Free online image dithering tool using Floyd-Steinberg and other dithering algorithms. Convert images to pixel-style art with black/white, 4-color, or 16-color palettes. Adjustable dithering intensity. Pure frontend Canvas processing, no image upload. Perfect for pixel art creation and retro-style image processing.
📁 Drop image here or click to select
Enter Content: Type, paste, or upload the content you want to process. All processing happens locally in your browser — no data is uploaded to any server.
Configure Options: Adjust available options and settings to customize the output format.
Get Results: Copy the result or download the file once processing is complete. All processing is done locally — no data is uploaded to any server.
Image dithering is a technique that creates the illusion of more colors by mixing pixels of different colors. It exploits the human eye's tendency to blend nearby colors. The most common is Floyd-Steinberg dithering, which spreads quantization error to neighboring pixels, creating smooth transitions with limited colors.
Floyd-Steinberg (1976) is the classic error diffusion dither — it spreads 7/16 of the error to adjacent pixels. Atkinson dithering, invented by Apple engineer Bill Atkinson, only diffuses 1/8 of the error, producing cleaner results with more noticeable pixelation. Simple threshold takes the nearest color without error diffusion for the hardest effect.
① Pixel art creation ② Retro game-style image processing ③ Low-bandwidth image optimization ④ E-ink display optimization ⑤ Print halftone processing ⑥ NFT art stylization ⑦ Data visualization texture differentiation
B&W (2 colors) creates the most extreme stylized effect. 4-level gray or 4-color palettes retain more detail with moderate stylization. 16-level gray balances detail and effect. Different palettes suit different needs: B&W for retro looks, 4-color for GIF-style, grayscale for elegant monochrome effects.
No. All image processing uses the Canvas API in your browser. Your images never leave your device. All pixel-level operations (quantization, error diffusion, resampling) are done locally for complete privacy.