How it works
The Image Blur tool applies a Gaussian blur to soften an image, reduce noise, create depth-of-field effects, or obscure sensitive visual information. The blur radius controls the intensity — radius 1 is a subtle softening, radius 20 is a strong frosted-glass effect, and radius 50+ creates an almost entirely unrecognizable blur.
Blur has both creative and functional uses. Photographers use shallow-depth-of-field blur to separate subjects from backgrounds. UI designers use blurred images as frosted-glass backgrounds in cards and modals. Privacy-conscious users blur faces or sensitive areas before sharing screenshots. Marketing teams blur upcoming product images to create teaser campaigns.
How to use it: upload your image. Set the blur radius using the slider (1–100). The preview updates in real time. For large images, the blur may take 0.5–2 seconds to compute. Download the blurred result.
Gaussian vs. box blur: Gaussian blur produces a naturally smooth, bell-curve-weighted blur that resembles defocus in real cameras — it's the perceptually correct blur. Box blur (a simpler average of neighboring pixels) is faster but produces blocky artifacts at high radii. This tool uses Gaussian blur.
Performance note: Gaussian blur is computationally intensive — O(n × r) where n is pixel count and r is radius. A 3000×2000px image with radius 20 involves ~240 million operations. Processing runs in a Web Worker to prevent freezing the page.
Selective blur: for blurring only part of an image (a face, a license plate, a document number), use the Privacy Blur tool which lets you draw a selection area before blurring.
Privacy: no pixel data leaves your browser. This is especially relevant for users blurring screenshots with personal information.
Frequently Asked Questions
- A radius of 15–20 is sufficient to obscure facial features in most photos. For absolute certainty (security-critical redaction), use a radius of 30–50. For a simple privacy softening, radius 10 is usually enough to prevent casual identification while preserving context.
- Yes, significantly. Gaussian blur removes high-frequency detail (edges, textures, noise) which are the components that are hardest to compress. A blurred JPG is typically 30–60% smaller than the original at the same quality setting.
- Gaussian blur spreads each pixel's value equally in all directions (circular blur kernel), simulating defocus. Motion blur spreads only in one direction (linear kernel), simulating camera or subject movement. This tool applies Gaussian blur. Motion blur is available in dedicated image editing software.
- Use the Privacy Blur tool to draw a selection over the specific area you want to blur — it applies the same Gaussian blur only within your selected region, leaving the rest of the image sharp.