What Sample Peak Meters Actually Measure

A standard peak meter in your DAW reads the highest absolute value among the digital samples in your audio file. At 44.1 kHz, there are 44,100 discrete samples per second per channel. Your peak meter checks each one and reports the largest value it finds.

This seems thorough, but it is fundamentally incomplete. Digital audio is a series of discrete snapshots. The actual sound wave is continuous. When a digital-to-analog converter reconstructs the continuous waveform from those snapshots, the analog signal between two samples can be higher than either sample value. Your peak meter never sees this because it only looks at the snapshots, not the reconstructed wave.

This gap between what your meter reads and what actually comes out of the converter is where intersample peaks hide.

Intersample Peaks: The Hidden Distortion

An intersample peak (ISP) occurs when the reconstructed analog waveform between two adjacent samples exceeds the value of either sample. Imagine two consecutive samples at -0.5 dBFS and -0.3 dBFS. The sine wave curve connecting them can easily peak above 0 dBFS in between, even though neither individual sample clips.

ISPs are most common in material with dense high-frequency content: bright cymbals, distorted guitars, heavily limited masters, and anything with lots of energy near the Nyquist frequency. The more aggressive your limiting, the more likely you are creating intersample peaks, because the limiter is catching sample peaks but not the reconstructed waveform between them.

How bad can it get? Intersample peaks can exceed the sample peak by up to 3 dB in extreme cases. A master that reads -0.1 dBFS on a sample peak meter could have true peaks at +2.9 dBFS, causing hard clipping in any DAC or codec that encounters it.

The distortion from ISP clipping is particularly unpleasant because it occurs at the output stage, after all your careful processing. It introduces harsh, digital-sounding artifacts that were never part of your mix or mastering chain.

True Peak vs Sample Peak: The Critical Difference

A true peak meter does not just read sample values. It oversamples the audio, typically by 4x, to estimate the continuous waveform between samples. By interpolating between the original samples, the meter can detect peaks in the reconstructed signal that a sample peak meter misses entirely.

The measurement unit for true peak is dBTP (decibels True Peak), defined in the ITU-R BS.1770 standard. A reading of -1 dBTP means the estimated true peak of the reconstructed analog signal is 1 dB below full scale.

The difference between sample peak and true peak can be surprisingly large:

Why It Matters for Streaming and Encoding

When you upload a master to a distributor, your file gets transcoded to lossy formats: AAC for Apple Music, Ogg Vorbis for Spotify, Opus for YouTube. These codecs decode to a slightly different waveform than your original. The encoding and decoding process can shift peaks, sometimes pushing them higher than the original.

If your master is already at 0 dBFS sample peak with intersample peaks above that, the transcoded version will clip even harder. The clipping occurs on the listener's device, producing audible distortion that you never heard in your studio. This is why every major streaming platform and broadcast standard requires true peak measurement, not sample peak.

Apple's Mastered for iTunes specification explicitly requires true peak measurement and recommends a ceiling well below 0 dBTP. Spotify, Tidal, and YouTube all normalize to specific LUFS targets and expect masters to be free of true peak clipping.

The -1 dBTP Standard and Why It Exists

The widely adopted mastering standard is a true peak ceiling of -1 dBTP. This provides 1 dB of headroom above the highest point of the reconstructed analog waveform, accounting for the additional peak shifts introduced by lossy codec transcoding.

Some engineers go further:

There is no benefit to mastering with a true peak ceiling of 0 dBTP. You gain a fraction of a dB of loudness at the cost of guaranteed clipping on at least some playback systems. The -1 dBTP standard exists because the math demands it.

How to Fix It: True Peak Limiters in Practice

A true peak limiter works like a standard brick-wall limiter but with one critical addition: it oversamples the input signal to detect intersample peaks and applies gain reduction to prevent the reconstructed waveform from exceeding the ceiling. The limiter catches peaks that a standard limiter cannot see.

Here is how to use one correctly in your mastering chain:

  1. Place it last in your chain. The true peak limiter should be the final processor before dithering. Any processing after the limiter can reintroduce peaks above the ceiling.
  2. Set the ceiling to -1 dBTP. Not -1 dBFS. Make sure your limiter is operating in true peak mode, not sample peak mode. Many limiters have a toggle for this.
  3. Use a look-ahead setting of 1 to 5 ms. Look-ahead allows the limiter to anticipate transients and apply gain reduction before the peak arrives, resulting in cleaner, less distorted limiting.
  4. Monitor with a true peak meter. Verify that your output never exceeds -1 dBTP. If your limiter is working correctly, the meter should confirm it.
  5. Do not push excessive gain into the limiter. If you are seeing more than 3 to 4 dB of constant gain reduction, you are asking the limiter to do too much. The result will sound squashed and distorted regardless of the peak ceiling. Back off the input gain or address the problem in the mix stage.

LuvLang uses true peak limiting as a standard part of its mastering chain, with real-time true peak metering displayed alongside LUFS measurement. Every export is guaranteed to respect the -1 dBTP ceiling, so your masters are safe for every streaming platform, every codec, and every playback device without any manual configuration.