File Metadata Explained: What It Reveals and What Can Be Removed

Metadata is not one thing

“Metadata” is a convenient umbrella term, but the data is not stored in one universal layer. A JPEG may use EXIF, XMP, and IPTC structures. A PDF can have a document information dictionary as well as an XMP packet. An MP3 can carry ID3 frames, while an MP4 or MOV organizes information into boxes or atoms.

That difference matters because a tool can remove one structure and leave another intact. It is more accurate to ask which fields, in which container, are detected and rewritten than to ask whether a file has been made “metadata-free.” The CIPA EXIF specification, for example, describes a specific family of image metadata—not every possible non-pixel structure in an image.

Privacy-sensitive data and technical data are different

Privacy-sensitive metadata may include GPS coordinates, author names, device identifiers, timestamps, comments, copyright fields, or editing software. Those fields can add context a recipient did not need.

Technical properties such as image dimensions, codec names, sample rate, duration, or color information often remain because software needs them to render or play the file. Their presence does not automatically mean a cleaning operation failed. The practical goal is to reduce unwanted context while preserving a usable file.

A format-by-format metadata map

For a closer look at the main image standard, read what EXIF data contains. If your immediate goal is action rather than background, use the format-specific workflows for photos, PDF files, video, or audio.

What metapeel currently handles

metapeel reads files in the browser and chooses a cleaner based on the detected format. Image and document work uses browser JavaScript and a Web Worker. Audio and video cleaning loads FFmpeg WebAssembly on demand and remuxes supported containers locally. File contents are not uploaded to a metapeel processing server.

These boundaries are why the site uses “remove supported metadata” rather than “sanitize every hidden element.” The distinction is especially important for professional documents.

Metadata cleaning is not hidden-content sanitization

A PDF annotation, an Office tracked change, a spreadsheet's hidden worksheet, and a document author field can all reveal information—but they are not the same technical object. Clearing document properties does not inspect every place sensitive content may live.

Before sharing a sensitive Office file, use the application's own inspection workflow in addition to metadata cleaning. Microsoft's Document Inspector guidance explicitly treats revisions, comments, hidden content, and properties as separate categories. For PDFs, visually inspect annotations, attachments, layers, forms, and redaction results as a separate task.

A safer workflow before sharing

1. Work on a copy. Keep the original in a controlled location until the cleaned file has been checked.
2. Inspect before cleaning. Review which supported fields the tool can actually detect.
3. Clean with the format-specific path. Use the dedicated image, document, video, or audio tool.
4. Inspect the output again. A second pass confirms what the same reader can still detect.
5. Use an independent checker for higher-risk files. Format-native tools can reveal structures a browser cleaner does not cover.
6. Review visible and hidden content separately. Metadata removal does not replace redaction or document inspection.

This workflow is deliberately less dramatic than a “one-click anonymity” promise. It is also much more useful: you know what was checked, what changed, and what still deserves attention.