Why Your Photo Looks Worse After Compressing (And How to Avoid It)
Compression damage isn't random — it shows up in specific, predictable places once you know what to look for, and almost all of it is avoidable.
Blocky squares near sharp edges
JPEG compresses images in fixed 8×8 pixel blocks. At low quality settings, those block boundaries become visible as a faint grid, especially around high-contrast edges like text on a photo or the outline of a face against a plain background. This is the single most recognizable sign of over-compression.
Color smearing and banding
JPEG stores color information at lower resolution than brightness information, because human eyes are more sensitive to brightness detail than color detail — a legitimate optimization most of the time. Pushed too far, it causes color to "bleed" across edges, and smooth gradients (skies, shadows) can break into visible bands instead of a smooth transition.
The real cause: repeated compression, not just low quality
A single compression pass at a reasonable quality setting is usually fine — the loss is real but rarely visible. The damage compounds when an image gets compressed multiple times: downloaded, edited, re-saved, uploaded to a platform that recompresses it again, downloaded again. Each pass discards more, and the artifacts from each pass stack on top of the last one's.
How to avoid it
- Compress from the original file, not from a copy that's already been compressed once
- Target a specific file size or a quality setting around 75-85% rather than the lowest setting available — the size difference between "low quality" and "very low quality" is often small, but the visible damage is not
- For images with text or sharp lines, prefer PNG or WebP over JPEG entirely — that class of image is where JPEG artifacts are most visible
- If a platform (social media, a CMS) will recompress your upload anyway, don't pre-compress aggressively yourself — you're just adding a second lossy pass on top of theirs