What peak level measures
A peak meter follows the highest sample values in a signal and reports how close they come to the digital ceiling. In a dBFS scale, 0 dBFS is the maximum representable sample level. Values below it are negative, so a peak at −6 dBFS has more headroom than a peak at −1 dBFS.
Peak level is not the same as average loudness. A short snare transient can reach a high peak while the rest of the track remains moderate. Conversely, a heavily compressed signal may have lower transients but sound continuously loud because its average energy is high.
Sample Peak and True Peak
A standard sample-peak meter reads the values stored in the digital file. True Peak estimation reconstructs the waveform between samples and can reveal inter-sample peaks that appear during digital-to-analog conversion or lossy encoding.
This distinction matters near the ceiling. A file that peaks at −0.1 dBFS on a sample meter can still exceed 0 dBTP after conversion. For final masters, podcasts and encoded delivery, inspect both sample peak and True Peak instead of relying on only one meter.
Peak level, headroom and clipping
Headroom is the distance between the current peak and the maximum level. If the highest sample is −8 dBFS, there are 8 dB of sample headroom. This space protects against unexpected transients and leaves room for EQ, compression, saturation and summing during mixing.
Digital clipping occurs when processing tries to push samples beyond the maximum available value. The waveform is flattened or otherwise distorted, and lowering the volume afterward does not restore the missing shape. Preventing clipping is easier and cleaner than repairing it.
Useful targets in recording and mixing
There is no universal peak target for every source. During recording, healthy peaks around −18 to −10 dBFS are often practical because they leave safety margin without creating a noise problem in modern equipment. Very dynamic sources may need even more room.
During mixing, the master bus does not need to approach 0 dBFS. Leaving several decibels of headroom makes later processing easier. A mastering engineer can increase level cleanly; removing accidental clipping or excessive limiting is much harder.
Peak level in mastering and export
At the final stage, the appropriate ceiling depends on the destination and codec. Lossy formats can create new peaks, so a small True Peak margin is useful. Streaming services may normalize loudness, but they do not repair clipping already present in the file.
Always measure the exported file, not only the live session. Resampling, dithering, limiting and encoding can slightly change peak values. Reopen the final WAV, MP3, AAC or OGG and verify it with the same analysis tools used for quality control.
How to analyze peaks correctly
Analyze the complete file, then inspect the loudest sections and isolated transients. Compare the waveform with peak, True Peak, RMS and LUFS readings. If a single click creates the highest peak, solve the click rather than reducing the whole song unnecessarily.
Match playback loudness when comparing processed and unprocessed versions. A louder result often seems better at first, even when transients, stereo image or clarity have been damaged. Objective measurements are most useful when paired with level-matched listening.
Common mistakes and checklist
Do not confuse peak level with perceived loudness. Do not record unnecessarily close to 0 dBFS, and do not assume that normalization fixes clipped audio. Avoid setting a final ceiling without checking True Peak after encoding.
Before delivery, confirm that no unwanted clipping is present, the loudest transient is intentional, sufficient headroom remains for the current production stage, and the final exported file has been measured from beginning to end.