How to blur a face, a number plate or a whole image
Blur or pixelate part of a photo before you share it, why a black box in a viewer is not enough, and how strong the effect has to be to actually hide text.
· 6 min read
Blurring part of a photo replaces the pixels in that area: what was there is written out of the file, not covered up. That is the whole point, and it is what separates a real blur from the black rectangle people draw over a screenshot in a viewer, which leaves the original underneath for anyone who moves the shape.
The quickest route here: open Blur Image, drop the photo, switch What to blur to Selected areas, drag the box over the face or the plate, resize it by its corner handle, add more boxes with Add an area, then pick Pixelate at a high strength and click Apply effect. Nothing is uploaded — the picture is processed on your own device, which is the point when it is a photo of your children or a screenshot of your bank.
Blur the whole picture, or only the parts that matter
| What you want | Setting | Why |
|---|---|---|
| A soft background behind text | Whole image, blur, high strength | Nothing needs to stay readable |
| A face, a plate, a name, a password | Selected areas, pixelate, high strength | The rest of the photo stays useful |
| A screenshot you want to show the layout of | Selected areas over each piece of data | A reader can still follow the interface |
| A low-detail placeholder image | Whole image, blur | It is meant to look like a preview |
In Selected areas mode each box is processed together with a margin of the image around it, so the blur has real pixels to sample at its edges. Without that margin the edge of the box fades towards transparent and the rectangle announces itself; with it, the blurred patch sits in the photo instead of on top of it.
Blur or pixelate?
Blur looks softer and suits a background. Pixelate is the better choice over anything you are hiding on purpose, for two reasons. It destroys more detail at the same strength, and it reads as a deliberate act: nobody looking at your photo wonders whether you simply missed the focus, so nobody squints at it trying to work out what is behind it.
Strength runs from 1 to 50 and is scaled by the size of your picture, so the same number looks much the same on a 600-pixel thumbnail and a 4000-pixel photo. For something that must stay hidden, go high — 30 and above — and then look at the result at full zoom before you share it.
How strong is strong enough
This is where people get caught out, and it is worth being precise rather than reassuring.
The pixels really are replaced, so there is no sharp copy of the face hidden inside the output file. But a weak effect over text can sometimes be undone. The attack is simple: the attacker knows the font and the layout, renders every plausible word, blurs each one with the same settings, and compares. Against a six-digit number or a short name that is quick work, and it has been done to published documents more than once. Faces are harder — there is no small list of candidates to test — but a light blur over a face still leaves the hair, skin tone and head shape, which can be enough to recognise someone in context.
So, in order of how well they hold:
- Crop it out. If the thing you want gone is at the edge of the frame, Crop Image removes it from the picture entirely. Nothing survives what is not in the file.
- Pixelate at high strength, then check at full zoom that no edge of a letter is still legible.
- A heavy blur, which looks tidier but destroys less structure at the same setting.
- A light blur or a mosaic with large visible blocks, which is decoration. Do not use it on a password, an account number or an address.
What blurring does not do
It does not remove the location of the photo. GPS coordinates, the camera model and the date sit in the file's metadata, not in its pixels. You can blur every face in a picture and still publish the street it was taken in. Run the file through Remove Image Metadata as well before posting it anywhere public.
It does not help with a PDF. A blurred rectangle over text in a PDF leaves the text in the page content, selectable and copyable. Redact PDF removes the characters themselves; the guide How to delete pages from a PDF covers when to remove a whole page instead.
It does not survive being the only copy. The tool writes a new file and leaves your original alone — which is what you want, until you send the wrong one. Rename the blurred version as soon as it lands in your downloads.
Common problems, and what they mean
The blurred box has visible hard edges. That usually means the area sits right at the edge of the picture, where there is no surrounding image to sample. Make the box slightly smaller than the edge, or crop the edge away instead.
The face is still recognisable. Raise the strength, switch from blur to pixelate, and make the box bigger — hair and the shape of a head carry more recognition than people expect, so cover the whole head rather than just the face.
The file came out much larger than the original. A PNG of a blurred photo can be bigger than the sharp JPEG it came from, because PNG stores every pixel exactly. Compress Image brings it back down, or save as JPEG in the first place.
The colours changed slightly. JPEG output is re-encoded at quality 92, which is visually lossless at normal viewing size but not bit-for-bit identical. A PNG or WebP source keeps its format and avoids that.
It is slow on a big photo. Everything runs in your browser, so a 40-megapixel image is real work for the device. It finishes; give it a moment, or shrink the photo first if you were going to anyway.
Questions people ask
Can I blur just one face and leave the rest sharp? Yes. Switch to Selected areas and put a box over each face. Only what is inside a box is changed.
Is the blur reversible? Not from the output file — the original pixels are not in it. What can sometimes be reversed is the content of blurred text, by guessing and comparing, which is why strength matters on text and why cropping is safer still.
Are my photos uploaded? No. The effect is applied in your browser, so the photo never leaves your device and nothing is stored on a server.
Does it work on a phone? Yes, and dragging the boxes with a finger works as well as with a mouse. Very large photos are slower on a phone.
How many areas can I blur at once? As many as you like, and they can overlap.
Can I blur several photos in one go? In Whole image mode, yes — drop a batch. Selected areas works on one photo at a time, because the boxes have to go in the right place on each one.
Is there a watermark? No watermark, no account, no limit on the number of images.
Next steps
- Take the information out of the frame entirely: Crop Image
- Strip the GPS location and camera details: Remove Image Metadata
- Hide text in a document instead of a photo: Redact PDF
- Shrink the result for an upload: the guide How to reduce image file size
Tools used in this guide
- Blur ImageBlur a whole image, or just the parts you pick.
- Crop ImageCut an image to the area and aspect ratio you need.
- Remove Image Metadata (EXIF Remover)Strip GPS location and EXIF data from photos.
- Redact PDFHide names, numbers and addresses in a PDF for good.
- Compress ImageShrink JPG, PNG and WebP images without visible quality loss.