Mild softness sharpens well by hand
Photos that are slightly soft from a basic lens, a small focus miss or light resizing respond well to a gentle unsharp mask. This is the ideal case for traditional sharpening.
A LIGHTER TOUCH
Want to unblur an image without AI? Traditional sharpening adds definition to soft edges while keeping every pixel your camera captured. Here is how to do it by hand.
Start with the original file for the best result.
or drag and drop it here
JPG, PNG or WebP · Up to 20 MBNon-AI sharpening is not built into this page. Sharpen by hand with the steps below, or use the AI tool on the homepage.

01
Non-AI sharpening is arithmetic, not guesswork. A filter such as unsharp mask compares each pixel with a slightly blurred copy of the photo and raises the contrast wherever the two differ, which is exactly where edges are. Nothing new is invented: no hair strands, letters or skin texture are drawn in. That makes the traditional approach predictable and honest, and it is why editors, journalists and archivists still prefer it for pictures that must stay faithful to the original.
The trade-off is reach. A manual pass can make mild softness disappear, but it cannot rebuild detail the camera never recorded. When a photo is badly out of focus, an AI model that predicts missing texture, like the one in our unblur image tool, usually goes further, at the cost of adding details that were not really there.
02
Start with realistic expectations. An edge can look crisper without recovering the detail that was lost.
Photos that are slightly soft from a basic lens, a small focus miss or light resizing respond well to a gentle unsharp mask. This is the ideal case for traditional sharpening.
Long streaks and double outlines spread each edge over many pixels. Edge contrast alone cannot pull them back into one line, and pushing harder only creates halos.
A filter cannot establish what unreadable letters, distant faces or tiny patterns originally looked like. If the information is gone, a non-AI method will not bring it back.
03
You do not need paid software. Each editor below has a classic sharpening filter, and the settings are a safe starting point for a 12-megapixel photo viewed on screen.
Open Filters → Enhance → Sharpen (Unsharp Mask). Try a Radius of 1–2 and an Amount of 0.5–1.0, leaving Threshold near zero. Switch the before-and-after view on and off, and zoom to 100% before you apply it.
Use Filter → Sharpen → Unsharp Mask with Amount 80–120%, Radius 0.8–1.2 px and Threshold 2–3. In Photoshop, Smart Sharpen set to Remove: Lens Blur is the closest classic option for slightly missed focus.
Duplicate the layer, apply High Pass at 1–3 px, set the blend mode to Overlay or Soft Light and lower the opacity. Because it sits on its own layer, you can mask it away from skin and sky.
04
Three numbers control almost every traditional sharpening filter. Understanding them saves a lot of trial and error.
Radius is how far around each edge the extra contrast reaches. Use 0.5–1 px for fine texture and text, 1–2 px for portraits and landscapes. Go higher and you get thick, glowing outlines.
Amount is how much contrast is added. Raise it slowly while zoomed to 100%, and stop the moment edges start to look drawn on.
Threshold skips pixels whose difference is small, so skin, sky and noise are left alone. A small threshold is the easiest way to sharpen a face without making pores and grain louder.
05
Most failed sharpening comes from the same few habits. Look for halos, the bright or dark rims along high-contrast edges such as a roofline against the sky. Watch the noise, because sharpening amplifies grain as readily as detail, so reduce noise first on low-light photos. Avoid sharpening twice: a phone has usually sharpened the file already, and each extra pass multiplies artifacts. Finally, sharpen last. Resize to your final dimensions first and then sharpen for that size, since detail that looks right at 100% can turn harsh after downscaling.
06
Each type of picture has a different weak spot, so adjust the traditional approach to the subject.
Use a mask or a threshold so eyes, lashes and hair get the sharpening while skin stays smooth. A manual method never changes someone’s face, which is why many portrait photographers prefer it.
Scan at 600 dpi or higher, remove dust first, then apply a small radius. Film grain is real texture; sharpen it lightly instead of trying to erase it.
GIMP opens every GIF frame as a layer. Apply the same settings to each frame and keep them gentle, because the 256-color palette turns heavy sharpening into speckles.
Convert to grayscale, raise contrast with Levels, then sharpen with a radius under 1 px. For badly blurred wording, the blurred text image page covers better options.
07
Choose traditional sharpening when accuracy matters more than impact: evidence, archives, product shots and any picture where invented detail would be a problem. Choose AI when a photo is badly soft or small and you want the most readable result; the unblur image tool sharpens and enlarges up to 4× in one pass.
If the photo is sharp but simply too small, an image upscaler is the better fit. For a walkthrough of every method, from Canva to Photoshop, read how to unblur a picture.
GOOD QUESTIONS. CLEAR ANSWERS.
A few things to know before you begin.
No. Traditional sharpening changes edge contrast; a learned restoration model estimates a clearer image from patterns in its training data. Neither can guarantee that lost information is recovered accurately.
UnblurPic’s own processing uses an AI model. For a non-AI result, follow the steps on this page in a free editor such as GIMP or Photopea. The photo check above confirms your file’s format and size first.
Not reliably. If the source no longer contains distinguishable letter shapes, increasing sharpness does not establish the original words.
GIMP and Photopea both include unsharp mask and high-pass techniques at no cost. GIMP runs on your computer and Photopea runs in a browser; both give you the same core controls of radius, amount and threshold.