Question 1

How does PNG achieve lossless compression under the hood?

Accepted Answer

PNG uses the DEFLATE algorithm — a combination of LZ77 pattern matching and Huffman coding — preceded by a prediction filter that guesses each pixel's value from its neighbors. The filter stage doesn't compress anything itself; it rearranges the data so DEFLATE can find more redundant patterns. This two-step approach typically achieves 30–50% compression without altering a single pixel value.

Question 2

How does PNG compression work?

Accepted Answer

PNG is normally lossless, so traditional quality sliders don't apply. This tool uses color quantization — reducing the number of unique colors in the image (e.g., from millions to 256). Fewer colors means a dramatically smaller file while preserving transparency.

Question 3

What does the Colors slider do?

Accepted Answer

The slider controls how many unique colors the output PNG will contain, from 2 to 256. Fewer colors produce smaller files. For most images, 128–256 colors looks nearly identical to the original. For simple graphics, even 16–32 colors can look great.

Question 4

Does this preserve transparency?

Accepted Answer

Yes. Alpha transparency is fully preserved during color quantization. Your transparent PNGs will remain transparent.

Compress PNG

About this tool

Frequently Asked Questions