JPG to PNG
Your images never leave this device
Converting a JPG to PNG produces a file that some upload forms and asset pipelines insist on. What it does not do is improve the image or give it a transparent background. It stores exactly the pixels the JPG decoded to, losslessly, in a file that is usually five to ten times larger.
How to convert JPG to PNG
The most common expectation of this conversion is the one it cannot meet. JPEG’s compression artefacts, the faint ripple around hard edges and the blockiness in flat areas, are ordinary pixel values by the time the browser has finished decoding. PNG has no way to tell them apart from real detail and preserves every one of them perfectly. Nor does the conversion create transparency: the background is opaque, exactly as it was, and removing a background is an editing job rather than a format change.
There is a second, less obvious reason the PNG comes out so large. PNG compresses by predicting each pixel from its neighbours and then deflating the residual, which works brilliantly when neighbouring pixels are similar. A decoded JPEG is not smooth in that way. The quantisation that made it small left a fine, structured noise across every eight-by-eight block, so the prediction is worse than it would have been on the camera original and the residual compresses worse. Most of the bytes in that PNG are being spent storing the JPEG’s own compression artefacts, faithfully.
There are real uses for it all the same. Some print services and some icon build steps accept PNG and nothing else. And if you are about to make several rounds of edits, converting to PNG once at the start stops every save adding another generation of JPEG loss on top of the last.
What people use it for
- Satisfying an upload form that accepts PNG and nothing else
- Feeding an icon or asset pipeline that expects PNG input
- Meeting a print service’s PNG-only file spec
- Getting a photo into a document or slide template that rejects JPG
- Freezing an image before editing so no further JPG loss stacks up
Questions
No. That is an editing operation, not a conversion. The PNG will have the same opaque background the JPG had.
No. It preserves what is there, artefacts included. Nothing can recover detail JPG already discarded.
Partly because PNG is lossless, and partly because the JPEG’s own block noise defeats the prediction PNG relies on. For a photograph, five to ten times the JPG is normal.
No. PNG is lossless, so there is nothing to trade. The slider appears only for the lossy output formats.
No further loss is introduced. The pixels the browser decoded are exactly the pixels written into the PNG.
When a system refuses anything but PNG, and when you want a stable starting point for several rounds of editing.
It is dropped, GPS coordinates included, because the file is rebuilt from a canvas rather than copied.
The image is drawn into an sRGB canvas and saved with no profile attached. An ordinary JPEG is already sRGB, so nothing changes; a wide-gamut one is converted down.
No. Everything runs in this page.