Why convert from a lossy format to a lossless one
At first glance, converting JPG to PNG might look pointless — PNG’s core advantage is lossless compression, but the JPG source has already lost whatever detail JPG’s encoding discarded, and that loss is permanent. The value isn’t in recovering quality; it’s in not adding any more. Once an image is in PNG form, it can be opened, edited, cropped, layered, or resaved indefinitely without any further generational quality loss, whereas re-saving a JPG as a JPG (even at “the same quality”) re-runs lossy compression and compounds artifacts each time. Converting to PNG before a multi-step editing workflow — several rounds of cropping, color adjustment, or compositing in different tools — protects against that compounding loss during the editing process itself, even though the very first JPG encoding’s loss can’t be undone.
When a workflow requires PNG specifically
Design and editing software frequently standardizes on PNG (or a similar lossless format) internally, because intermediate editing steps benefit from not losing quality between saves, and because PNG’s alpha channel is often needed partway through a workflow — for example, isolating a subject from its background typically produces a PNG with transparency, which requires the working file to support that channel from the point transparency is introduced onward. Some content management systems and print workflows also specify PNG for anything containing text, line art, or sharp graphic elements, since PNG avoids the ringing artifacts JPG produces around hard edges.
What actually happens during conversion
The tool decodes the JPG into raw, uncompressed pixel data (this decode step is where the JPG’s original lossy compression’s effects become “locked in” — whatever detail is present in the decoded pixels is all that will ever be available going forward) and then re-encodes those exact pixels using PNG’s lossless compression scheme, which uses filtering and DEFLATE-based compression to shrink the file without discarding any of the decoded pixel values. The result is visually identical to the source JPG, just stored in a format that won’t degrade further and that supports transparency if you add it later.
Setting expectations on file size
Because PNG can’t discard detail the way JPG does, a PNG version of a photographic JPG is very often several times larger in file size for a visually equivalent image — this is normal and expected, not a bug or a sign of an oversized export. If file size matters and you don’t specifically need PNG’s losslessness or transparency support, keep the file as JPG (or convert to WebP, which offers a middle ground of better compression than JPG at similar quality). Use this converter specifically when a downstream tool or requirement calls for PNG, not as a general-purpose way to “upgrade” image quality — that isn’t something any format conversion can do.