How is the waveform drawn?

Each horizontal pixel represents a bucket of samples. The tool finds the minimum and maximum sample in that bucket and draws a vertical line between them — the standard peak-waveform rendering used by audio editors.

What does the red colour mean?

Red marks pixel columns that contain clipped samples — audio that has reached or exceeded full scale (±1.0). Clipping causes audible distortion, so red regions are worth re-checking against your source.

What is the peak dBFS value?

It is the loudest sample in the whole file expressed in decibels relative to full scale, computed as 20·log10(peak). A value near 0 dBFS means the audio is close to clipping; very negative values mean it is quiet.

No. The file is read into an ArrayBuffer and decoded locally with AudioContext. Nothing is sent to any server, so the viewer works offline and keeps your audio private.

Why is the waveform mono even for stereo files?

For display, the channels are averaged into a single mono waveform so the whole file fits one view. Peak level and clipping statistics still consider every channel separately.

Audio Waveform Viewer

See your audio as a waveform, right in the browser

A waveform viewer turns an audio file into the familiar shape that audio editors show: amplitude over time. It is handy for spotting silent gaps, fades, transients, and — importantly — clipping, all without installing software or uploading your file anywhere.

How it works

The file is decoded to raw PCM with AudioContext.decodeAudioData, then the channels are mixed to a single mono Float32Array. To draw, the Canvas width is divided across the samples so each pixel column covers a fixed bucket. For that bucket the tool computes the minimum and maximum sample value and draws a vertical line between them — this is a peak (min/max) waveform, the same approach used by DAWs.

Clipping is detected when any sample in a column reaches ±0.999 or beyond; those columns are drawn in red. The overall peak level is reported in dBFS using 20 · log10(peak), where the loudest possible sample is 0 dBFS.

Tips and notes

Use the zoom slider to expand the start of the file; it makes count-in clicks, pops, or DC offset easy to see.
Red columns indicate clipping. If you see them, reduce gain on the source and re-export before mastering or publishing.
A peak well below 0 dBFS (for example -6 dBFS) leaves headroom — useful before applying further processing.
Because everything runs locally, large files consume browser memory; trim very long recordings if the page slows down.