Sharpen Image

This tool sharpens a PNG, JPG, WebP, GIF, BMP or AVIF image in your browser using an unsharp mask — the same technique photo editors call "Sharpen". A blurred copy of the picture is compared with the original, and the difference — the contrast the blur threw away — is added back, scaled by the amount you choose. Where the picture is flat the difference is zero and nothing changes; at an edge the lighter side is pushed lighter and the darker side darker, which is what reads as crispness. It is for making a soft scan or a resized photo look more defined, for putting a little bite back after an image has been shrunk, or for a product and social photo that needs to pop next to others.

The most important thing to understand is what sharpening is not. It cannot recover detail that was never captured. A blurred, out-of-focus or heavily compressed photo stays blurred — sharpening exaggerates the contrast that is there, so a soft edge becomes a more contrasty soft edge, not a truly sharp one. It also amplifies whatever noise and JPEG blocking is already in the picture, because those are small details too. Sharpening is contrast, not new information; a genuinely sharper large picture can only come from a sharper or larger original. This tool says so on the page rather than selling a miracle.

Two numbers control the effect. The amount is how much of the difference is added back: 1 is the standard strength, below 1 is a gentle touch-up, and at an amount of 2 or more the bright rims along high-contrast edges — halos — become visible, so the tool warns you there. The radius is the size of the edges that get the boost, in whole pixels: 1 px works on the finest detail, while a larger radius (say 5–20 px) fattens edges and increases local contrast, which can look satisfying on hazy landscapes but is exactly what creates halos if pushed. For a normal photo, an amount around 1 and a radius of 1–2 px is a safe, natural starting point. (Unlike many editors, this tool has no threshold setting, so a smooth area is never left alone — keep the amount modest on noisy images.)

Because the maths runs on the raw pixels rather than a browser filter, the result is the same on every device. Your browser decodes each image (turning a phone photo upright with its EXIF orientation), draws it onto a canvas at full size, and rewrites the red, green and blue values in JavaScript. Transparency is never touched: the alpha channel is left exactly as it was, so a transparent PNG stays transparent. The canvas is re-encoded and offered as a download, and because the image is redrawn on a canvas, EXIF metadata — camera, date and GPS location — is not copied into the result.

"Same as input" keeps PNG as PNG, JPG as JPG and WebP as WebP. GIF, BMP and AVIF are saved as PNG, because browsers can read those formats but cannot reliably write them from a canvas, and an animated file keeps only its first frame. You can also pick PNG, JPG or WebP explicitly: PNG is lossless, so the sharpened pixels are saved exactly, while JPG and WebP are re-encoded at the quality you choose. JPG cannot store transparency, so transparent areas are filled with the colour you pick.

Limits: up to 50 images at a time, 50 MB each, and about 100 megapixels (16,384 px per side). iPhones and iPads allow a canvas of only about 16.7 megapixels, so a very large photo may fail there with a message saying so. Your images stay on your device: the Network tab in your browser's developer tools shows no request carrying them.

1. The file's size is checked (50 MB) before it is read, and its first bytes confirm it is PNG, JPG, WebP, GIF, BMP or AVIF; the browser then decodes it upright. 2. The image is drawn onto a canvas at full size and its RGBA pixels are read. 3. A blurred copy is made with the same three-pass box blur the Blur Image tool uses — three passes of a box blur of the radius you set, in px, horizontally then vertically, with edges clamped and colours handled premultiplied by alpha. It is a close approximation of a Gaussian blur. 4. Each red, green and blue channel is replaced, per pixel, by the unsharp mask formula: sharpened = original + amount × (original − blurred), rounded to the nearest byte and clamped to 0–255. A flat area has original = blurred, so it is unchanged; at an edge the values overshoot, which is the visible sharpening. 5. The alpha channel is never written, so transparency is preserved; only colour is sharpened. 6. The pixels are written back, the canvas is encoded as PNG, JPG or WebP, and the file is named like photo-sharpened.jpg or logo-sharpened.png.