Redact, flatten or print to PDF
Faced with a document that needs something removed, most people try one of five things. Two of them work, three of them do not, and the difference is not visible on screen. Here is what each one actually does.
Drawing a box, then flattening
Adding a black rectangle puts a drawing instruction on top of the text. Flattening then merges annotations into the page content — which sounds like it should help and does not, because the text operators are still in the content stream underneath the newly merged rectangle. In some tools flatten does rasterise, in which case it works; in others it only locks form fields. The word does not tell you which, and that ambiguity is why the method keeps failing for people who did everything their tool asked.
Print to PDF
Printing to a PDF re-renders the document through the print pipeline, which in many cases does discard annotations and produce a page whose content is only what was visible. It is genuinely better than flattening. But it is not reliable as a redaction method: whether the text survives depends on the printer driver, the viewer and whether the black box was drawn as an annotation or as page content. If it was page content, printing preserves both the text and the box on top of it. Sometimes safe, never certain.
Print and rescan
Printing the document on paper and scanning it back in genuinely destroys everything that was not visible: there is no text, no metadata from the original, no annotations, no layers. It is the most reliable low-tech method and it is why it is still used for sensitive disclosure. The costs are obvious — the output is an image, it is unsearchable, quality drops, it takes time, and if the scanner runs OCR you have quietly reintroduced a text layer that needs checking.
Rasterise the affected pages, then verify
What this app does is the digital version of print-and-rescan, applied only where it is needed. A page carrying a redaction box is converted to an image, the boxes are painted onto it, and the page is rebuilt from the image, so no text object survives underneath. Pages without boxes are copied across untouched and keep their real text. Then the output is reopened and searched twice — extracted text and raw streams — for the terms you removed. The verification matters as much as the method: it is what turns "this should be gone" into "this is gone, and here is the check".
| Method | Removes the text? | Notes |
|---|---|---|
| Black box on top | No | Text stays; copy-paste reveals it |
| Black or white highlight | No | Annotation over the page; sometimes removable |
| White-coloured text | No | Invisible, not absent |
| Crop the page | No | Hides only; crop can be undone |
| Flatten | Depends on the tool | May merge, rasterise or only lock fields |
| Print to PDF | Sometimes | Depends on driver and on how the box was added |
| Print and rescan | Yes | Also destroys metadata; output is an image |
| Rasterise affected pages | Yes | What this app does, then verifies the output |
Frequently asked questions
Is print to PDF good enough for redaction?
Sometimes, and you cannot tell which time it is without checking the output. If the black box was page content rather than an annotation, printing keeps the text under it.
Is printing and rescanning still a valid method?
Yes, and it is very thorough — it destroys metadata too. The costs are quality, time and an unsearchable result, plus an OCR layer if your scanner adds one.
What does flatten actually do?
It varies by tool: merging annotations into page content, rasterising, or just locking form fields. Only one of those removes text.
Why not rasterise the whole document?
Because it would throw away real text on pages that never needed touching, making the file bigger and unsearchable for no benefit.
Does rasterising hurt print quality?
Slightly, on the affected pages, since they become images at a fixed resolution. Pages that were not redacted are untouched.
How do I know which method my tool used?
Test it: redact something, produce the output, then select and copy the redacted area. Five seconds tells you what a paragraph of documentation will not.
Remove the text, then prove it is gone. Redact takes the words out of the page rather than covering them, then reopens the file it just made and searches it twice — by text extraction and by raw stream scan — before letting you download. The file never leaves your browser.
Redact a PDF in the browser