How to Blur Faces in Photos: Select, Check and Download
Choose which detected faces to cover
Process a photo locally in your browser, review the result and download a free PNG.
Before you download
To blur faces in a photo, choose the image, review the detected faces, apply a covering effect and inspect the downloaded result. A face without a detection box will not be covered just because you choose a stronger effect. Check who is selected as well as how the mask looks.
FaceBlurify's photo tool processes the image locally in your browser. The photo file is not sent to the server for processing, and the PNG download is free. Paid video credits are a separate workflow.
Blur faces in a photo, step by step
- Open the photo face-blurring tool and choose your original image. JPEG, PNG and WebP are practical choices for still photos. Keep the original separately so you can return to it if a correction is needed.
- Wait for detection to finish. Review the boxes against the photo itself: count foreground people, small background faces, profiles and reflections. A detected-face count is not confirmation that every face was found.
- Decide who should remain visible. All detected faces initially receive the effect. Clicking or tapping a face marks it to stay clear; tapping again restores masking. Check each face's selection state before continuing.
- Choose Blur, Pixelate, Black Box or Emoji. For Blur or Pixelate, open Advanced options to check the privacy level and intensity. High is a starting preset, not a guarantee that the result meets your needs.
- Select Apply effect and inspect the before/after comparison. Check the processed side at the face boundaries, not only the overall appearance of the picture.
- Select Download, then open the PNG on your device. Confirm that you saved the processed image and that everyone intended to be covered is covered. If it needs another attempt, return to the original photo and revise the selection or effect before sharing.
The current photo workflow exports PNG, including when the input is JPEG or WebP. It retains the source image dimensions after orientation handling; it does not recreate detail missing from a small or compressed original.
Keep one person visible in a group photo
Suppose a group photo contains six people and only the presenter should remain visible. First check whether all six have detection boxes. Leave the five other faces selected for masking, then tap the presenter to keep that face clear.
After applying the effect, inspect all six faces again. This catches two different mistakes: accidentally leaving another detected face clear, and overlooking a face that was never detected. The selection controls operate on detected faces; they are not a way to draw a new mask over a missed face.
For an image intended for a public page, also consider whether the caption, a name badge or a recognizable setting identifies someone whose face you covered. Face selection alone cannot resolve those details.
Choose an effect by inspecting the result
| Effect | What it changes | What to inspect |
|---|---|---|
| Blur | Softens detail inside the detected region | Whether facial features remain recognizable, especially at larger viewing sizes |
| Pixelate | Replaces detail with blocks | Whether the blocks obscure useful facial detail and the region covers the whole face |
| Black Box | Covers the detected region with a solid mask | Whether any face edge or uncovered area remains outside that region |
| Emoji | Places a graphic over the detected face | Whether the graphic's shape and edges actually cover what you intend to hide |
A solid covering removes visible detail within its area, but the area still needs to be correct. Blur and pixelation depend on their settings and the particular photo. There is no universal setting that makes every image anonymous, and an emoji is not a privacy certification.
Judge the result at the size people will see it and at a larger size if the original download will be available. A face that appears obscured in a small thumbnail may still show useful detail when enlarged.
What to do when a photo face is missed
If a visible face has no box, changing intensity will affect other detected faces rather than create the missing selection. Start by checking whether the source image is correctly oriented and whether a clearer original is available. Small faces, strong side angles, coverings and poor lighting can make detection difficult.
If the same face remains undetected:
- Crop that person out when doing so preserves the purpose of the image, then inspect the crop.
- Use an image editor that lets you draw a manual opaque mask over the whole face.
- Choose another photo or keep the image private until the correction is complete.
Do not assume the video's detection-sensitivity control exists in the photo tool. For moving footage, use the separate missed-face video troubleshooting guide.
Check more than the face mask
A useful final review covers the image and the way it will be shared:
- Every intended face: scan the foreground, background, mirrors and screens.
- Selection mistakes: verify that only the intended people stayed clear.
- Other identifiers: inspect badges, documents, addresses, distinctive clothing and the caption.
- File details: choose a neutral filename if the original name reveals personal information. The download name can retain the original filename.
- The shared copy: attach the processed PNG, not the original. Check the uploaded version if the platform resizes or recompresses it.
The photo export creates a new PNG without carrying over the input's EXIF metadata. This does not remove text visible in the image, information in its filename or details a publishing platform adds separately.
For a sequence of moving images, use the video anonymization checklist instead of assuming one reviewed frame validates a whole clip.
Frequently asked questions
Can I blur just one face and leave the others visible?
Yes, among the faces the tool detects. Keep the intended face selected for masking and mark the other detected faces to remain clear. Verify every selection in the processed result.
Are photo processing and download free?
Yes. The photo workflow runs locally in your browser and produces a free PNG download. The sign-in and paid-credit requirement for watermark-free video exports does not apply to this photo workflow.
Does a successful download mean all faces were detected?
No. Download success confirms that an output file was produced. You still need to compare the actual people in the image with the covered regions.
Can I use the photo tool without any internet connection?
Do not rely on that. The image is processed locally, but the page and detection assets must be available in your browser. Local processing is not the same as a guarantee of offline availability.