How to Unblur an Image (and When You Cannot)

First work out which kind of blur you have

Three completely different problems all get called blurry, and the fix is different for each. Two minutes identifying which one you have saves you an hour of trying tools that were never going to work.

Motion blur. Streaks that follow a direction. Either the camera moved or the subject did. Look at a hard edge: if it is smeared along one axis while other parts of the frame are sharp, the subject moved. If everything is smeared the same way, the camera moved.

Out of focus blur. Soft everywhere, no direction, and points of light become discs rather than streaks. The lens focused on the wrong distance.

Low resolution softness. Not really blur at all. The image is simply too small for the size it is being shown at, so everything looks mushy and edges look blocky when you zoom in.

The third is the most common and by far the most fixable.

What can honestly be recovered

Being straight about this saves disappointment.

Low resolution softness: usually yes. This is what AI upscaling is genuinely good at. The model has seen a very large number of real photographs, so it can make a well informed guess at what the missing detail should be and redraw it larger. Hair, fabric, foliage and skin texture come back convincingly.

Mild out of focus: sometimes. A slightly soft photo often sharpens up well. A badly defocused one does not, because the information about fine detail was never recorded.

Motion blur: rarely. The detail was smeared during the exposure. Sharpening a smear gives you a crisper smear. If the streaks are long, no tool available to anyone will bring that photo back.

The uncomfortable truth is that a lot of blurry photos cannot be rescued, and any tool promising otherwise is either sharpening edges to fake it or inventing detail that was not there.

Try it on your photo

The fastest way to find out which category yours falls into is to run it and look.

Drag the slider on the result. If the difference is real, you had resolution softness. If it barely changes, you are probably looking at motion blur, and the honest answer is that the photo is not recoverable.

Sharpening is not the same as unblurring

Worth understanding, because it explains why the sharpen slider so often makes things worse.

Sharpening does not add detail. It finds edges that already exist and increases the contrast on either side of them, which reads to your eye as crisper. Applied gently to an already good photo, it helps. Applied hard to a blurry one, it produces bright halos along every edge and makes the image look processed and worse, while adding no actual information.

AI enhancement works differently. Instead of exaggerating what is there, it reconstructs what was probably there at a higher resolution. That is why it can genuinely improve a soft photo where a sharpen slider cannot.

If your photo is already sharp and you just want more bite, use sharpening. If it is soft, use an image enhancer.

Get the best version before you start

Most people try to fix the wrong file.

Use the original, not a shared copy. Messaging apps recompress images. The version in your camera roll is nearly always better than the one that came back through WhatsApp or Instagram.

Do not crop first. Cropping removes pixels, and pixels are what the reconstruction has to work from.

Do not sharpen first. Sharpening exaggerates edges, and the AI will faithfully rebuild those exaggerated edges as though they were real.

Do not enhance twice. Running a result back through usually amplifies whatever the first pass got wrong.

If you need it bigger as well as clearer

Clearing up softness and adding real size are related but separate jobs. If the end goal is a print or a large screen, start with the AI image upscaler, which is built to add resolution rather than repair damage.

For prints, aim for roughly 300 pixels per inch at the finished size. A 12 inch wide print therefore wants about 3600 pixels across.

FAQ

Can you really unblur an image?

It depends entirely on why it is blurry. Softness from low resolution can usually be improved a great deal. Mild focus blur sometimes improves. Motion blur generally cannot be fixed by anything, because the detail was smeared while the shutter was open and was never recorded.

What is the difference between sharpening and unblurring?

Sharpening increases contrast along edges that already exist, which looks crisper but adds no new detail and produces halos if overdone. AI enhancement reconstructs detail at a higher resolution based on what the model has learned from real photographs, which is why it can improve a soft image where sharpening cannot.

Can I unblur a photo to read a number plate or a document?

No, and this matters. If the text is unreadable in the original, an AI tool will produce letter shapes that look like text but are not the real characters. The result will be confidently wrong rather than blank, so it should never be relied on to read something you could not already read.

Why does my photo look worse after I sharpen it?

Sharpening exaggerates the contrast at every edge, including noise and compression artefacts. On a soft or already compressed photo that produces visible halos and blotches. Enhancement is the better route for a soft image.

Is there a free way to unblur an image?

Yes. The tool on this page runs three images a day free, with no account, no email and no watermark on the result.

Related guides

• How to fix pixelated photos
• How to restore old photos