Compression fundamentals

Lossy vs Lossless Image Compression: A Practical Guide

Learn how lossy and lossless compression differ and how to choose for photos, graphics, editing and the web.

Reviewed 13 September 20268 min readEditorial policy

What lossless means

Lossless encoding reconstructs the represented pixel data without compression loss. PNG is common. This does not make every PNG an untouched original; earlier resizing or conversion may already have changed it.

What lossy means

Lossy encoders approximate less noticeable information. Sensible settings can look excellent; aggressive settings cause blocking, ringing, smeared texture and colour banding.

Choose by workflow stage

StageApproach
MasterKeep the original or a lossless/editable master
EditingAvoid repeated lossy saves
Web deliveryVisually test a lossy photo copy
Logo/screenshotPrefer lossless output

How to test quality

Inspect at intended size and difficult areas at 100%: hair, foliage, small type, gradients and hard edges. Use the lowest setting that meets the real requirement; quality numbers differ between encoders.

Practical steps

  1. Keep a high-quality master.
  2. Choose compression for the workflow stage.
  3. Test several settings on representative content.
  4. Inspect difficult areas before publishing.

Common mistakes

  • Calling maximum-quality JPG lossless.
  • Using lossless output when delivery size is the primary constraint.
  • Recompressing a lossy file repeatedly.

Frequently asked questions

Is lossless always higher quality?

It preserves represented pixels, but a damaged source remains limited by that source.

Can lossy compression be reversed?

No. A later PNG save cannot reconstruct discarded detail.

Authoritative references