Can blurred text be unblurred? Why we use solid covers

Can blurred text be unblurred? Sometimes, yes. A blur or a pixelation doesn’t remove what it hides; it rearranges it. The pixels you see are still made from the original ones, and that is enough for someone to work backward. That is why Dress Docs has no blur and no pixelation at all. A solid cover keeps nothing of what is underneath.

Can blurred text be unblurred? Blur keeps part of the original

A blur replaces each pixel with a mix of its neighbors. A pixelation replaces each block of pixels with their average color. Either way, the result is calculated from the original. The detail is smeared, not gone.

For short, predictable text, such as an account number, a date or a name from a known list, someone can blur every candidate the same way and see which one matches. A light blur can sometimes be sharpened back into readable text.

Official guidance says the same. The UK’s ICO, in its guidance on redacting information: “It might be possible for someone to reverse redactions made using simple image editing tools to blur or pixelate information.” The NSA’s guide Redacting with Confidence (2006) puts the rule in one line: sensitive information must not be “just visually hidden or made illegible, but is actually removed”.

In a PDF there is a second problem. A blur effect is often just one more layer on top of the page, with the original text still underneath, the same flaw as a black box drawn over a PDF.

Why a solid cover is the only cover

Dress Docs covers are solid shapes, rectangles or ellipses, in solid black by default. On export, each page becomes one flat image with your covers in it. The pixels under a cover are replaced by the cover’s color. Nothing of the original is left to calculate from, and there is no text layer to copy.

This matches the ICO’s method for images: mask the information “by obscuring part of an image with a solid block of colour”, then export it to a simple image format. Dress Docs exports to PDF, PNG or JPG, with the page flattened every time. See everything the export does on the features page.

Screenshots and photos: same rule

A screenshot of a chat with a phone number smudged out, or a photo of a letter with the address blurred, has the same weakness. If the hidden part can be guessed from what is left, it isn’t hidden. Cover it with a solid shape instead.

Photos carry a second risk: the camera writes location, device and time into the file. The Dress Docs export starts from a fresh image, so none of that comes along. More on that in the metadata you forget to remove. On a phone, the free app covers photos of documents the same way: redact an ID photo on your phone.

Check it yourself

For the full method, start with why a black box isn’t enough. Or open Dress Docs and cover your next document with something that holds.