A developer known as Angel2mp3 released AudioAuditor on July 28, 2026—a free, open-source Windows application that bundles a full-featured audio player with a forensic analysis engine capable of sniffing out upsampled FLAC files, identifying AI-generated music, and inspecting spectral details that most players ignore. The tool, released under the Apache 2.0 license, aims to give everyday users and power users alike a single dashboard for verifying the authenticity and quality of their local music collections.

What AudioAuditor Can Actually Do for Your Music

AudioAuditor isn’t just another media player. It’s built around a set of verification tools that answer questions most apps never ask: Is that high-resolution FLAC really lossless, or was it converted from a low-bitrate MP3? Does a “stereo” track actually contain unique left and right channels? Are there hidden clipping problems or silence gaps that could ruin a playlist or podcast? And, in a beta feature, can we spot telltale signs that a song was generated by AI?

The application combines audio analysis, metadata editing, spectrogram viewing, file comparison, batch processing, and playback into one interface. Its feature set is unusually broad for a free tool:

  • Spectral analysis with a 4096-point Hanning-windowed FFT by default, and a Hi-Fi mode using a 16384-point Blackman-Harris-windowed FFT for deeper inspection
  • Fake-lossless detection that looks for telltale frequency cutoffs suggesting an upsampled origin
  • Fake-stereo detection to identify tracks where both channels are essentially duplicates
  • Clipping and loudness measurements, including True Peak, LUFS, and dynamic range
  • MQA and MQA Studio detection
  • AI music detection (beta), checking for watermarks, byte patterns, C2PA provenance, and optional cloud-based verification
  • Multi-mode waveform and spectrogram comparison with stacked, overlay, and wipe views—overlay even shows a pixel-level difference heatmap
  • Full metadata editor with batch enrichment via MusicBrainz and AcoustID fingerprinting
  • Batch tools for deduplication, metadata stripping, playlist import, cue sheet parsing, and auto-renaming
  • Report exports to CSV, TXT, PDF, XLSX, and DOCX

These aren’t just academic exercises. For anyone who has ever downloaded a 24-bit/192kHz FLAC that sounded suspiciously thin, or who manages a sprawling music library full of mystery files, AudioAuditor provides evidence you can actually act on.

A Real Player, Not an Afterthought

The audio player component holds its own, too. It supports native 96kHz and 192kHz playback (with fallback resampling), gapless playback, a 10-band parametric EQ, crossfade controls, peak normalization, multiple visualizers, and Windows System Media Transport Controls integration. There’s even spatial audio processing with HRTF-style crossfeed. This means you can listen critically while inspecting a track without switching programs.

Who Should Care About This Tool?

Home Music Collectors and Audiophiles

If you’ve built a library from CDs, downloads, and rips, AudioAuditor helps you separate genuine high-resolution files from upsampled ones. The fake-lossless detection isn’t foolproof—some recordings naturally have limited high-frequency content—but it gives you a starting point. You can then inspect the spectrogram, compare versions, and decide whether to keep or replace a file. The duplicate detection and automated tagging also make library maintenance faster.

Podcasters, DJs, and Video Editors

Silence-gap detection is a practical highlight. AudioAuditor can find dead air that might throw off a mix, disrupt a playlist, or ruin a podcast edit. Clipping and loudness checks let you spot tracks that were pushed too hard during mastering. The batch tools can process folders of files, stripping unnecessary metadata or renaming from tags.

Archivists and IT Administrators

For anyone responsible for long-term audio storage, the ability to verify file integrity through spectrogram comparison is invaluable. You can compare an original archival copy to a derivative to check for accidental degradation. The privacy design—no telemetry, offline mode, and clear documentation of all network activity—also makes it suitable for sensitive collections.

Curators and Rights Managers (with caution)

The AI music detection feature is particularly relevant for those vetting submissions or managing content platforms. But as the developer warns, results can be inaccurate. The tool provides a confidence score (≥70% = “Yes”, 35–70% = “Possible”, below 35% = “No”) by averaging several detectors. It’s best used as a triage and evidence-gathering step, not a final verdict. A positive result backed by recognizable watermarks or C2PA provenance is strong; a “Possible” result should lead to manual review, not immediate rejection.

