Blur & Sharpen
Seven precision effects — Box, Gaussian, Motion and Zoom blur, plus Sharpen, professional Unsharp Mask, and Edge Enhance — with a live before/after preview that updates instantly as you drag any slider. Batch process a whole folder and download individually or as one ZIP — all inside your browser, nothing uploaded.
How to Blur or Sharpen an Image
Features
Two Opposite Effects, One Tool
A blur and sharpen tool changes how much local detail is visible in an image — blurring averages neighboring pixels together to soften detail and noise, while sharpening amplifies the contrast between neighboring pixels to make edges look crisper. Real photo-editing software offers several distinct blur and sharpen algorithms rather than one generic slider, because each is genuinely suited to a different problem: a portrait background needs different blur math than a fast-zoom motion effect, and a quick touch-up needs different sharpening math than a professional print-ready pass. This tool implements seven of them, each with its own correct underlying algorithm.
If you want to change exposure or color rather than detail, use Image Brightness for precise brightness/contrast/saturation/hue control, or Image Filters for one-click artistic looks. Run the result through the Image Compressor afterward if file size matters.
How Each Effect Works Technically
Box Blur and Gaussian Blur are both separable convolutions — processed as one horizontal pass and one vertical pass, which is far faster than a full 2D kernel — but with different weights. Box Blur weights every pixel in the radius equally; Gaussian Blur uses a real Gaussian (bell-curve) function to weight pixels near the center more heavily, which is what produces its smoother, more photographic look. Motion Blur samples pixels along a line through each pixel at the angle you set, and Zoom Blur samples along a line from each pixel toward the image center — both average those samples into the final pixel value. Sharpen and Edge Enhance use small convolution kernels that subtract a weighted sum of neighboring pixels from an amplified center pixel, which is what makes edges look more defined. Unsharp Mask is the most involved: it blurs a copy of the image with Radius, subtracts that blur from the original to isolate edge detail, discards differences smaller than Threshold (protecting flat areas from noise), and adds the remainder back scaled by Amount.
Every effect runs through the exact same processing function for both the live preview (computed at a small capped size for speed) and the final export (computed once at full resolution) — so the preview you see while adjusting sliders is always an accurate representation of what you'll download, just rendered smaller.
Who Actually Uses Blur and Sharpen
Anyone sharing a photo with sensitive information visible — faces in a crowd, a license plate, an address on an envelope, account numbers in a screenshot — where Box or Gaussian Blur at a strong radius obscures the detail while keeping the overall image legible.
Photographers with slightly soft images from camera shake or a missed focus point, where Unsharp Mask recovers a meaningful amount of apparent detail that's technically present in the pixel data but too soft to read clearly.
Designers and marketers creating a sense of speed or motion in product photography or ad creative using Motion or Zoom Blur, a technique common in automotive, sports, and electronics advertising.
Anyone preparing a batch of scanned documents or graphics where a light, consistent sharpening pass across every image improves legibility without manually adjusting each file.
From Hiding a Background to Rescuing a Soft Photo
Blurring faces before posting a group photo: Upload the photo, select Gaussian Blur, set Radius to 25-35px and Intensity to 100%, and the faces become unreadable while the overall scene composition stays recognizable.
Rescuing a slightly out-of-focus product shot: Select Unsharp Mask, set Amount around 150%, Radius around 3-5px, and Threshold around 3-5 to sharpen genuine edges without amplifying sensor noise in smooth surfaces.
Giving a static car photo a sense of speed: Select Motion Blur, set the Angle to match the car's direction of travel, Length around 20-30px, and Intensity around 50-60% for a believable motion effect rather than an obviously artificial one.
Every Adjustment Happens Right on Your Device
Because every effect is computed locally using the Canvas API, your images are never transmitted to a server — there's no upload step to intercept and no server-side copy left behind. This matters for photos you're blurring specifically to protect privacy, since the unblurred original never leaves your device in the first place. It's also faster: the live preview updates instantly as you adjust sliders, with no round trip to wait on.
Getting a Natural-Looking Effect
- Use Gaussian, not Box, for anything meant to look natural — Box Blur\'s equal-weighting can look slightly blocky at high radius; Gaussian\'s bell-curve weighting almost always looks smoother.
- Start Unsharp Mask with a low Amount and small Radius — 100-150% Amount and 2-4px Radius is a reasonable starting point for most photos; push higher only if the result still looks soft.
- Raise Threshold if sharpening makes noise or skin texture too obvious — a Threshold of 5-10 protects smooth areas while still sharpening real edges.
- For privacy blurring, err on the side of too strong — a face that\'s "mostly" blurred can sometimes still be recognizable; push Radius higher than feels necessary.
- Match Motion Blur\'s angle to the subject\'s implied direction — an arbitrary angle reads as an obvious filter rather than genuine motion.
- Use Intensity to soften any effect that looks too strong — rather than re-adjusting every individual slider, dial Intensity down to blend back toward the original.
Why the Effect Might Look Too Strong or Too Weak
My blur looks the same at "Box" and "Gaussian". At small radius values (1-3px) the visual difference between the two is genuinely subtle — the distinction becomes much clearer at radius 15+ where Gaussian\'s smooth falloff versus Box\'s uniform averaging is easy to see side by side.
Unsharp Mask is amplifying noise in the sky or skin. Raise the Threshold slider — it\'s specifically designed to ignore small pixel-to-pixel differences (which are usually noise, not real edges) and only sharpen areas with genuine local contrast.
Motion Blur doesn\'t look realistic. Real motion blur is rarely uniform across an entire frame — this tool applies it globally, so very strong settings can look like an obvious filter rather than camera motion. Lower the Intensity to 40-60% for a more believable, subtle result.
The result looks over-processed. Reduce Intensity rather than the effect\'s own strength slider — it blends smoothly back toward the original at any percentage, which is usually a faster way to dial in the right amount than re-tuning Radius, Amount, or Strength directly.
What Blur and Sharpen Can't Fix
Don't rely on blur alone for irreversible privacy protection in a high-stakes context — a strong enough blur applied here is very difficult to reverse, but for extremely sensitive material, cropping or removing the region entirely is safer than any blur strength.
Don't expect sharpening to recover detail that was never captured — Unsharp Mask and Sharpen amplify existing edge contrast, they don't invent detail that isn't present in the original pixel data. A severely out-of-focus or low-resolution source will not become sharp no matter how the sliders are set.
Don't use this for animated GIFs you need to keep animated — Canvas processing only reads the first frame, so an animated GIF becomes a single still image after processing.