Why convert a JPG to PNG at all
JPG and PNG solve different problems. JPG, standardised in 1992, was designed for photographs: it throws away detail the eye is unlikely to notice and gets files 5 to 20 times smaller than raw pixels. PNG, created in 1996 as a patent-free replacement for GIF, is lossless: every pixel you save is exactly the pixel you get back.
Converting does not make a photo look better, but there are solid practical reasons to do it:
- Editing in several rounds. Every time a JPG is opened, changed and saved again, it is recompressed and loses a little more quality. Working on a PNG copy stops that cycle, and you export to JPG once at the very end.
- Upload forms and software that insist on PNG. Some app stores, icon generators, print services, game engines and older CMS plugins accept only PNG.
- Screenshots and graphics that were saved as JPG by mistake. Text, diagrams and flat-colour artwork suffer from JPG artefacts. Converting will not remove artefacts already there, but further edits such as adding arrows or labels will stay crisp.
- Preparing a base for transparency. PNG supports an alpha channel. A JPG has none, so the converted file is fully opaque, but it is now in a format where you can erase the background in an image editor and keep the result.
How the conversion works here
Each file is decoded by your browser's own image decoder and drawn onto a canvas at its original pixel dimensions, then encoded as PNG. Where the browser supports it, this happens in a background worker so the page stays responsive while a batch is processing. Nothing is uploaded; the images exist only in your tab's memory until you download them.
There are no settings to choose because PNG has no quality slider: the output is lossless by definition. Files are processed one after another, each row shows the size before and after, and you can download files individually or all at once as a ZIP.
WebP is accepted as input too. Unlike JPG, WebP can carry transparency, and that alpha channel is preserved in the PNG.
What changes and what does not
The pixels you see are carried over exactly, but a few things around them do not survive a canvas re-encode:
- Metadata is removed. EXIF data such as camera model, exposure settings, date taken and GPS coordinates is not written into the PNG. That is useful if you plan to share the image publicly.
- Orientation is baked in. Phone photos often store rotation as an EXIF flag rather than rotating the pixels. Modern browsers apply that flag when decoding, so the PNG comes out upright and no longer depends on the flag.
- Colour profiles are not embedded. The PNG is written in standard sRGB. Photos taken in a wide-gamut colour space, such as Display P3 on recent phones, may look slightly less saturated on high-end screens.
- Dimensions stay the same. A 4000 by 3000 pixel JPG becomes a 4000 by 3000 pixel PNG.
Expect much larger files
PNG compression works by finding repeated patterns, which photographs have very few of. A 2 MB phone photo commonly becomes a 10 to 20 MB PNG. Simple graphics, logos and screenshots grow far less, sometimes only by 20 to 50 percent.
If the PNG is too heavy for where it is going, these steps help:
- Reduce the pixel dimensions with the image resizer. A 1600 px wide image has less than a fifth of the pixels of a 4000 px original, and PNG size drops roughly in proportion.
- If the destination accepts either format and the image is a photo, keep it as JPG instead. For final delivery a high-quality JPG is usually the better choice, and you can convert back with PNG to JPG after editing.
- For graphics with few colours, the PNG compressor can shrink the file considerably without visible change.
Limits
Each file can be up to 50 MB, and there is no fixed limit on how many files you add at once. Very large images, for example 50 megapixel camera files, need a lot of memory to decode, so phones may struggle with big batches. If a conversion fails, try fewer files at a time or a desktop browser.