Make the text in a photo readable

Every step runs on the corrected picture after the measured light pass, and every one is a switch. Clarity and colour boosts are off in the Document preset, because local contrast on a page draws haloes around letters.

The most common reason text in a photo is hard to read is not focus but light: a grey page with a darker corner and a yellow cast from indoor light. Flattening removes the gradient, text contrast removes the grey, and the measured white balance removes the cast. If the photo is also soft, the unblur route's deconvolution can be switched on in the same Clean-up panel.

Sign on plain white paper with a dark pen, photograph it flat in daylight and crop close. Document mode with Ink to black turns the paper white and the ink black, and the Target panel fits the result to the portal's pixel and KB band, measuring the saved bytes.

• No OCR, no page-edge detection and no perspective correction in this release.
• Very faint pencil can be lost by the black-and-white threshold; use Keep greys for it.

Does it read the text (OCR)?

No. It makes the picture of the text easier to read — by you, or by whatever reads it next. It does not turn it into editable text, and it does not detect the page edges or straighten a photo taken at an angle.

What does flattening the paper do?

A phone photo of a page is brighter where the light falls and darker where your hand or phone shadowed it. Flattening estimates the paper's brightness everywhere — by looking for the brightest surface around each point, wider than any letter — and divides it out, so the whole page comes out the same white.

Keep greys or clean black and white?

Keep greys for pages with photos, stamps or pencil; it keeps every tone and just makes paper white and ink dark. Clean black and white for printed or typed text; it decides per pixel using the local brightness and contrast (Sauvola's method), which copes with faint text and shadows far better than one global threshold.

Can I use it for my signature on a form?

Yes: turn on Ink to black, then use the Target panel to fit the portal's pixel and KB limits. If the ink is mostly blue the page says so — a form that asks for black ink is asking about the pen you signed with.

Why does a screenshot behave differently?

A screenshot has no paper, no shadow and no sensor grain, so flattening and denoising would only damage it. When the picture has very few distinct tones it is treated as a screenshot: text contrast only, and the Size panel offers a pixel-exact enlargement that keeps text edges hard.

• Form photo enhancerFix the light, then fit sourced portal pixel and KB limits, measured on the saved file.
• Increase image size in KBQuality first, then pixels, until the file clears the minimum. Never padded.
• Unblur imageA measured deconvolution for slight softness — and a plain warning when the blur is motion.
• Improve image qualityA quality report first: grain, JPEG quality, blocking, edge energy, clipping and resolution, each with its fix.