Catches music sold as lossless that was secretly compressed. It accuses only when independent lines of evidence agree, because a false accusation costs more than a miss.
-
Updated
Oct 19, 2026 - Python
Catches music sold as lossless that was secretly compressed. It accuses only when independent lines of evidence agree, because a false accusation costs more than a miss.
Desktop audio spectrum analyzer: spectrograms, an automated lossless/transcode verdict, whole-library audit with a persistent cache, duplicate detection and folder diffs by acoustic fingerprint, encode comparison, and integrity/loudness checks. Avalonia / .NET.
Detect lossless audio files that were re-encoded from a lossy source
Find audio files whose real frequency content doesn't match their format: low-bitrate rips re-encoded at 320 kbps, lossy files passed off as FLAC.
Fake FLAC detector & spectrogram viewer: finds MP3/AAC/Opus transcodes, upsampled hi-res and padded 24-bit in your lossless music library. Names the source codec, scans albums, EBU R128/DR/clipping. Self-hosted web app (Docker), HTTP API.
To associate your repository with the transcode-detection topic, visit your repo's landing page and select "manage topics."