What Is dBTP? True Peak in Plain English
Published · updated · by Loopin
dBTP stands for decibels relative to full scale, true peak. It answers a question ordinary peak meters cannot: how loud does this file actually get once it becomes sound again?
Digital audio is dots, not a line
A digital file stores a series of samples — snapshots of the waveform taken tens of thousands of times a second. Your peak meter reads those snapshots. But playback does not output dots; a converter draws a smooth curve through them and sends that to your speakers.
That reconstructed curve can rise higher than any of the individual samples it passes through. The peak lives between the dots. A sample peak meter reads 0.0 dBFS and is telling the truth about the file, while the actual sound hits +0.8 dB and clips.
How true peak is measured
A true-peak meter upsamples the audio — commonly four times or more — reconstructing the signal between the existing samples and measuring the result. That is the number reported in dBTP.
Resolution matters here. Cheap true-peak detection using a short filter under-reads on dense, bright material, sometimes by a decibel or more, which is exactly the material most likely to have the problem in the first place.
Why −1 dBTP is the standard advice
Two reasons. First, inter-sample peaks: leaving a decibel of room means the reconstructed waveform still has somewhere to go. Second, lossy encoding — the MP3 or AAC that streaming platforms actually deliver is not a bit-perfect copy, and the encoding process routinely shifts peaks upward.
A master sitting at 0 dBFS will often come back from an encoder above full scale. The listener hears distortion that was never in your file, on a device you will never test on. One decibel of headroom is cheap insurance, and normalisation means it costs you nothing in perceived loudness.
When to go lower than −1
Amazon Music asks for −2 dBTP. Very dense, heavily limited material benefits from −1.5 or −2 because the denser the audio, the larger the inter-sample overshoots tend to be.
If you are delivering to a specific platform with a published spec, use their number. If you are delivering everywhere at once, −1 dBTP is the safe default and −1.5 is a reasonable belt-and-braces choice for loud genres.
How to check it
Any loudness meter worth using reports true peak alongside integrated LUFS. Measure the finished, exported file rather than the output of your session — dither and format conversion happen after your meter in the chain.
If your export reads above your ceiling, do not just turn the whole file down; that changes your loudness too. Lower the limiter’s ceiling and re-render so the limiter itself catches the overshoots.
Frequently asked questions
What is the difference between dBFS and dBTP?
dBFS measures the highest individual sample in the file. dBTP measures the highest point the reconstructed waveform reaches during playback, including peaks that fall between samples. dBTP is always equal to or higher than dBFS.
Is -1 dBTP always enough?
For most material, yes. Very dense or heavily limited tracks can benefit from -1.5 or -2 dBTP because dense audio produces larger inter-sample overshoots, and Amazon Music specifically asks for -2 dBTP.
Can I just turn my master down to fix true peak?
It works but it also lowers your loudness by the same amount. Better to lower the limiter ceiling and re-render, so the limiter catches the overshoots while your integrated loudness stays where you set it.