Compression Quality

The quality slider is where most file-size savings come from. This page explains what it changes and how to pick a good value.


What the quality slider does

For JPG and WebP output, the quality slider controls the tradeoff between visual fidelity and file size.

  • Higher quality means fewer visible artifacts and larger files
  • Lower quality means smaller files and more visible compression damage

The slider ranges from 1 to 100.


When the slider applies

Output mode Does quality apply?
JPG Yes
WebP Yes
PNG No - PNG export is lossless
Same as Input Only for JPEG and WebP source files

If you switch output to PNG, the app disables the quality slider because PNG compression is lossless.


Quality Typical use
90-100 Minimal quality loss, larger files
80-89 Best default balance for most photos
65-79 Stronger compression when file size matters
1-64 Aggressive compression, usually only for throwaway previews or web thumbnails

The default value in BatchCompress Image is 80.


What artifacts to look for

When quality is too low, you may notice:

  • Blockiness in detailed textures
  • Smearing in foliage, hair, or fabric
  • Haloing around sharp edges
  • Text or UI elements becoming fuzzy
  • Banding in gradients or skies

The live preview is the fastest way to catch these issues before you process the whole batch.


Photo vs graphic content

Photos

Photographs usually compress well as JPG or WebP. Start around 80 and only move lower if the preview still looks clean.

Graphics, logos, screenshots, and text

These often look worse at low JPG quality because hard edges show artifacts quickly. Prefer PNG for these images when preserving crisp edges matters more than file size.


Why preview matters

Different images react differently to the same quality value. A portrait with a soft background may still look excellent at 70, while a screenshot full of fine text may look poor even at 90.

BatchCompress Image previews one file from the folder using your current settings so you can tune the tradeoff before running the full batch.