Posterize an Image Online

It reduces the number of distinct brightness values each colour channel can take. Instead of 256 possible reds, you might allow only four, so every red in the image snaps to the nearest of those four. The result is flat areas of solid colour with visible boundaries between them, which is the look of screen printing and vintage poster art.

Two gives a hard, graphic two-tone per channel. Three or four gives the classic poster look and is where most people land. Above about eight the effect becomes subtle, and by twelve it mostly reads as slight banding rather than a deliberate style. The right number depends entirely on whether you want it obvious.

Levels apply per channel, so the total palette is levels cubed. Four levels means up to 64 possible colours, three means 27, and two means 8. Switching on greyscale collapses that to just the number of levels, so two levels in greyscale is pure black and white.

Dithering scatters the rounding error across neighbouring pixels so your eye blends the few available levels back into an apparent gradient. Without it, a smooth sky posterized to four levels becomes four hard bands. With it, the same four levels are mixed in a fine pattern that reads as a continuous transition — using no extra colours at all.

Floyd-Steinberg pushes each pixel's error into the pixels not yet processed, which produces an organic, slightly random-looking texture that hides banding very well. Ordered dithering uses a fixed 4 by 4 matrix, giving a regular crosshatch pattern that is unmistakably retro and reproduces the look of early games and printed halftones.

No, and that is the point. We verified this on a smooth gradient: posterized to four levels, the output contained exactly four distinct values whether dithering was on or off. Dithering changes where those four values are placed, not what they are. The apparent extra shades come entirely from your eye averaging neighbouring pixels.

Because posterizing depends on where your tones sit. A flat, hazy photo has most pixels clustered in the middle of the range, so they all snap to the same one or two levels and the result is nearly blank. Raising contrast first spreads them across the full range so each level actually gets used. It runs before quantising for exactly that reason.

Yes. The quantisation is built so that pure black stays 0 and pure white stays 255 at every level count. With four levels the allowed values are 0, 85, 170 and 255 — evenly spaced and anchored at both ends, so your blacks stay black and your whites stay white.

No. Quantisation and dithering both run in your browser using the canvas API. Nothing is sent anywhere, and the tool keeps working if your connection drops after the page has loaded.

In PNG, usually a lot — large flat areas of identical colour compress extremely well, and a posterized image without dithering can be a fraction of the original size. With dithering the fine pattern works against compression, so the saving is smaller. In JPEG the effect is less predictable, because JPEG handles hard edges poorly.

PNG, in almost every case. Posterized images are made of flat areas with hard boundaries, which is exactly what PNG is designed for and exactly what JPEG handles worst — JPEG will add ringing around every boundary and mottle your flat areas.

Yes. Set greyscale on with two or three levels and no dithering, and you get clean, hard-edged tone separations that can be traced or cut. Raise contrast first so the separations land where you want them rather than where the photo happened to sit.

Yes. Add as many as you like and identical settings are applied to each. Download individually or take the whole batch as a single ZIP, built in your browser.

Any format your browser can decode: JPG, PNG, WebP, GIF (first frame), BMP, and AVIF or HEIC where supported. You can keep the original format or export as PNG, JPG or WebP.

Yes for PNG and WebP output. The alpha channel is untouched and only visible pixels are quantised. JPG has no alpha, so transparent areas are filled with the colour you choose first.

Images with strong shapes, clear subject separation and a good spread of tones. Portraits against a plain background, architecture and bold silhouettes all work. Busy scenes with lots of similar mid-tones tend to collapse into an unreadable mass, because once you have only three or four levels there is nothing left to separate the elements.

No. Reducing 256 levels to four discards information permanently — every value that snapped to 85 is now indistinguishable from every other. Keep your original file; the tool never modifies it, but the downloaded result cannot be reversed.

No to both. No sign-up, no email wall, no daily limit and no watermark. The output is your image with the levels reduced and nothing else added.