2017 with explicit True Peak limiting mode — a dedicated oversampled limiting algorithm that prevents the output from exceeding the set dBTP ceiling at the reconstruction stage, not just at the sample level — which became a reference tool in mastering studios globally.
The AES (Audio Engineering Society) formalized additional research on codec-induced True Peak overshoot through papers published between 2012 and 2018, particularly work by researchers including Florian Camerer (Austrian Broadcasting Corporation, ORF) and the EBU's PLOUD group, who demonstrated empirically that AAC encoding consistently added 0.5–2.0 dB of True Peak amplitude relative to the input PCM. This research directly influenced streaming platform delivery specifications and the incorporation of codec simulation into metering tools. By 2020, True Peak compliance had become a baseline expectation for any professional mastering delivery, with distributors such as DistroKid, TuneCore, and CD Baby all publishing True Peak ceiling requirements in their technical specifications for submitted masters.
In a mastering session, True Peak governs the ceiling setting of the final limiter or maximizer. The standard workflow places a True Peak-compliant limiter — such as FabFilter Pro-L 2 with True Peak mode enabled, iZotope Ozone Maximizer with the True Peak checkbox active, or Waves L2/L3 at conservative levels — at the end of the master chain, with its output ceiling set to −1.0 dBTP. A BS.1770-compliant meter (iZotope Insight 2, Nugen Audio VisLM, TC Electronic LoudnessPilot, Youlean Loudness Meter 2) sits after the limiter on the monitor chain to confirm that no signal is breaching the ceiling in real time and to hold the session maximum for final compliance verification. This is not a single check at export time; the metering runs throughout the entire master playback so that any transient event — a snare hit, a cymbal wash, an 808 sub transient — that triggers inter-sample overshoot is caught and addressed during the session.
For drum-heavy music — trap, drum and bass, metal, pop with heavy programmed drums — True Peak management is particularly demanding. Snare transients, cymbal crashes, and clap layerings contain dense high-frequency energy that interacts aggressively with reconstruction filters. A trap master at −8 LUFS with a sample peak of −0.2 dBFS may reconstruct to +1.5 dBTP or higher without True Peak limiting. The mitigation is twofold: apply transient shaping or high-frequency saturation before the limiter to reduce the sharpness of the most problematic peaks (which reduces ISP without necessarily increasing sample peak level), and engage True Peak mode on the limiter so the output is oversampled-constrained rather than sample-constrained. Expect that True Peak mode will reduce perceived loudness by 0.5–1.5 dB compared to a sample-peak-limited master at the same ceiling, because the limiter must work harder to enforce a lower effective ceiling.
Bass-heavy music — where 808 sub-bass, synth bass, or bass guitar occupies significant low-frequency headroom — creates True Peak challenges at the opposite end of the spectrum. Sub-bass waveforms with steep attack phases (especially pitch-bent 808s that sweep rapidly) can cause inter-sample overshoot in the low-frequency reconstruction, though the magnitude is typically less than for high-frequency transients. More practically, sub-bass energy uses up so much headroom that after True Peak limiting at −1.0 dBTP, the loudness ceiling achievable in LUFS terms may be lower than genre expectations. Engineers working on bass-heavy music sometimes apply gentle multiband limiting or dynamic EQ on the sub-bass (below 60 Hz) before the final limiter to reclaim headroom, allowing the limiter to reach the LUFS target without being forced into aggressive True Peak action on low-frequency peaks.
For vocal-led music — singer-songwriter, acoustic pop, podcast production — inter-sample peaking is less severe because the content has less extreme transient density and less high-frequency saturation. However, True Peak compliance remains essential because streaming normalization will be applied regardless of genre. A vocal master at −14 LUFS with a True Peak of −0.8 dBTP is borderline; Spotify's normalization headroom calculation may encounter edge cases. The safe practice is −1.5 dBTP on vocal-primary content to provide margin for any additional processing the platform applies. Notably, podcast and spoken-word content distributed through Apple Podcasts is subject to EBU R128 broadcast standards when the platform applies normalization, meaning the same −1.0 dBTP rule applies outside of music contexts as well.
One email a week. The techniques behind the terms — curated by working producers, not algorithms.
Abstract knowledge becomes practical when you can hear it in music you know. These tracks demonstrate true peak used intentionally, at specific moments, for specific purposes.
