Frequency Sweep Detector

Detect and characterize frequency sweeps, chirps, and siren patterns in real time using your microphone or by uploading an audio file. Measure sweep direction, rate, start/end frequencies, and sweep type — all processed locally in your browser.

Sweep Detector Tool

Understanding Your Results

Sweep Status

The sweep status shows "Detected" when the tool identifies a consistent frequency change over consecutive analysis frames. A sweep is detected when the peak frequency moves monotonically (consistently in one direction) by more than the sensitivity threshold over multiple frames. When no sweep is active, the status reads "None".

Direction Indicator

The direction arrow shows the sweep's trajectory: Ascending (frequency rising over time, arrow pointing up), Descending (frequency falling, arrow pointing down), or Bidirectional (alternating up and down, as in many siren patterns). The arrow animates in real time as sweeps are detected.

Sweep Rate

The sweep rate measures how quickly the frequency changes, expressed in Hz per second. This is calculated using linear regression on the peak frequency values over the sweep duration. A rate of 500 Hz/s means the dominant frequency shifts by 500 Hz every second. Higher rates indicate faster sweeps like emergency sirens; slower rates suggest gradual tonal shifts.

Start & End Frequency

These values show the frequency at the beginning and end of the current or most recently detected sweep. Together with the sweep rate and duration, they fully characterize the sweep's trajectory through the frequency spectrum.

Sweep Type

Sweeps are classified as Linear (constant Hz/s rate — the frequency changes by the same number of hertz each second) or Logarithmic (constant octaves/s rate — the frequency changes by the same ratio each second, common in audio test sweeps and many natural sounds). The classification compares the variance of the Hz/s rate against the variance of the octaves/s rate.

Siren Pattern Detection

The siren detector looks for periodic alternation between ascending and descending sweeps, which is characteristic of emergency vehicle sirens. When a repeating up-down pattern is detected, the siren indicator activates and estimates the siren cycle period.

How Frequency Sweep Detection Works

A frequency sweep (also called a chirp) is a signal whose instantaneous frequency changes over time. Sweeps appear in emergency sirens, radar signals, acoustic testing, musical effects, and many natural phenomena. This tool detects sweeps by tracking the peak frequency across consecutive FFT analysis frames.

Peak Frequency Tracking

Each analysis frame computes a Fast Fourier Transform (FFT) of the audio signal using a 4096-point window, yielding a frequency resolution of approximately 10 Hz at 44.1 kHz sample rate. The bin with the highest magnitude is identified as the peak frequency for that frame, with quadratic interpolation applied to achieve sub-bin accuracy.

Sweep Detection Algorithm

The detector monitors the peak frequency across a sliding window of consecutive frames. If the peak frequency changes monotonically (consistently increasing or decreasing) for a minimum number of frames, a sweep is declared. The sensitivity control adjusts how many consecutive monotonic frames are required and the minimum frequency change per frame.

Rate Calculation via Linear Regression

Once a sweep is detected, its rate is calculated by fitting a linear regression line to the peak-frequency-vs-time data points. The slope of this line gives the sweep rate in Hz/second. The sign of the slope determines direction: positive = ascending, negative = descending.

Linear vs. Logarithmic Classification

To classify the sweep type, the tool computes two rate series: the Hz/s rate (frequency difference between consecutive frames) and the octaves/s rate (log2 ratio between consecutive frames). If the Hz/s rate has lower variance relative to its mean, the sweep is classified as linear. If the octaves/s rate has lower variance, it is classified as logarithmic.

Siren Detection

Siren patterns are identified by detecting periodic alternation between ascending and descending sweeps. The tool tracks sweep direction changes and measures the intervals between them. If the direction change intervals are approximately equal (within a tolerance), a periodic siren pattern is declared.