The PNG format
Lossless, transparent, and the right answer for graphics — and the wrong one for photographs.
What is PNG?
PNG was created in the mid-1990s as a free replacement for GIF. It uses DEFLATE, the same lossless compression as ZIP, so the decoded image is bit-for-bit identical to what was encoded. Nothing is ever discarded.
Its second defining feature is the 8-bit alpha channel: every pixel carries its own opacity, which allows genuinely soft edges over any background. GIF could only do a single fully transparent colour, which is why cut-outs used to have jagged white fringes.
Lossless compression works by finding patterns. Flat colour and repeated structure compress extremely well; the near-random pixel variation of a photograph does not, which is why photos stored as PNG are so large.
Common uses
- Logos, icons and UI assets that need transparency
- Screenshots, where text must stay sharp
- Charts, diagrams and line art
- Intermediate files in an editing pipeline, to avoid generation loss
Advantages
- Completely lossless — re-saving never degrades the image
- Full 8-bit alpha transparency with smooth edges
- Sharp text and hard edges reproduced exactly
- Universal browser and application support
Limitations
- Very large files for photographic content — often 5–10× a comparable JPEG
- No animation (APNG exists but support is inconsistent)
- No quality dial, so the only ways to shrink a PNG are resizing or changing format
- Larger than lossless WebP for the same image
Browser support
Universal, including full alpha transparency. Safe to use anywhere without a fallback.