Audio Forensics

How to Use the Audio Forensics Tool

1. Upload Audio: Drag and drop an audio file (MP3, WAV, OGG, FLAC, M4A) onto the upload zone, or click to browse your files.
2. Review Metadata: The tool instantly extracts file metadata — name, MIME type, size, duration, sample rate, channels, and total samples.
3. Examine Waveform: A visual amplitude waveform is automatically rendered, showing the audio signal's loudness variations across the entire duration.
4. Live Spectrogram: Click "Start live analysis" to arm the real-time frequency spectrogram, then press play on the custom player — the spectrogram visualizes how energy is distributed across frequencies over time while you keep full control of playback (seek, volume, speed). Click "Stop analysis" any time; you can restart it freely.
5. Custom Player: The built-in player replaces the browser default with play/pause, a seek bar, time readout, volume, playback speed, and keyboard shortcuts (Space, arrow keys). Clicking anywhere on the waveform jumps to that position.

Technical Overview & Use Cases

This audio forensics tool uses the Web Audio API's AudioContext and AnalyserNode to perform real-time frequency domain analysis. The waveform visualization reads raw PCM samples from the decoded audio buffer and renders min/max amplitude per pixel column on an HTML5 Canvas. The spectrogram uses Fast Fourier Transform (FFT) with a 2048-point window to decompose the audio signal into its constituent frequencies, painting a time-frequency heatmap that reveals hidden tonal patterns, splices, or anomalies invisible in the raw waveform.

Real-world use cases:

• Audio tampering detection: Identify splices, cuts, or inserted segments by spotting discontinuities in the spectrogram that indicate post-production edits.
• Voice analysis: Examine frequency content to identify speaker characteristics, background noise patterns, or synthetic speech artifacts.
• Music production QA: Verify mastering quality by inspecting frequency distribution, checking for clipping, and validating sample rate consistency.

Privacy & Security Guarantee

This tool is part of the FAK LAB ecosystem, founded by Faizan Ahmad Khan Khichi. Your audio files are processed 100% client-side using the Web Audio API built into your browser. Audio data is decoded locally into memory, analyzed, and visualized — it is never uploaded to any server. No recordings are stored, no audio fingerprints are generated remotely, and no third parties have access to your files.

Frequently Asked Questions

What does the spectrogram reveal that the waveform cannot?

A waveform shows amplitude over time — how loud the signal is at each moment. A spectrogram adds a frequency dimension, showing which specific frequencies carry energy at each point in time. This reveals hidden patterns: a splice may look smooth in the waveform but show abrupt spectral discontinuities in the spectrogram.

Why does the spectrogram require clicking "Start Live Analysis"?

The spectrogram uses real-time FFT analysis via the Web Audio API's AnalyserNode, which requires the audio to actually play through the AudioContext. This creates a MediaElementSource from the audio player and routes it through the analyser — the audio must be playing for frequency data to flow. This is a browser security requirement, not a limitation.

Can I detect if an audio file has been re-encoded or compressed?

Yes. Multiple compression passes leave characteristic artifacts: spectral band limiting (frequencies cut off at the encoder's filter point), pre-echo smearing in the spectrogram, and loss of high-frequency detail. A WAV file that was previously MP3 will show a hard spectral cutoff even though it's now in a lossless container.