Glossary Design & Media

Image compression

Image compression makes an image file smaller, either exactly (lossless) or by discarding detail (lossy).

Last reviewed:

AdSense Placeholder
Slot: header_reference_page

In more detail

Lossless compression (PNG, GIF, lossless WebP) finds repeated patterns and rebuilds every pixel exactly, so it works best on screenshots, logos and flat colour. Lossy compression (JPEG, lossy WebP, AVIF) throws away detail the eye is least likely to notice, such as fine colour variation, and achieves much smaller files for photos. Lower quality settings shrink lossy files further but add visible artefacts: blocks, blur and halos around edges.

Why it matters

Large images slow down web pages and hit email and upload limits. Each time a lossy file is edited and saved again, it is recompressed and loses a little more quality (generation loss), so keep an original and export once. Resizing usually saves more than lowering quality: resize to the size you will display before compressing.

Example

A 4000 × 3000 photo has 12 million pixels; at 3 bytes per pixel it is 36,000,000 bytes uncompressed, while a good-quality JPEG of it is typically a few megabytes. Resizing it to 2000 × 1500 for a web page cuts the raw data to 9,000,000 bytes, a quarter, before any compression is applied.

Common confusion

Compressing makes the file smaller at the same pixel dimensions; resizing changes the Resolution and PPI itself.

AdSense Placeholder
Slot: footer_leaderboard