The punishing sub-bass that opens 'bad guy' is a textbook case of bass-heavy content that stresses True Peak budgets. The 808-style bass transient attacks are steep enough to generate significant inter-sample overshoot even at the relatively modest integrated loudness of the master (~−14 LUFS on Spotify's normalized stream). Mastering engineer John Greenham's work on this track demonstrates that a restrained loudness target allows room for clean True Peak management without visible limiting artifacts. Listen on a mobile phone without EQ boost — the sub hits cleanly without harshness, evidence of careful True Peak ceiling management at delivery.
The opening snare crack in 'HUMBLE.' is one of the most-cited examples in mastering forums for inter-sample peak analysis. The snare transient is extremely sharp, with minimal attack smoothing, concentrated energy between 2–8 kHz — exactly the frequency range where reconstruction-filter ISP is most severe. At the track's commercial loudness (approximately −8 LUFS integrated on the album version), this type of transient demands True Peak limiting rather than sample-peak limiting to avoid distortion on consumer playback. Compare the streaming version with a loud reference monitor decode to hear how the mastering chain has controlled the peak without dulling the crack.
Released at the cusp of the streaming normalization era, 'Get Lucky' illustrates a well-managed True Peak on a dense, mid-loudness pop master. The rhythm guitar, bass, and drum combination is rhythmically dense without being transient-extreme, making ISP management more tractable than on purely electronic content. The master sits comfortably within −1.0 dBTP on modern streaming analysis tools, representing mastering engineer Florian Lagarde's careful attention to peak compliance on what became one of the most-played streaming tracks of its era. Use this as a reference track in your metering setup to calibrate expectations for commercial pop True Peak behavior.
The distorted, saturated vocal processing and abrasive synthesizer stabs in this track create exactly the kind of high-frequency saturation content that generates elevated ISP readings. Despite the aggressive sound design, Greenham's master maintains True Peak compliance — listen for the way the distorted elements retain their textural aggression without audible inter-sample crackle on earbuds, which would be the tell-tale sign of an uncontrolled True Peak ceiling. This track is useful as a stress test for your True Peak limiter: run it through your mastering chain and compare your TP meter reading to the published delivery master.
Dedicated hardware loudness and True Peak meters, used primarily in broadcast post-production facilities, compute True Peak in real time via dedicated DSP and display both integrated LUFS and TP ceiling compliance on front-panel displays. The TC Electronic LM2 is the industry standard for EBU R128 broadcast compliance checking. Hardware metering eliminates the latency and plugin-load concerns of DAW-based metering when monitoring live broadcast chains.
Software plugins that implement BS.1770-compliant True Peak measurement within a DAW session. These range from free tools (Youlean Loudness Meter 2 free tier) to professional broadcast-grade solutions (Nugen Audio VisLM, priced for commercial post-production). For music production, iZotope Insight 2 is widely adopted for its comprehensive display of True Peak, LUFS, and dynamic range alongside spectral analysis, all in a single insert.
A category of limiter that operates with oversampled internal processing — typically 4× to 32× — so that the gain reduction algorithm responds to the reconstructed waveform, not just the sample values. Enabling True Peak mode in FabFilter Pro-L 2 or iZotope Ozone Maximizer means the limiter will reduce gain whenever the oversampled signal would exceed the set ceiling, preventing ISP from breaching the dBTP limit at delivery. The Weiss DS1-MK3 hardware mastering limiter is considered the reference for transparent True Peak limiting on high-end mastering chains.
A specialized analysis mode that simulates the amplitude changes introduced by AAC or MP3 encoding and displays the estimated post-codec True Peak level. These tools allow mastering engineers to predict whether a master that complies at −1.0 dBTP in PCM form will still comply after AAC 256 kbps encoding by Spotify or Apple Music. Given that codec processing routinely adds 0.5–1.5 dB of True Peak overshoot, codec-aware True Peak analysis has become an important final check in professional mastering workflows.
Broadcast-specific True Peak implementations are governed strictly by ITU-R BS.1770-4 and EBU R128, with additional requirements for loudness range (LRA) and maximum True Peak enforced by playout systems before transmission. Jünger Audio's Level Magic system performs real-time loudness and True Peak normalization in broadcast playout chains, ensuring that submitted content meets regulatory requirements before it reaches the viewer. Mastering for broadcast requires understanding both the −1.0 dBTP peak ceiling and the specific LUFS target of the receiving broadcaster.
These MPW articles put true peak into practice — specific techniques, real tools, and applied workflows.