ضغط PNG دون فقدان الجودة

What "lossless PNG compression" actually does

Lossless compression does not touch a single pixel. It reduces the file by encoding the same image data more cleverly — choosing better PNG filter types per scanline, and applying a stronger DEFLATE pass over the result. Every pixel of the output is identical to the input, byte for byte after decoding.

Typical savings are 10–30% for photographic PNGs and rather less for images that are already well optimised, because a PNG that has been through a good encoder once has little redundancy left to find.

How to tell it worked

Compare file sizes, not pixels. If the output is smaller and the image looks identical at 100% zoom, it worked. Because the process is lossless, there is no quality slider to worry about and no such thing as "too aggressive" — the only variable is how long the encoder is allowed to search.

If lossless is not enough

When the file is still too large after lossless optimisation, the choice is between accepting some quality loss and reducing the image itself. In order of how little they hurt: reduce the pixel dimensions to what is actually displayed, reduce the colour depth if the image has few colours, then move to a lossy format such as WebP or JPEG.

Resizing is almost always the bigger win. A 3000-pixel-wide PNG displayed at 800 pixels is wasting more bytes than any encoder setting can recover.