Clearing the author and company fields takes one click and solves one problem. A PDF also keeps the review comments, the earlier draft embedded inside it, the invisible text layer and the edit timeline. Wiping the properties dialog leaves all of that intact. This removes the whole set, then proves it is gone.
Metadata and hidden-object cleaning runs locally in your browser. No upload required.
The word is used loosely. In a PDF there are distinct places data lives, and a tool that clears one has not touched the others.
The standard fields: who created it, which organisation, which application produced it, and the creation and modification timestamps. This is what most metadata removers clear — and all they clear.
Many PDFs carry an XMP stream alongside the classic info dictionary. Clear one and the other still answers queries about the author, the source tool and the edit history.
Stored as separate objects, not as page text. A file that looks clean can still carry every margin note, including the ones asking whether a clause should be softened.
Attachments and embedded documents travel with the PDF. The version you sent may still contain the version you rejected.
Text placed outside the visible page box, or rendered in the background colour, is invisible on screen but fully extractable by any parser.
Dark watermarks are low-contrast markings embedded for traceability. They indicate the file is tracked to a recipient, which matters before you forward it.
Removing properties is step two of five. The reason it matters is everything around it.
See what the file carries before you decide what to remove. Detection reports the object references, not a yes/no guess.
Metadata, XMP, comments, hidden objects and embedded files removed at object level — not covered over.
Permanently delete the sensitive text itself where the content, not just the properties, is the problem.
Add review notes, stamps or Bates numbers without reintroducing the metadata you just removed.
Re-parse the output and confirm the fields are empty. A cleanup you cannot verify is a cleanup you have to assume.
You do not have to trust a tool's own report. Any clean PDF should return nothing from these two commands.
# what properties remain?
pdfinfo cleaned.pdf
# is the "deleted" text still extractable?
pdftotext cleaned.pdf - | grep -i "confidential"If the second command prints anything, the text was covered, not removed. That is the single most common failure in document cleanup: a black rectangle over the words, with the words still underneath and still selectable.
Drawing a shape over text, or setting it to white, leaves the underlying text object in the file. It can be selected, copied, searched and extracted. It also survives most format conversions.
Deleting the text object itself means there is nothing left to recover. The correct output fails the extraction test above — which is exactly how you verify it.
No. Metadata is one of six places a PDF holds information. Clearing the properties dialog leaves comments, embedded files, hidden objects, revision remnants and watermarks untouched. If you only do one, you have reduced the surface, not closed it.
No. The visible pages are unaffected. Only the non-visual data is removed.
Yes. Metadata removal is a standard operation. PDFnoted re-parses the output after writing to confirm the file opens and the fields are empty before you download it.
Metadata and hidden-object cleaning runs locally in your browser, so no upload is required. Where a feature genuinely needs server-side processing, that is stated before you confirm it.
Yes. Batch cleanup applies the same removal across a set of documents and reports per-file results, which matters when a disclosure involves hundreds of files.
Property and object cleanup works on any PDF. If a scan has no text layer at all, OCR is needed first before text-based checks are meaningful.
Remove metadata, comments, hidden objects and embedded drafts in one pass — and verify the output instead of hoping.
Clean a PDF now