P●privacycleanLAB◌   LOCAL-FIRST

PRIVACYCLEAN / INFORMATION

What Metadata Can Still Remain in Photos and Files? | PrivacyClean

Understand ICC profiles, thumbnails, hidden document content, unsupported tags and why basic metadata removal is not anonymity.

Photo metadata is not one thing

JPEG stores metadata in several marker segments. PNG stores information in named chunks. WebP can contain EXIF and XMP RIFF chunks. Removing recognized fields is useful, but a file may also contain proprietary segments, color profiles, comments, previews and third-party extensions that basic cleaners do not understand.

Why color profiles may be retained

ICC profiles can be necessary for accurate colors on some displays and printing systems. PrivacyClean generally retains them to avoid unexpected appearance changes. Retaining a profile means the resulting file is not literally stripped of every possible piece of metadata.

Visible pixels cannot be anonymized automatically

GPS data may be absent while the image still shows a recognizable house, location, face, branded uniform or vehicle number. Likewise, a PDF can disclose information through visible text, layers, annotations, forms, revisions or attachments rather than only through its document-info dictionary.

Why PDF scanning has narrow coverage

The PDF check is a heuristic scan for readable metadata keywords, not a complete parser or sanitizer. Compressed streams, incremental updates, encrypted files, object streams and embedded documents may conceal metadata from basic text scans. Do not interpret an empty detection result as proof that a PDF is private.

Browser re-encoding can change a file

Image re-encoding can alter orientation, transparency, pixel precision, animation, color fidelity or compression quality. Use the metadata-specific cleaner when you need to preserve original encoded image data; still check its output carefully. Different applications may add metadata again when exporting or uploading.

Text redaction also has blind spots

Regular expressions may miss nicknames, unstructured postal addresses, uncommon account identifiers, multilingual names or credentials hidden inside encoded values. Scan your output manually, including CSV headers and JSON property names—not just their values.

Practical verification

Compare source and output dimensions, file type, visual appearance and intended content. Inspect the cleaned file with an independent application, and check the context in which it will be published. For confidential or regulated data, a formal disclosure-review process is more appropriate than a free browser-only utility.

READ SUPPORTED FEATURESSupported formats ↗