Oscilloscope - Real-Time Audio Waveform Visualizer

How to use this tool

Choose Time/div, Amplitude/div, and Rising, Falling, or Auto (off) trigger settings.
Click "Start Oscilloscope", allow microphone access, then speak or play a sound.
Read the 10×8 grid and sample-rate/window readout; use Auto if you want an untriggered trace.
Click "Stop Oscilloscope" to release the microphone.

Reading the live waveform

This oscilloscope plots microphone amplitude vertically against time horizontally. The browser's Web Audio API supplies a fresh block of time-domain samples, and the canvas draws those samples as a line. The 10-by-8 grid helps compare timing and relative amplitude while you speak, whistle, test a speaker, or inspect a steady musical note. A sine tone looks smooth; clipped or buzzy sources often have flatter peaks, and speech or room noise produces a more irregular trace.

Time and amplitude scales

Choose Time/div to set the horizontal window. Ten divisions at 1 ms/div show about 10 ms; at a 48 kHz sample rate that is 480 samples. Ten divisions at 10 ms/div show about 100 ms, or 4,800 samples at 48 kHz. The readout reports the selected window, sample count, analyser buffer size, and actual sample rate. The buffer is rounded up to a supported power of two, so the displayed window can use only part of it.

Amplitude/div controls vertical zoom in normalized digital amplitude units. At the default 0.25 per division, the grid spans approximately −1 to +1, which is near the full range returned by many microphone inputs. At 0.5 per division the grid spans −2 to +2, so a full-scale signal reaches about halfway from the center to an edge. Narrower settings can clip the trace; wider settings show it smaller. Microphone samples are not calibrated volts, decibels, or sound-pressure level, and their level varies with the device and browser.

Edge trigger and free-running fallback

Rising and falling trigger modes scan each sample window for the selected crossing of the normalized threshold. By default, a rising crossing at zero is placed about one division from the left edge, so repeated cycles line up at a stable horizontal position. Move the threshold to follow another part of a waveform, or select Auto (off) to always show the newest samples without trigger alignment.

An edge trigger can only lock when the chosen crossing is present in the current sample window and enough samples remain to draw the full trace. If the microphone is silent, a tone is too low for a short time window, or the selected threshold is not crossed, the display automatically free-runs through the newest samples instead of freezing. The trace can still look unstable with noisy or changing audio. Microphone inputs also differ in their built-in gain and processing. The tool asks the browser to turn echo cancellation, noise suppression, and automatic gain control off when supported; a browser may ignore those requests, and if it rejects them the tool retries with browser defaults.

Common waveform shapes

Shape	How it looks	Sound or source
Sine	Smooth repeating curve	Pure tone or tuning fork
Square	Fast transitions with flat high and low parts	Digital buzzer or synth
Sawtooth	Steady ramp followed by a sharp reset	String-like synthesizer tone
Triangle	Straight rising and falling slopes	Mellow synthesizer tone
Noise	Irregular movement without a repeating cycle	Hiss, wind, or room noise
Complex	Jagged or changing pattern	Speech, music, and everyday audio

Frequency and Lissajous figures

A waveform's period is the time for one complete cycle, and frequency in hertz is one divided by that period. A 440 Hz tone completes a cycle in about 2.27 ms. If one cycle covers two divisions at 1 ms/div, its period is about 2 ms and its frequency is about 500 Hz. This is a visual estimate; this page does not calculate a frequency readout.

Lissajous figures show the relationship between two signals when plotted against each other: equal-frequency signals can form a line, ellipse, or circle depending on phase, while different frequency ratios create more complex loops. This microphone oscilloscope has no X/Y mode, so those figures are background theory rather than a display available here.

Local microphone privacy

Audio is analyzed and drawn in this browser session only. The tool does not record or upload microphone audio and has no audio file picker. If you explicitly choose a microphone, the shared media-device preference stores its device-selection metadata in the browser under toolcroft:media-device:audioinput:v1 so other Toolcroft media tools can remember that device. The preference does not contain audio.

Frequently asked questions

What does the waveform show?

It plots normalized digital microphone amplitude on the vertical axis against time on the horizontal axis. At the default 0.25 amplitude/div scale, the grid spans approximately −1 to +1. These are not calibrated volts or sound-pressure levels, and the input level depends on the microphone and browser.

Is my audio recorded?

No. Audio is visualized locally in real time using the Web Audio API; it is not recorded or uploaded. This page has no audio file picker.

Why does the waveform sometimes move instead of locking in place?

Rising and falling modes align a visible crossing of the selected normalized threshold at a fixed point near the left side of the grid. If that edge is absent from the current sample window, including during silence or a low tone at a short time scale, the display free-runs instead of freezing. Auto (off) always free-runs.

Are the microphone processing controls guaranteed to be off?

The tool requests echo cancellation, noise suppression, and automatic gain control off when supported. A browser may ignore those requests; if it rejects them, the tool retries capture with browser defaults. Device processing can still affect the waveform.