WKRD Studio AI Detector (SAD-100) + Metadata Inspector

Loca tool for MP3 and WAV files. Inspect visible metadata, hidden/custom tag blocks, scan metadata for known AI music maker names, edit common fields, add custom text metadata, write a WKRD Master SHA / Authorship Project Code into the export, preserve the automatic file fingerprint, and download a JSON report.

Private • Local • No Uploads
Edit Metadata
Extracted / Hidden Data
AI / SHA Check
Audio Watermark
Advanced Edit
Raw Inspector
Help
Load a file to begin editing.
Work on a copy when testing. Exporting creates a new file and does not overwrite your original.

2. Edit Common Metadata

Copyright / Authorship Rights

These editable rights lines are written into the clean export metadata and included in the JSON report. Keep the token private unless you intentionally use it to verify ownership.

3. Custom Text Metadata

MP3 export writes these as TXXX custom ID3 frames. WAV export writes keys with 4-character INFO codes. Use plain labels like Website, Publisher, ISRC, WKRD, or NOTE.

Extracted Metadata and Hidden/Custom Data

File Type
—
Size
—
Editable Fields
—
Suspicious / Hidden
—
Source Type Key / ID Value / Preview Bytes Risk Action
Load a file to inspect metadata.

Studio AI Detector (SAD-50) Quick AI Metadata Scan

This checks metadata text, filename text, visible tag values, selected raw text areas, and common hidden blocks for AI music maker names. A match means the name was found somewhere in the file data. It does not prove the whole song was AI-made. No match does not prove the song is human-made.
AI Name Matches
—
SAD-50 Result
—
WKRD Master SHA
Not set
Automatic File Fingerprint
—
Audio Review
—

WKRD Master SHA / Authorship Project Code

This is the main editable WKRD project identifier. Enter the code created by your WKRD SHA Maker and keep it with the song as stems, vocals, instruments, revisions, and artist contributions move between collaborators and states.

This project code can remain consistent across the entire collaboration. It is written into the edited audio metadata and included in the JSON report.

Automatic File SHA-256 Fingerprint

This technical fingerprint is calculated from the exact bytes of the loaded audio file. It changes whenever the file bytes change and remains locked so it can accurately verify that exact file version.

AI Music Maker Keyword List

You can add more names here later. Put one maker or alias per line.

Audio Watermark Detector & Review Editor

This local scan looks for persistent narrow-band tones, unusually stable high-frequency carriers, repeating level beacons, and left/right channel differences that can be consistent with an audio watermark. These are heuristic clues only: mastering, synths, noise shaping, codecs, and test tones can produce similar patterns.
Watermark Review
—
Strongest Carrier
—
High-Band Energy
—
Repeating Beacon
—

Detector Settings

Detected Clues

Load a file to scan for possible audio watermark clues.

The detector does not identify a watermark vendor or decode proprietary watermark payloads. It reports signal patterns for manual review.

Repeating Tone Cleanup Editor

This cleanup section is for audio you own or are authorized to edit. It targets audible or near-audible repeating tones by frequency; it does not decode or defeat a proprietary authentication system. The original file is never overwritten. Cleanup preview/export renders a new 24-bit PCM WAV.
Detected Low Tone
—
Detected High Tone
—
Cleanup Preview
Not rendered
Output
24-bit WAV

Narrow Tone Filters

Enter up to three fixed tones. A blank frequency disables that slot. Higher Q values make the cut narrower so less surrounding music is affected.

Optional Band Limits

Load a file, scan it, then load or enter the tones you want to reduce.

For an alternating two-tone signal, put the low frequency in one slot and the high frequency in another. Start with a 12–24 dB reduction and a Q around 18–30, then listen before making the cut stronger.

Advanced Hidden Data / Raw Edit

Advanced raw edits patch bytes directly. For safety, raw patches must be the exact same byte length as the original range. To remove hidden blocks completely, use the privacy-clean export.

Selected Extracted Item

Click an Edit button in the Extracted / Hidden Data table.

Text metadata is safest to edit by adding it to the metadata editor and exporting a clean copy. Binary/private frames should normally be removed with privacy-clean export unless you know the exact bytes you want to patch.

Raw Byte Patch Editor

No range selected.
No raw patches queued.

Raw Technical Inspector — Editable Workspace

You can edit and save this inspector text as part of the browser draft and JSON report. Editing this text does not directly change the audio bytes. Use the Advanced Edit tab to patch actual file bytes by offset.

What This Tool Can and Cannot Do

MP3: Reads ID3v2 frames, ID3v1 tags, common text fields, comments, custom TXXX fields, private PRIV frames, unique IDs, embedded artwork, URL frames, padding, and some trailing metadata signatures.

WAV: Reads RIFF chunks, LIST/INFO metadata, BEXT broadcast metadata, iXML XML metadata, embedded ID3 chunks, cue/marker chunks, JUNK/PAD chunks, and unknown custom chunks.

Editing: You can edit normal text metadata and custom text fields. You can also select extracted hidden items, add edited text back into the clean metadata export, or use the raw byte patch editor for exact same-length binary edits by offset.

AI maker scan: The SAD-50 scan looks for known AI music maker names in metadata and selected raw text areas. It is a clue finder, not a court-proof AI detector.

Quick audio analysis: The report also runs a local browser analysis for duration, sample rate, peak/RMS level, rough dynamic range, clipping, transient activity, spectral balance, stereo width, and synthetic/electronic clues. This can suggest possible electronic instruments or heavy processing, but it cannot prove AI generation by waveform alone.

WKRD Master SHA: The main WKRD Master SHA / Authorship Project Code is editable and can stay consistent across collaborators, stems, revisions, and final assembly. A separate automatic SHA-256 fingerprint is calculated from the exact file bytes and remains locked for file-version verification.

Audio watermark detector: Locally samples decoded audio for persistent narrow-band carriers, high-frequency energy, repeating level patterns, and stereo-difference clues. The Watermark tab lets you adjust sensitivity, edit the review report, set a manual review status, add notes, and save the review into the JSON report or metadata editor. It does not decode proprietary watermark payloads and does not prove a watermark is present.

Repeating tone cleanup: The Watermark tab also includes a local frequency cleanup editor for audio you own or are authorized to edit. It can reduce up to three fixed tones with narrow adjustable cuts, optionally apply gentle high-pass/low-pass filters, provide original/cleaned A/B listening, and export a separate 24-bit PCM WAV without overwriting the source.

Privacy: The file stays in your browser. There is no upload, no server, and no external library loaded from the internet.