Remove image and PDF metadata before sharing
A clean-looking file can still describe where it was created, which device or application made it, who authored it, and when it changed. Metadata is useful for organizing and editing files, but it deserves the same deliberate review as visible text when a copy leaves your device or organization.
Put this guide into practice: Remove source metadata while redacting a PDF. Requires a visible redaction and creates image-based pages.
About this guide
By Derek Brumby · Product behavior checked with synthetic FreeRedact test files
Published · Updated · 9 minute read
How this guide was checked
The process was checked with synthetic JPEG images containing non-sensitive camera-style metadata and PDFs containing fictional title and author properties. Results vary by source application and format.
The process below uses only fictional sample information. Test the workflow with harmless data before relying on it for a sensitive file.
Know what may travel with the file
Images can contain date and time, device details, software names, dimensions, captions, and GPS coordinates. PDFs can contain title, author, subject, keywords, creation and modification dates, producer software, comments, attachments, forms, and hidden layers.
Not every field is sensitive in every situation. A camera model may be harmless for a product photo but identifying in a confidential investigation. Review metadata according to the audience, purpose, and risk of the specific share.
Inspect before you remove
Check file properties in the operating system and the source application. For a photo, review its information panel and map. For a PDF, review document properties and panels for comments, forms, layers, links, and attachments. Preserve a private original before removing anything because metadata cleanup can be destructive.
Filename and folder name are not embedded metadata, but they are still disclosed when you attach or upload a file. Rename the shareable copy if the name contains a person, account, project, location, or internal status.
A fresh export limits inherited information
FreeRedact builds a new image or image-only PDF from the visible pages you reviewed. It does not copy the original image EXIF block or PDF properties, comments, forms, attachments, and layers into the new file. This is broader than changing only one author or GPS field.
Use the source application's own inspection tool when you must preserve editable text, tags, forms, layers, or other document capabilities. Microsoft, Adobe, Apple, and other vendors provide format-specific controls that may better fit that requirement.
Metadata removal does not anonymize the content
The visible image may still show a street sign, house number, badge, face, reflection, screen name, or distinctive workspace. A document may still identify its author through writing, letterhead, signatures, or unique facts. Treat metadata cleanup and visible redaction as two separate review passes.
Tested workflow
Inspect and record non-sensitive test fields
On a copy, note the filename and any author, title, software, date, location, comment, attachment, or form entries you can see.
- A private original is preserved
- The test fields are harmless
- The intended recipient needs are clear
Create a sanitized copy
Review visible content, apply necessary redactions, and download a new image or flattened PDF. Give it a neutral filename.
- Visible identifiers are handled separately
- The export has a new name
- Only the export will be shared
Inspect the exported copy
Repeat the operating-system and application property checks. For important shares, inspect with a second application or a local metadata viewer you trust.
- GPS and author fields are absent
- No comments or attachments remain
- Filename reveals no private context
Example: Synthetic site-inspection photo
A fictional JPEG shows a blank demo equipment label. Its test metadata includes author “Morgan Vale,” location “Demo Campus,” and software “Example Editor.”
The fictional source contains
DEMO UNIT 4821Inspection complete
Redact
- DEMO UNIT 4821 if it identifies an internal asset
- Any badge, face, map, or address visible in the photo
Keep when needed
- Inspection complete when it is the point of the image
Expected result: The new image has only the reviewed visible pixels, a neutral filename, and no inherited author, GPS, caption, or editor fields in ordinary property inspection.
Verification checklist
Run these checks against the downloaded file, not only the editor preview.
- Inspect the downloaded file's details in the operating system.
- Check the photo information/map panel or PDF document properties in a second application.
- Look for author, title, subject, keywords, GPS, device, software, creation date, and modification date.
- Check comments, forms, links, layers, previews, and attachments when the format supports them.
- Review the filename and any accompanying cloud-link title or description.
- Inspect the visible content separately for faces, signs, addresses, reflections, and distinctive surroundings.
Limitations and decisions that remain yours
- Different applications display different metadata fields; absence in one properties window is not a universal proof.
- Rebuilding a PDF as page images removes useful accessibility and interactive features.
- Cloud platforms and messaging services may add their own dates, labels, account names, or sharing metadata.
- Removing metadata cannot undo a previous share or conceal identity revealed by the visible content itself.
Sources and further reading
These sources support the file-format and privacy practices discussed above. Product-specific behavior is described from FreeRedact’s documented workflow and synthetic tests.
Related guides and tools
Ready to check a file?
Open FreeRedact, review every suggestion, add anything the scan missed, and inspect the downloaded copy before sharing. No signup is required.
Remove source metadata while redacting a PDF