Frequency Drift Monitor

Monitor frequency stability over long periods — from minutes to hours. Track drift rate in Hz/minute and PPM/hour, lock onto a target frequency and watch deviation in real time, set alert thresholds, and export timestamped data to CSV — all processed locally in your browser.

🔒 Your audio never leaves your device — 100% local processing, zero uploads. Chrome Firefox Safari Edge

Space Start/Stop R Reset C Calibrate F Freeze

Microphone:

Target Frequency

Hz

Alert Threshold

± Hz

Time Scale

Sensitivity -55 dB

📁 Drag & drop an audio file here, or browse WAV, MP3, OGG, FLAC — analyze frequency drift from a recording

—

Hz

Deviation (Hz)

—

Deviation (PPM)

—

Drift Rate

— Hz/min

PPM / Hour

PPM/hour: —

Alert Status

Idle

Session Statistics

Mean

—

Std Dev

—

Min

—

Max

—

PPM (overall)

—

Elapsed Time

00:00:00

Frequency vs Time

Deviation from Target

Frequency Histogram

Data Log

Time	Frequency (Hz)	Deviation (Hz)	PPM

How to Use the Frequency Drift Monitor

Set Your Target Frequency

Enter the expected frequency you want to monitor (e.g., 1000 Hz for a calibration tone, 440 Hz for concert A, or 50/60 Hz for mains). Set the alert threshold to the maximum acceptable deviation in Hz.
Start Monitoring

Click "Start Monitoring" and grant microphone permission. Select your preferred microphone from the dropdown. The tool begins tracking frequency immediately and logging data points with timestamps.
Watch the Drift Chart

The main chart shows detected frequency over time with auto-scaling axes. Use the Time Scale selector to zoom between 1 minute and 1 hour views. The deviation chart below shows how far the signal drifts from your target.
Monitor Alerts and Statistics

The alert indicator turns red when deviation exceeds your threshold. The statistics panel shows mean, standard deviation, min, max, and overall PPM drift across the entire session.
Export Your Data

Click "Export CSV" to download a complete log with timestamps, frequency readings, deviation, PPM, and drift rate. Use this for documentation, calibration records, or further analysis.

Understanding Your Results

Current Frequency (Hz)

The large display shows the real-time detected frequency from your audio input. This is the peak frequency extracted via FFT analysis with parabolic interpolation for sub-Hz accuracy. The reading updates continuously while monitoring is active.

Deviation (Hz and PPM)

Deviation in Hz is the simple difference between the measured frequency and your target: measured - target. PPM (parts per million) normalizes this deviation relative to the target frequency: (measured - target) / target × 1,000,000. PPM is useful for comparing stability across different frequencies.

Drift Rate (Hz/min)

The drift rate uses linear regression over the recent data window to calculate how fast the frequency is changing. A positive value means the frequency is rising; negative means it is falling. This helps identify systematic drift caused by temperature changes, component aging, or power supply issues.

Statistical Summary

The session statistics show the mean (average frequency), standard deviation (spread of readings), min/max (extremes), and overall PPM (mean deviation from target in parts per million). Lower standard deviation indicates better frequency stability.

Alert System

When the absolute deviation exceeds your configured threshold, the alert indicator turns red and an optional audio beep sounds. This is useful for unattended monitoring where you need to know immediately if the signal drifts out of specification.

Technical Background

Long-term frequency monitoring requires sustained accuracy over extended periods. This tool combines real-time FFT analysis with statistical methods to characterize both instantaneous frequency and long-term stability trends.

FFT Peak Tracking

The Web Audio API's AnalyserNode computes the FFT in real time. With an FFT size of 8,192 at a 48 kHz sample rate, the raw frequency resolution is approximately 5.86 Hz per bin. Parabolic interpolation around the peak bin refines this to sub-Hz accuracy, which is essential for detecting small drifts over time.

Linear Regression for Drift Rate

The drift rate is computed using least-squares linear regression on the time-frequency data within the visible window. The slope of the best-fit line gives the drift rate in Hz per minute. This method is robust against individual noisy readings because it considers the overall trend rather than just consecutive differences.

PPM Calculation

Parts per million (PPM) is a standard metric for expressing frequency accuracy and stability. It is computed as (f_measured - f_target) / f_target × 10⁶. For example, a 1 Hz deviation at 1 MHz is 1 PPM, while a 1 Hz deviation at 1 kHz is 1,000 PPM. This normalization allows meaningful comparison of stability across very different frequency ranges.

Standard Deviation and Stability

The standard deviation of frequency readings quantifies short-term stability (jitter). Lower values indicate a cleaner, more stable signal. For reference, a quartz crystal oscillator typically achieves stability better than 100 PPM, while MEMS oscillators are usually within 20–50 PPM.

Frequently Asked Questions

What is frequency drift and why does it matter?

Frequency drift is the gradual change in a signal's frequency over time. It matters in applications like radio communications, audio calibration, and electronic testing where frequency accuracy must be maintained within tight tolerances. Excessive drift can cause signal loss, detuning, or measurement errors.

How long can I monitor continuously?

The tool can run for many hours continuously. Data is stored in browser memory, so very long sessions (12+ hours at high logging rates) may eventually consume significant RAM. For extremely long monitoring, export CSV periodically and reset the session. The tool uses efficient data structures to minimize memory usage.

What does PPM mean and how is it calculated?

PPM (parts per million) expresses frequency deviation relative to the target frequency. The formula is: (measured - target) / target × 1,000,000. For example, if your target is 1000 Hz and you measure 1001 Hz, the deviation is +1000 PPM. PPM allows you to compare stability across different frequency ranges on a common scale.

How accurate is the drift rate calculation?

The drift rate uses linear regression over the visible data window, which is more accurate than simple point-to-point comparison. Accuracy improves with more data points and longer monitoring periods. For best results, ensure a clean signal with minimal noise, and use a monitoring window of at least 5 minutes.

Can I use this to test oscillator stability?

Yes, this is an ideal use case. Feed the oscillator output into your microphone input (using appropriate signal conditioning), set the target to the nominal frequency, and monitor over time. The standard deviation and PPM readings will characterize the oscillator's short-term and long-term stability. Note that accuracy is limited by your sound card's sample rate and the FFT resolution.

Is my audio data private?

Absolutely. All frequency detection and analysis runs 100% in your browser using the Web Audio API. No audio data is ever recorded, transmitted, or stored on any server. The tool works completely offline once loaded.

~