Pixelate a photo or selected areas
This image pixelator turns a photo into blocks of averaged color. Use the whole-image mode for a pixel-art effect, or select rectangles over faces, license plates and text while leaving the rest of the picture unchanged. The controls update the preview as you work, so you can compare block sizes before saving a PNG. You do not need an account or an image-editing application.
The image is decoded and processed in your browser. Choosing a file does not upload its pixels to an image-processing service. The tool accepts images your browser can decode, rather than promising support for every camera format. If a file cannot open, export a standard PNG or JPEG from your photo application and try that copy. Keep the original separately so you can return to it later.
How block averaging works
The image is divided into square blocks. Within each block, the red, green, blue and alpha channels are averaged separately, rounded to whole channel values, and written back across the block. In plain terms, channel average = sum of channel values / number of pixels in that block. This produces flat squares rather than the smooth transitions of a conventional blur filter.
Blocks at the right and bottom edges can contain fewer pixels; their averages use only the pixels actually present. In selected-area mode, the block grid starts at the selected rectangle's corner. Multiple rectangles are processed in selection order. Overlapping selections can therefore change an already processed part of the picture again. Undo the last area if that effect is unwanted, or clear all areas and draw separate rectangles.
A worked pixelation example
The example below was computed with the same pixelation library used by the preview. Consider a small opaque blue block whose red-channel values are 0, 1, 4 and 5. Green is 0, blue is 255 and alpha is 255 throughout. Averaging the block gives a red value of 2.5, which the library rounds upward to 3. Every output pixel in that block becomes RGBA (3, 0, 255, 255).
For a practical size example, open a 200 × 100 image and leave the block size at 12 pixels. The standard download remains 200 × 100 pixels, with the block effect painted into it. The Pixel-art size download is 17 × 9 pixels. That size follows grid width = ceiling(image width / block size), with the same rule for height. The final column and row represent partial edge blocks rather than stretching the source to fit an exact multiple.
Choose the right download
Download PNG preserves the working canvas dimensions and is useful when the processed picture must occupy the same space in a document or slide. Pixel-art size is available in whole-image mode and creates a genuinely smaller grid by taking a representative pixel from each processed block. It is useful for sprites or further pixel editing, but it does not include a sprite-sheet editor or animation timeline.
Very large source images are scaled down before editing to keep preview changes responsive. Read the dimensions on the download buttons instead of assuming the original camera resolution is retained. The tool exports the current canvas as PNG; it does not offer adjustable JPEG quality or a choice of output codecs. A small pixel grid may look soft in another application unless that application uses nearest-neighbor enlargement.
Original colors and retro palettes
Original colors keeps each averaged block's computed RGB color. A palette replaces that color with the nearest available palette color, using squared distance between RGB channel values. Game Boy offers four greens, PICO-8 offers sixteen colors, and the grayscale and black-and-white choices restrict the image further. These are color effects, not automatic line tracing or a guarantee of hand-drawn pixel-art quality.
Try the block size first, then the palette. A restrictive palette can merge nearby subjects that remain distinct in the original colors. Inspect faces, lettering and fine outlines after each change. If your goal is an equation-based drawing rather than a raster picture, the image-to-equations tool follows a different workflow. For practical selection ideas, read the pixelate an image guide.
Obscuring details needs a final review
Selected areas lets you drag boxes over details while keeping the rest of the image readable. Choose a box that covers the entire detail, including the edges of letters or identifying features. The control offers an undo action for the most recent rectangle and a separate clear action for all rectangles. Switching back to whole-image mode changes which parts receive the effect; check the active mode before downloading.
Pixelation is not guaranteed anonymization. Large blocks can still leave context, distinctive clothing, silhouettes or readable fragments, and small blocks may preserve enough detail to identify the subject. Review the downloaded image at its intended sharing size. For sensitive information that must be removed, use an editor that can replace it with an opaque solid area and inspect the final flattened file. Do not treat any particular block size as a privacy certification.
Common mistakes and next steps
A common mistake is judging the result only from a small preview. Open the download and inspect the selected details again. Another is expecting the small-grid export to retain fine lettering; discarding detail is the purpose of the effect. Transparent edges can also look different against different backgrounds because alpha is averaged along with the color channels.
For a different artistic result, keep an original copy and try the drawing-to-equations editor after sketching a simpler outline. Use the pixelator when a block-based bitmap is the output you need. It does not detect faces automatically, select objects by meaning, or restore hidden detail, so each rectangle and every final publishing decision stays under your control.