Spectro overview

Spectro is a macOS app that detects fake lossless audio files in your music library. If you've ever downloaded a WAV or AIFF track that sounds dull or distorted on a professional sound system, the culprit is likely an upconverted MP3 - a low-quality file masquerading as lossless.

Spectro performs spectral analysis on your audio files and delivers an instant verdict: LOSSLESS, LOSSY, or FAKE. No subscriptions, no cloud uploads - everything runs 100% offline on your Mac, keeping your unreleased tracks private.

Key features

Instant fake lossless detection for FLAC, WAV, AIFF, ALAC and MP3 files
Batch-process entire folders, analyze 1000+ tracks in under a minute
Visual spectrogram with frequency cutoff detection
Deep Finder integration: Quick Look, drag & drop, and color tags
One-time purchase, $39 USD, no subscription required
Free trial: scan 100 audio files for free
Fully offline, no internet connection needed

Who is it for?

Spectro is designed for club DJs preparing sets, record pool subscribers cleaning their library, and audiophiles who want to verify the quality of their collection before it hits the speakers.

Stop playing fake lossless files on systems that deserve better.

What's new in version 1.4.1

Added

Upsampling detection - identifies files upsampled from 44.1 kHz to 48 kHz previously marked as Lossless.
Band-energy analysis - detects MP3 320 kbps with quantization noise above 19 kHz that previously passed as Lossless.
Multi-segment analysis - analyzes 5 independent segments and uses the maximum cutoff, fixing silent passages in classical music dragging results down.
Gray-zone second pass - refines ambiguous 19-20 kHz verdicts using density analysis to promote genuine lossless files or demote transcodes.
Crest Factor, Clipping, and Stereo Correlation now shown in results and CSV export.

Fixed

Classical and acoustic recordings with natural high-frequency rolloff no longer misclassified as Fake.
Piano recordings with sparse high-frequency content now show correct cutoff and confidence values.