How JPG Compression Works (Lossy Compression Explained)

"Lossy" — the word that explains everything

JPG uses lossy compression: to save space it permanently throws some information away. That sounds alarming, but the genius of JPG is what it discards — detail your eyes are biologically poor at noticing. Done well, you lose bytes you'll never miss while keeping every detail you actually see.

Lossy vs lossless: a lossless format (like PNG) keeps every bit but stays large. JPG trades a sliver of unseen detail for an enormous size reduction — the right deal for photographs.

The four-step journey of a JPG

1. Colour conversion

The image is converted from RGB to a brightness-plus-colour model (YCbCr). This matters because human eyes are much more sensitive to brightness than to colour. Separating the two lets the encoder be stingier with colour without you noticing.

2. The Discrete Cosine Transform (DCT)

The image is split into 8×8 pixel blocks, and each block is transformed from "pixel values" into "frequencies" — broad, smooth gradients versus fine, rapid detail. Nothing is lost yet; the data is just described in a more useful way.

3. Quantisation — the lossy step

This is where compression happens. High-frequency detail (fine texture and sharp transitions) is rounded off aggressively, because the eye barely registers it. The quality slider simply controls how aggressive this rounding is: lower quality rounds away more, producing a smaller file and, eventually, visible artefacts.

4. Entropy coding

Finally the simplified data is packed efficiently using lossless techniques (run-length and Huffman coding). No further detail is lost here — it's just tidy storage.

Why over-compression looks "blocky"

Those 8×8 blocks are the culprit. Push quality too low and the rounding becomes coarse enough that the block edges show, especially in smooth areas like skies and around sharp text. That telltale grid of squares is JPG compression pushed past its limit.

Generation loss: why re-saving hurts

Every time you open a JPG, edit it, and save again, the lossy steps run again, throwing away a little more detail each round. After many saves the degradation stacks up — like photocopying a photocopy. The fix: always edit from the highest-quality original and export to JPG once, at the end.

Takeaway: JPG is brilliant for photos precisely because it discards the right things. Keep an original master, export once, and choose a quality around 75–85 for the best size-to-look balance.

Now put it into practice with our file-size reduction checklist, or revisit the compression overview.

FAQ

What does 'lossy' compression mean?

Lossy means some image information is permanently discarded to save space — specifically the fine detail your eye is least likely to notice. It cannot be recovered later.

Why does re-saving a JPG keep making it worse?

Every save re-applies lossy compression, throwing away a little more detail each time. This 'generation loss' is why you should edit from an original and export once.