How the Tech Works—and Where It’s Cautious

AudioAuditor’s spectral analysis uses standard signal-processing techniques that are familiar to audio engineers, but it presents them accessibly. The default 4096-point FFT with Hanning window provides a good balance of detail and speed for library scanning. When you need to zoom in on a suspicious cutoff, the 16384-point Blackman-Harris mode offers finer frequency resolution. You can switch between logarithmic and linear scales, view left/right channel differences, and export spectrograms as PNGs.

The fake-lossless detector looks for an abrupt high-frequency drop-off, typically around the cutoff of a lossy codec—for example, a sharp roll-off above 16 kHz or 20 kHz that suggests the file was converted from an MP3 or AAC. However, not every roll-off is fraud. Older recordings, certain mastering choices, and even some legit high-res transfers may lack ultrasonic content. The tool flags files for your attention; it doesn’t convict them.

AI detection is the most experimental part. It locally scans for embedded metadata (like a “created by” tag), byte patterns associated with known generation models, C2PA content credentials, and watermark identifiers. An optional SH Labs cloud service can add another layer, though raw audio never leaves your device. The developer is transparent about the beta nature: false positives and negatives are possible, and the tool should never be used to defame or harass. That’s a responsible posture for a field where even paid services struggle with accuracy.

Privacy is a strong point. AudioAuditor has no telemetry, no analytics, and no usage tracking. Network activity is limited to user-invoked features like MusicBrainz lookups, Discord Rich Presence, scrobbling, and the optional AI check. An offline mode disables all connections. Settings, caches, EQ profiles, and listening statistics live in %AppData%\\AudioAuditor\\, fully under your control.

How We Arrived at This Moment

The need for audio verification tools has been growing for years. As music moved from physical media to digital files, the problem of “fake FLACs” emerged: sellers and file-sharers upsampling low-quality sources and passing them off as high-resolution. Tools like Spek (a basic spectrogram viewer) and Fakin’ The Funk (a commercial checker) filled some gaps, but they often required switching between multiple applications. Meanwhile, the explosion of AI-generated music—some of it clearly labeled, much of it not—created a new headache for platforms, labels, and listeners.

AudioAuditor arrives as a comprehensive, open-source answer that consolidates many of these checks. Its developer, Angel2mp3, built it using .NET 8 with a Windows Presentation Foundation (WPF) front end, and a shared core that powers both the desktop app and a command-line interface (CLI) available for Windows and Linux. A macOS CLI is promised. The permissive Apache 2.0 license means the code is open to inspection, modification, and redistribution, though the branding assets are protected.

Getting Started: What to Do Now

  1. Download from the official source. The safest route is the project’s official website or GitHub repository. The developer warns against unofficial distributions, which could harbor malware. The desktop app requires Windows 10 or later (64-bit).
  2. Point it at your library. Launch the program and add the folders you want to audit. You can enable or disable individual analysis modules (fake-lossless, fake-stereo, clipping, AI detection, etc.) to speed up processing. Use the batch GUI for mass tagging or normalization.
  3. Learn to read the spectrogram. Even if you’re not an audio engineer, a quick visual inspection can reveal obvious upsampling. Look for a flat “shelf” at a specific frequency rather than a natural roll-off. The Hi-Fi mode and zoom are your allies when uncertain.
  4. Interpret AI results carefully. Treat a “Yes” verdict as cause for deeper investigation—check the details panel to see which detectors contributed. A “Possible” result is just that: possible. Use the comparison tools to examine the file alongside a known human performance if you have one.
  5. Don’t ignore the player. The EQ, crossfade, and visualization features can help you critically listen to suspect tracks. If a supposed high-res file sounds no better than a CD quality rip, the analysis might explain why.
  6. Stay offline if privacy matters. Enable Offline Mode if you don’t need internet-dependent features like MusicBrainz lookups or the optional AI cloud service. This ensures zero network communication.
  7. Keep expectations realistic. No automated tool can definitively “prove” a file’s provenance. Use AudioAuditor as a skeptical assistant, not an infallible judge. When in doubt, seek the original source.

The Road Ahead

AudioAuditor’s ambitious scope—player, analyzer, tagger, reporter—is both its strength and its risk. A young open-source project with so many features may face stability issues or bite off more than it can maintain. But the developer has shown strong attention to transparency, from the privacy model to the AI detection disclaimers. The CLI and planned macOS support suggest a commitment to broader usability. For Windows users who want to audit their music collections without paying a cent or stitching together a dozen utilities, AudioAuditor is already one of the most compelling options available. As AI music generation continues to evolve, tools that help verify authenticity—even imperfectly—will only grow more essential.