Chroma subsampling

How JPEG and other formats shrink files by storing less color detail than brightness detail.

Definition

Chroma subsampling reduces file size by storing color (chroma) information at lower resolution than brightness (luma) information. The human eye is more sensitive to brightness detail than color detail, so discarding color data is rarely noticed. Notation like 4:4:4, 4:2:2 and 4:2:0 describes how many chroma samples are kept per block of pixels.

In practice

4:2:0 is the default for most web JPEGs and halves the color data — usually invisible on photographs. 4:4:4 keeps full color resolution and matters for images with saturated hard edges: graphics, text, screenshots and product shots with strong color boundaries. If you see color bleeding on a compressed screenshot, that's chroma subsampling.

Compress your images now — free and private

Drop your files and ImageAlchemy compresses them in your browser. No uploads, no accounts, no limits. Batch process hundreds at once and download a ZIP.

Open the compressor
100% private No signup

Frequently asked questions

It's a chroma subsampling mode: for every 4-pixel-by-2-pixel block, 2 chroma samples are kept in the horizontal direction and 0 additional vertical resolution — effectively halving color resolution in both directions. It's the default for most web JPEGs and is invisible on most photographs.

4:4:4 keeps full color resolution; 4:2:0 halves it horizontally and vertically. 4:2:0 produces smaller files and is fine for photographs. Use 4:4:4 for graphics, text and screenshots where color bleeding along hard edges would be visible.

It reduces color detail, which the human eye notices far less than brightness detail. On photographs at normal viewing size, 4:2:0 is effectively invisible. On saturated hard edges — text, screenshots, flat graphics — it can cause visible color bleeding, which is why those images are better served as PNG or WebP.