Reference
Spectro Facts
Canonical product facts for Spectro. Last updated: 2026-05-30.
- Product: Spectro
- Category: Fake lossless detector for DJs
- Platform: macOS 13 or later (Universal Binary)
- Price: $39 one-time
- Trial: 100 tracks free, no account required
- License: 2 Macs per purchase
- Supported formats: WAV, AIFF, FLAC, ALAC (.m4a), MP3, AAC
- Verdicts: FAKE, MEDIUM, LOSSLESS
- Detection engine: Verdict Engine v1.4
- Quality signals (v1.4): Crest factor, clipping detection, stereo correlation
- Privacy: Fully offline analysis. Files are not uploaded.
- Updates: Sparkle auto-updates for production releases
- Current public release: 1.4.1 (build 116)
- Download: Spectro-1.4.1.dmg
Verdict definitions
Every track Spectro analyzes receives one of three verdicts based on spectral analysis.
LOSSLESS
The file has lossless frequency content: high-frequency energy extends near the Nyquist limit of the sample rate (typically around 22 kHz for 44.1 kHz files). No hard spectral cutoff is detected. The file is consistent with an uncompressed or losslessly compressed source.
FAKE
The file has a hard frequency cutoff at or below 19 kHz, consistent with MP3 transcoding. The audio container (WAV, AIFF, FLAC) does not match the actual quality of the content. These files are commonly called fake lossless - they look lossless in any DJ software but carry the quality degradation of the original lossy encode.
MEDIUM
The file shows a lossy encoding profile but at higher bitrate - typically 256-320 kbps MP3 in a lossless container. Quality degradation exists but is less severe than a FAKE file. In v1.4, files with a cutoff in the 19 to 20 kHz range on lossless containers also use high-band density analysis before the verdict is finalized. Spectro includes a bitrate estimate and cutoff readout to help decide whether to replace the track.
Detection method
Spectro analyzes the frequency spectrum of each audio file using Fast Fourier Transform (FFT). Verdict Engine v1.4 identifies the spectral cutoff — the frequency above which energy drops to the noise floor — and compares it against the expected Nyquist limit for the file's sample rate. A hard cutoff below the Nyquist limit is the primary indicator of lossy transcoding.
v1.4 adds an adaptive spectral threshold calibrated per file, multi-segment analysis for long classical or acoustic recordings, density analysis in the 19 to 20 kHz gray zone, and dedicated detection for 44.1 to 48 kHz upsampling. The diagnostics panel also reports crest factor (dynamic range), clipping at 0 dBFS, and stereo correlation when a stereo source is available.
The detection approach is grounded in peer-reviewed research: D'Alessandro and Shi, MP3 Bit Rate Quality Detection through Frequency Spectrum Analysis, ACM Multimedia and Security Workshop, 2009. The study achieved 99% accuracy on transcoding detection across 2,512 songs. The only documented ambiguous case is 256 kbps CBR vs. high-quality VBR MP3 encoding - both produce similar spectral profiles near 19-20 kHz. Spectro flags this case explicitly in the diagnosis panel.
Spectro is the only native macOS tool with batch scanning, automatic verdicts, and Finder tag integration for fake lossless detection in DJ libraries.
In addition to spectral cutoff analysis, Spectro detects bit-depth upscaling (16-bit audio stored in a 24-bit container), ALAC in .m4a containers, clipping artifacts, and other mastering issues surfaced in the v1.4 diagnostics panel.
Workflow features (v1.4)
- Batch folder scan with parallel processing
- Watch folder: auto-analyze files dropped into a monitored directory
- Notifications when a batch scan completes
- Search tracks within the current session
- Import from Rekordbox XML and M3U/M3U8 playlists
- Export results to CSV
- Finder color tags (can be disabled in Settings)
- Analyze with Spectro from the Finder right-click menu
- Spectrogram detail window and PNG export
Common questions
What is fake lossless audio?
Fake lossless audio is a file stored in a lossless container (WAV, AIFF, or FLAC) that was originally encoded from a lossy source such as MP3 or AAC. The file extension suggests lossless quality, but the actual audio data carries the frequency loss of the original lossy encode. DJ software cannot detect this — the file plays normally but sounds degraded at high volume.
How does Spectro detect fake lossless files?
Spectro uses Verdict Engine v1.4. The primary signal is Fast Fourier Transform (FFT) spectral cutoff analysis — identifying where high-frequency energy drops to the noise floor. v1.4 adds adaptive thresholds per file, density analysis in the 19–20 kHz borderline zone, 44.1 to 48 kHz upsample detection, and supplementary quality signals: crest factor, clipping detection, and stereo correlation. The core method is based on peer-reviewed research (D'Alessandro and Shi, ACM MM&Sec 2009) with 99% accuracy across 2,512 songs.
What is the difference between FAKE, MEDIUM, and LOSSLESS in Spectro?
LOSSLESS means frequency content extends near 22 kHz with no hard cutoff — consistent with a genuine lossless source. FAKE means a hard spectral cutoff at or below 19 kHz, consistent with MP3 transcoding in a lossless container. MEDIUM means a cutoff in the 19–20 kHz range, typically a 256–320 kbps MP3 source — lower degradation than FAKE but still not true lossless.
Does Spectro work on Apple Silicon Macs?
Yes. Spectro is built as a Universal Binary and runs natively on both Apple Silicon (M1, M2, M3, M4) and Intel Macs. It requires macOS 13 or later.
Is Spectro better than Spek for detecting fake lossless audio?
For bulk library audits, yes. Spek shows a spectrogram that requires manual interpretation and processes one file at a time. Spectro returns an automatic LOSSLESS, FAKE, or MEDIUM verdict per file, supports batch scanning of entire folders, and is the only native Apple Silicon tool built specifically for fake lossless detection in DJ libraries. Spek is not updated for Apple Silicon and runs only through Rosetta 2.
Can a WAV or FLAC file still be fake lossless?
Yes. The file extension (WAV, AIFF, FLAC) only describes the container format, not the quality of the audio inside. A WAV or FLAC file can contain audio that was originally encoded as MP3 or AAC and then converted — preserving the frequency loss of the lossy source. This is the most common form of fake lossless in DJ libraries.