Losselot Docs Decision Graph

Audio forensics: spectral FFT + MP3 binary analysis in WebAssembly

Drop audio files here

or click to browse

Supports: MP3, FLAC, WAV, OGG, M4A, AIFF

filename.flac

Analysis
Spectral
Binary
Raw JSON
Sample Rate: 44100 Hz
Duration: 0.0s
Channels: 2
Format: -
Encoder: -
Lowpass: -
Bitrate: -
VBR: -
Issues Detected
Transcoding Evidence
Analysis indicates prior lossy encoding
Estimated Quality Retention
0% (Destroyed) ~70% 100% (Original)
Source
Lossy Encode
Re-encode

Spectrogram (Time vs Frequency)

Analysis Summary

OK
Verdict
Analysis pending...
Combined Score
0/100
Avg Cutoff Frequency
- Hz
Cutoff Variance
-
Rolloff Slope
- dB/kHz
CFCC Cliff Detection
Not detected
Natural Rolloff
-

How it works

Full Rust analysis in WebAssembly - spectral FFT + MP3 binary parsing. For MP3 files, we extract LAME headers, encoder signatures, lowpass filters, and detect re-encoding chains. Your files never leave your device.

The analyzer performs: 8192-sample windowed FFT, CFCC codec cliff detection, MP3 frame parsing, LAME/Xing header extraction, encoder signature counting, and lowpass-to-bitrate mismatch detection.