FAQ
Last updated: 2026-05-22
A fake lossless file is an MP3 or AAC that has been re-encoded and saved as WAV, AIFF or FLAC. The file size is larger but the audio quality is identical to the original compressed file - the high frequencies above the MP3 cutoff are permanently lost. Playing a fake lossless on a professional PA system means lost high frequencies and a set that doesn't hit the way it should.
Spectro runs a full FFT spectral analysis on each file and measures where the frequency content actually ends. A genuine lossless file has energy up to 20-22kHz. A fake lossless - an MP3 or AAC re-encoded as WAV - shows a hard cutoff, typically around 16-20kHz.
Yes. Spectro processes files from any source — WAV, AIFF, FLAC, ALAC, MP3, and AAC.
Yes. In Spectro 1.3 and later, import your full Rekordbox collection directly via File → Import from Rekordbox — no drag and drop required. Spectro reads your XML export and loads all tracks automatically. You can also drop any folder manually. Analyzes WAV, AIFF, FLAC, ALAC, MP3, and AAC and returns a LOSSLESS, FAKE, or MEDIUM verdict per track.
Yes. Spectro 1.3 added full ALAC support. Apple Lossless Audio Codec files (.m4a) are analyzed the same way as WAV and FLAC — spectral content determines the verdict, not the container format.
Yes. Use File → Import Playlist to load any M3U or M3U8 file. Spectro resolves the track paths and loads them automatically. This works with playlists exported from Traktor, VirtualDJ, and any other software that supports the M3U format.
Yes. Use File → Export CSV to save your full batch results. The export includes verdict, file format, declared bitrate, spectral cutoff frequency, and confidence score for every track — useful for auditing large libraries or sharing results.
No. Spectro runs entirely on your Mac. Nothing is uploaded, no account is needed, and no internet connection is required for analysis. Your unreleased promos and unannounced tracks never leave your machine.
Spectro is free for your first 100 files - no time limit, no account required. Download and try it on your own tracks before buying.
You receive a license key by email. Enter it in the app to unlock unlimited scans with a 2 Mac license.
WAV, AIFF, FLAC, ALAC (.m4a Apple Lossless), MP3, and AAC. Spectro reads the actual frequency content, not just the file extension or container format.
The frequency point where audio content stops. A genuine lossless file has energy up to 22kHz. A fake lossless shows a hard cutoff around 16-20kHz - the fingerprint of lossy encoding that Spectro detects instantly.
Yes. Spectro is a modern macOS alternative to Spek - built natively for Apple Silicon and Intel, with batch processing, automatic LOSSLESS / FAKE / MEDIUM verdicts, and Finder Tag integration. No manual interpretation needed.
macOS 13 Ventura or later. Optimized for Apple Silicon and Intel Macs.
Spectro's detection method is based on peer-reviewed research. A 2009 study published at the ACM Multimedia & Security Workshop analyzed 2,512 songs across 94 artists and achieved 99% accuracy detecting transcoded audio - files re-encoded to a higher bitrate to disguise their true quality. The only known ambiguous case is 256 kbps CBR versus high-quality VBR (such as MP3 VBR-0), where spectral analysis alone reaches 92.4% accuracy. When Spectro encounters this edge case, it returns a MEDIUM verdict with an explicit note explaining the ambiguity, rather than making a confident wrong call.
Yes. When an MP3 is converted to a lossless format like FLAC or WAV, the lost high-frequency information is not recovered - it stays gone. The spectral cutoff left by the original MP3 encoder remains permanently embedded in the file, regardless of the container it's placed in. Spectro reads that cutoff and returns a FAKE verdict. A 128 kbps MP3 converted to FLAC will show the same spectral signature as the original 128 kbps MP3.
Both 256 kbps CBR and VBR-0 (the highest VBR setting) produce a spectral cutoff in the same 19-20 kHz range. This is a known limitation of frequency-based analysis documented in peer-reviewed research - it's the only bitrate pair that consistently falls below 97% classification accuracy. In this case Spectro shows a MEDIUM verdict and includes a diagnostic note: "Cutoff in the 19-20 kHz range - consistent with 256 kbps CBR or high-quality VBR; spectral analysis cannot distinguish between the two." Both formats are acceptable quality for DJ use.
Results may be less reliable. Streaming platforms re-encode uploaded audio using their own codec pipeline - typically AAC or Opus at platform-defined bitrates - before making it available for download or offline listening. If you analyze a track that has passed through platform encoding, Spectro is reading the quality of the platform's output, not the original source file. For the most reliable results, analyze files from direct purchase sources such as Beatport, Bandcamp, or your own CD rips.
No. Converting an MP3 to WAV changes the container format and increases the file size, but does not restore any lost audio data. The high-frequency information removed during MP3 encoding is permanently gone — the conversion preserves the defects, not the quality. A WAV converted from a 128 kbps MP3 is acoustically identical to the original 128 kbps MP3.
Yes, if the source quality is comparable. A genuine 320 kbps MP3 from a high-quality master preserves more real audio information than a fake FLAC made from a low-quality source — for example, a 128 kbps MP3 re-encoded as WAV. The file format label (FLAC, WAV, AIFF) only describes the container, not the quality of the content inside. Format label matters less than source quality.
Often yes, especially at high volume on professional systems. Fake lossless files carry the frequency loss of their original lossy source — typically a hard cutoff in the 16–20 kHz range that removes the air and detail from cymbals, hi-hats, and high-frequency transients. On a club PA at loud levels these losses become more noticeable: dull highs, smeared transients, reduced stereo detail. On consumer speakers or at low volume the difference may be hard to detect.