The honest use cases
Three, and they are all technical:
- A tool, printer or platform that accepts PNG only
- You are about to edit and do not want another round of JPEG loss
- You need transparency in the output and will cut the background yourself
What it will not do
It will not sharpen anything, recover detail, or remove JPEG artefacts. Lossy compression is one-way: the information is gone from the file, and a lossless container simply stores the damaged version faithfully and at greater size.
If the goal is a better-looking image, the answer is the original file, not a conversion.
Expect the size
| Content | Typical change | | --- | --- | | Photograph | 4–10× larger | | Screenshot with text | Roughly the same, sometimes smaller | | Flat graphic, few colours | Often smaller |
PNG wins on flat colour and loses badly on photographic noise. If the file is destined for a web page, convert to WebP instead — lossless when you need it, and a fraction of the size either way.
Questions
Will converting to PNG improve the image?+
No. The JPEG already discarded detail, and PNG preserves what is left exactly — including the artefacts. The file gets larger and looks the same. Convert for a technical reason, not for quality.
Why is the PNG so much bigger?+
Because PNG is lossless and photographs have no repeating structure to exploit. A 400 KB JPEG photograph routinely becomes a 3 MB PNG. That is the format working correctly, not a fault.
Does the PNG gain transparency?+
The file supports an alpha channel, but the pixels are opaque — JPEG had no transparency to carry over. You get a canvas you can erase into with an editor, which is the usual reason to convert.
When is this actually the right move?+
When a tool demands PNG, when you are about to edit and want no further generation loss, or when you need to cut a background out afterwards. For putting an image on a web page, [WebP](/t/webp-converter) is smaller and lossless too.