PDF for researchers
For blind submission, anonymising the document properties is not enough.
What carries your name without you noticing
Document properties are the obvious part: author, title, producer, the software you used. Most people know to clear those.
The part that catches people is that embedded images carry their own metadata. A figure exported from your analysis software can hold your username, your machine name or a file path with your name in it, and clearing the document properties does not touch it.
The file name travels with the submission too. Smith_NatureComms_v12_FINAL.pdf is not anonymous no matter how clean the metadata is.
On top of that, most venues have a size limit and a minimum resolution for figures, and both are checked at upload.
The tools for it
- Remove PDF metadataStrip the author, software and timestamps a file carries without showing them.
- Sanitize a PDFStrip the script, triggers and launch actions out of a PDF you were sent.
- Check a redactionFind out whether the words you blacked out can still be pulled out of the file.
- Print checkFind out why a print shop would reject your PDF — before you send it.
- Compress PDFMake a heavy file small enough to email, without wrecking it.
What this does not do
We can strip metadata and show you what remains, but we cannot know what counts as identifying in your field. An acknowledgement section, a self-citation phrased as “our earlier work”, or a distinctive method can identify you as clearly as a name, and no tool can catch that.
Check the text yourself. Our part is the machine-readable layer — the properties, the embedded metadata, the attachments and the annotations that most people forget are in there.
In every case above, the document is processed without ever being transmitted. Nothing is uploaded, which for confidential filings, student records, unpublished manuscripts and salary data is not a convenience but the point.