How to Upscale a Photo Without Losing What's Real

You found the perfect photo — and it's 800 pixels wide. It needs to be a print, a banner, a full-bleed hero. Drag the corner to enlarge it and you get the familiar disappointment: soft edges, mushy text, skin that looks like wax.

Upscaling done badly is worse than not upscaling at all. Done well, it's the difference between a usable archive and a dead one.

Why naive resizing fails

When you stretch an image, the software has to fill in pixels that were never captured. The old approach — bicubic resizing — just averages neighbouring pixels. It can make an image bigger, but it cannot make it sharper. Every edge softens, because an average is always blurrier than the things it averages.

That's why a naively enlarged photo looks "off" even before you can name the problem. There is genuinely less information per pixel than your eye expects at that size.

What good upscaling actually does

Modern AI upscaling doesn't average — it reconstructs. Trained on millions of image pairs, it has learned what a sharp eyelash, a crisp brick edge, or a clean letterform looks like, and it rebuilds those structures at the new resolution.

The important word is plausible. The model isn't recovering the exact pixels that were lost — those are gone. It's generating detail consistent with what the photo shows. For a landscape or a product that's exactly right. For a face or a document, it asks for judgement.

Upscaling restores sharpness, not facts. It will not recover a licence plate that was four pixels wide. Treat the result as a faithful enlargement, not new evidence.

Match the strength to the photo

Not every photo wants the maximum setting:

• Light — a gentle pass for a photo that's already decent and just needs more pixels. Safest for faces.
• Crisp — the everyday default. Noticeably sharper, still natural.
• Full — the strongest reconstruction, for small or soft sources headed for large prints. Check faces carefully; aggressive upscaling can drift a likeness.

When in doubt, go one step gentler. An honest, slightly-soft enlargement beats an over-sharpened one that looks artificial.

Where upscaling earns its keep

• Old digital photos from early phones and compact cameras, shot at tiny resolutions.
• Product shots that need to work as both a thumbnail and a zoom.
• Cropped images — once you crop in, you've thrown away pixels; upscaling buys some back.
• Print — screen-resolution images that need 300 DPI on paper.

The one rule: start from the best source you have. Upscaling amplifies what's there — including compression artefacts and noise — so a clean small photo beats a damaged larger one.

Open Photo Upscale and start with the image you most wish were bigger.