Choosing a text-covering method
Blur, Pixelation, or a Solid Box for Private Text?
For text you do not want a recipient to read, prefer a fully opaque cover rendered into a separate image over light blur or pixelation. Softening can leave clues to the characters. Inspect the exported image, mask edges, and repeated text; a box visible only in an editable project is not the same as a prepared image attachment.
By Qingxin Studio, the developer of HushPic · Updated
Match the method to the disclosure
Softening a background and concealing an account number are different tasks. Ask whether you are making text less noticeable or deliberately removing its readable representation from the sharing image. A visually attractive blur is not evidence that a short private string is protected.
The 2016 PETS study on mosaicing and blurring demonstrated text recovery risks under its tested conditions. It supports caution about remaining character structure, not a claim that every blurred image can always be reconstructed. Do not rely on softness as the protection for sensitive text.
Check fill, coverage, and edges
In a local editor, choose a filled, fully opaque shape rather than a translucent highlighter or outline. Cover the complete text with a little margin so the first character, descenders, or end of a line do not remain visible. Inspect tilted and wrapped text separately.
Look for duplicate values in headings, confirmation fields, QR codes, tooltips, or another screenshot. Covering only half a token intentionally leaves information visible. Make that disclosure choice deliberately rather than calling any partial mask complete.
Export pixels, not an editable cover layer
Save a separate image export and reopen it outside the editor’s working view. Confirm the cover remains in the image you will attach. Keep the original and editable project private; do not include them as additional files alongside the export.
A flattened raster image differs from a presentation, layered project, or PDF with a shape over live text. This article concerns supported image files, not PDF sanitization or every embedded preview. An opaque-looking image still needs checks of its file and your sharing choices.
Inspect the saved image at reading size
Zoom into the saved output and examine every region’s boundary. Check for exposed letters, weak opacity, a line outside the cover, or an export that restored earlier content. A tiny attachment thumbnail is not a sufficient coverage check.
Review neighboring labels and context. Hiding a name while leaving a unique profile or identifier may still identify someone. Solid fill covers selected pixels; it does not guarantee that a document or conversation becomes anonymous.
Keep metadata verification separate
Covering text does not remove hidden photo fields. Our browser checker reads supported metadata categories but does not look at pixels or create redacted copies. Use it only for that separate question, respecting its parsing and format limits.
HushPic’s on-device flow applies suggested masks and clears supported private metadata on a separate copy. Open Review / Adjust for missed regions, regenerate after changes, and inspect the current result. Suggestions may miss details; output verification is bounded by supported checks, not a certification that all sensitive text was found.
Sources
This guide is published by Qingxin Studio, the developer of HushPic. Product claims describe our supported workflow, not independent certification. Our review checklists are practical first-party guidance; linked platform sources support only the mechanisms they describe. Automatic checks may miss details. Read our content and corrections policy.