Blur vs black bar on a screenshot

Blur is a filter. A black bar is a decision that those pixels get no future.

Use blur when the region is a face in a crowd and you only need “not identifiable at a glance.” Use a bar when the region is a secret you would rotate: password, recovery code, card, phone, session cookie.

Product teams like blur because it keeps layout. The screenshot still looks like the product. Attackers do not grade your aesthetic. They try contrast. PETS 2016 documented that mosaicing and blurring over document text often leave enough structure to recover characters. That paper is about printed redaction, not a consumer “unblur” site. The operational lesson for a PNG is the same: leftover structure is leftover secret.

Why does blur look more professional than a bar?

It preserves the silhouette of the UI. Marketing decks, bug reports, and App Store reviews all want the reader to recognize the screen. A black slab says “we hid something” and some teams hate that. Hide it anyway when the something is a token. You can keep the rest of the layout. You do not owe a stranger the glyphs.

iPhone highlighter is worse than blur. It is often a tint. Raise brightness and brilliance in Photos and the words can return. That is documented, not folklore: 9to5Mac in 2018, then Lifehacker telling people to stop using the highlighter as a privacy tool. A filled shape in Markup is closer to a bar. A translucent marker is not. Markup fail.

When is a light mosaic as bad as a polite blur?

When the alphabet is small and the string is short. Six SMS digits. Four digits of a card if the rest of the plastic is in frame. A 2FA backup code printed in a monospace font. Mosaic cells still follow character shape. Blur still leaves stems. Hill, Saorma, and Saul were not writing a how-to for criminals. They were measuring a redaction habit that looks finished in a PDF or a scan and is not.

A heavy mosaic over a long paragraph can still be a poor choice if you needed deletion. A solid fill is simpler to audit: zoom in, look for peeking pixels, look for anti-aliased crumbs at the edge of the box. If you can still name the letter, the box is too small or too transparent.

What does JPEG do to a blur you already applied?

Chat apps recompress. Ringing around a light blur can outline glyphs. Start with PNG so you are not adding your own lossy pass on top of theirs. A hard black edge can ring too, but the interior of the bar stays empty. A light smear has interior structure that compression can emphasize. PNG note.

Do not screenshot the editor window to “save” the result. That is a second resample. Use the download control. Do not flatten by sending through an app that turns PNG into a sticker with extra compression unless that is the only channel you have — and even then, the source file should still be the barred PNG, not a JPEG of a blur.

Can you mix blur and bars in one screenshot?

Yes. Faces in a standup photo: blur if policy allows it. The Slack sidebar with customer email: bar. The URL with token=: bar the chrome. The notification that landed mid-capture: bar the banner or recapture. Mixing is not a contradiction. It is matching the tool to the failure you care about. Glance-level identity is not the same as a rotatable secret.

If you are unsure, bar. You can still describe the UI in the ticket text. You cannot un-send a PNG that still has a session cookie in the address bar.

How do you check the bar actually replaced pixels?

Open the export, not the layered editor file. Zoom. Invert. Nudge brightness. If glyphs return, you used a overlay or a highlighter. If the region is uniform, you replaced samples. Keep the original off the thread. Rotate anything that was ever visible on a file you posted, even if you think the bar covered it.

Both modes live in the same tab. Drag a box, pick a heavy smear or a fill, download PNG: run blur and bar in BlurThis.