Unlike lossy formats (JPG), PNG uses DEFLATE compression (the same as zlib). But not all PNGs are created equal—many tools generate bloated PNGs with unnecessary chunks, inefficient filters, or unused color palettes.
[Design Software Export] │ ▼ [Run through Lossy Quantizer (e.g., TinyPNG)] ──► Reduces color palette bloating │ ▼ [Run through Lossless Stripper (e.g., ImageOptim)] ──► Wipes remaining metadata │ ▼ [Optimized Production Asset] png to png better
When processing a PNG to make it better, you have two primary choices: Optimization Type How It Works Best Used For Unlike lossy formats (JPG), PNG uses DEFLATE compression
npm install -g imagemin-cli imagemin input.png --out-dir=output --plugin=pngquant However, the standard encoders in most software make
The PNG format uses the compression algorithm (the same one used in ZIP files) to compress pixel data. However, the standard encoders in most software make a trade-off for speed rather than the ultimate file size. A lossless optimizer simply re-runs the DEFLATE algorithm with more aggressive settings. It can try different compression levels, filter strategies, and even rearrange the data to find the smallest possible encoding without losing any quality.
The phrase usually refers to optimizing an existing PNG file to make it "better"—meaning a smaller file size without losing any visual quality. Since PNG is a lossless format, you can re-compress it to strip out unnecessary metadata and more efficiently encode the pixel data. Why optimize a PNG to another PNG?