True Peak vs Sample Peak: Why Your Master Distorts on Phones
You did everything right. Your master's peak meter in your DAW never touches zero. There is headroom. It looks clean. And then someone plays it on their phone, or it comes back from streaming compression, and there is distortion, a harsh, crackly edge that was not there in your session.
You check the meter again. Still under zero. So where is the distortion coming from? The answer is one of the most misunderstood concepts in digital audio, and once you understand it, you will never ship a distorted master again.
Digital audio is made of samples
Digital audio represents a continuous sound wave as a series of individual samples, snapshots of the signal's level taken tens of thousands of times per second. At 44.1 kHz, that is 44,100 samples every second. Your DAW's peak meter looks at these samples and reports the highest one. If no sample exceeds zero dBFS, the meter says you are safe.
But here is the thing: the samples are not the actual sound. They are measurements of it. When the audio is played back, a converter reconstructs the continuous analog waveform that passes through those sample points. And that reconstructed waveform does not just connect the dots in straight lines. It curves smoothly through them, the way the original sound did.
Where the overshoot happens
Because the reconstructed waveform curves through the sample points, it can peak higher between two samples than either sample itself. Picture two adjacent samples, both sitting just under zero. The smooth curve that the converter draws between them can bulge above zero in the gap, even though neither sample did.
That overshoot is called an inter-sample peak, or true peak. It is a real level that occurs during playback, and it can exceed zero dBFS even when every single sample reads below it. Your sample peak meter cannot see it, because it only looks at the samples, not at the reconstructed waveform between them.
When that inter-sample peak exceeds what the playback system or encoder can handle, it clips, and clipping is distortion. This is why a master that looks clean in your DAW distorts on a phone, through cheap converters, or after the lossy compression that streaming platforms apply, which can push inter-sample peaks even higher.
Why streaming makes it worse
Lossy codecs like the ones streaming platforms use to compress audio can actually increase inter-sample peaks. The encoding and decoding process reshapes the waveform slightly, and a master that was already flirting with the ceiling can end up with true peaks well over zero after conversion. So even if you got away with it on your monitors, streaming can introduce the distortion for you.
This is why delivery specifications for streaming almost always call for a true peak ceiling below zero, commonly around -1 dBTP, giving headroom for inter-sample peaks and codec overshoot.
The fix: measure true peak with Decibel
The only way to catch inter-sample peaks is with a true peak meter, which oversamples the signal to reconstruct the waveform between the samples and reports the actual peak level, not just the highest sample.
Decibel includes true peak metering with a Max validator that lights up red the moment your true peak exceeds the ceiling you set. Instead of trusting your DAW's sample meter and hoping, you can see the real peak level of the reconstructed waveform and know for certain whether your master will clip on playback.
Decibel's Super Meter shows loudness, dynamics, and true peak together in one module, so you manage all three at once. It is compliant with the major broadcast standards, so whatever specification you are delivering to, you can meet its true peak requirement precisely. You can drop a finished file onto Decibel for instant offline analysis and read its true peak in seconds, and you can stream the meters to phones and tablets to keep them in your eyeline.
Step by step: a master that never clips on playback
- Put Decibel on your master bus and enable true peak metering.
- Set your true peak ceiling. For streaming, many engineers use -1 dBTP to leave room for inter-sample peaks and codec overshoot.
- Play the whole track through and watch the true peak reading. If the Max validator goes red, your master will clip on playback.
- If you exceed the ceiling, lower your limiter's output ceiling, or use a true-peak-aware limiter, until the reading stays under your target.
- Confirm with offline analysis by dropping the bounced file onto Decibel to read its final true peak before you deliver.
FAQ
What is the difference between true peak and sample peak? Sample peak is the highest individual sample value, which is what a standard DAW meter shows. True peak is the highest level of the continuous waveform reconstructed during playback, which can rise between samples, called an inter-sample peak. True peak can exceed zero even when every sample reads below it.
Why does my master distort on phones when my DAW says it's under 0 dB? Because your DAW's sample meter only measures the samples, not the waveform reconstructed between them during playback. Inter-sample peaks can exceed zero and clip on the playback device or after streaming compression, causing distortion your sample meter never showed. A true peak meter catches this.
What true peak ceiling should I use for streaming? Many engineers use -1 dBTP for streaming, which leaves headroom for inter-sample peaks and for the overshoot that lossy codecs can introduce during encoding. Some platforms and stricter specifications call for lower ceilings, so check your delivery target.
How do I measure true peak? Use a true peak meter, which oversamples the signal to reconstruct the waveform between samples and report the actual peak. Decibel includes true peak metering with a validator that alerts you when you exceed your ceiling, so you can catch clipping before delivery.
Stop shipping distortion you can't hear
Put Decibel on your master, set a true peak ceiling, and deliver a master that stays clean on every device and every platform.