JPG vs PNG: which one should you use?

By Andrew Butson · July 2, 2026 · 6 min read

JPEG and PNG have coexisted since the mid-1990s, and between them they cover almost every image on the internet, which is exactly why people confuse them. JPEG (1992) was built for continuous-tone photography: it uses lossy compression that throws away detail the eye barely notices, shrinking a photo to a fraction of its raw size. PNG (1996) was built to solve a different problem entirely — a licence-free, lossless alternative to GIF that could reproduce a screenshot or logo pixel-for-pixel and add proper transparency.

The result is two formats optimised for opposite kinds of image. Ask which is “better” and there’s no single answer; ask which suits a photograph versus a diagram, and the answer is immediate.

Lossy vs lossless: the difference in practice

JPEG’s compression is built around a fact about vision: people notice broad shapes and brightness far more than fine colour detail. The encoder splits the image into blocks, transforms them, and discards the highest-frequency information first: the detail a viewer is least likely to miss. Push the quality slider down and it discards more; push it up and less, but a JPEG always throws something away.

PNG never does. Its DEFLATE compression finds repeated patterns in the pixel data and encodes them more compactly, the same idea as a ZIP file, but the process is fully reversible. Decompress a PNG and you get back the exact pixels that went in, forever.

Why the same rule doesn’t apply to every image

That difference in mechanics is why the photo-vs-graphic rule of thumb holds up so well. A photograph is made of millions of pixels that all vary subtly (clouds, skin tones, foliage), and JPEG’s lossy transform handles that gracefully, throwing away variation nobody would spot. Try to losslessly compress that same photo and DEFLATE barely shrinks it, which is why a PNG photo is often several times the size of the JPEG.

A screenshot or a logo is the opposite case: large flat areas of identical colour and hard, high-contrast edges around text and shapes. DEFLATE compresses that superbly, so a PNG stays small. JPEG, on the other hand, smudges those same hard edges. The blocky “ringing” you sometimes see around text in a low-quality JPEG is the DCT transform struggling with exactly the kind of content it wasn’t designed for.

Transparency is the other half of the decision

JPEG has never supported transparency in any form. Every JPEG is a solid rectangle. If your source image has a see-through background and you convert it to JPEG, Morphr fills the transparent area with a solid colour rather than leaving a gap, because there’s nowhere else for it to go.

PNG’s alpha channel is the reason it became the default for logos, icons and any image that needs to sit cleanly on top of something else. It’s not a feature bolted on afterwards: full, graduated transparency was part of PNG’s design from 1996, specifically to go further than GIF’s crude one-bit transparency.

What actually happens when you convert

Converting a photo from PNG to JPEG is usually a straightforward win: Morphr re-encodes it with a quality slider, and the file typically shrinks to a fraction of its size with no visible difference. It also strips the EXIF and GPS data the original carried, so the JPEG leaves nothing about the camera or location behind.

Converting the other way, JPEG to PNG, is more of a formality than an upgrade. PNG can’t recover detail the JPEG’s encoder already discarded. It only stores, exactly, whatever pixels it’s handed. The result is a PNG that looks identical to the JPEG but takes up considerably more space. That’s worth doing when a tool insists on PNG input, but it won’t turn a JPEG back into an untouched original.

A history that explains the rivalry

JPEG dates to 1992, a joint effort by the ISO and ITU-T committees that gave it its name, and it arrived at exactly the moment digital photography needed a format that could make camera and scanner files small enough to store and share. PNG followed in 1996 for an unrelated reason entirely: GIF’s LZW compression was covered by a Unisys patent, and developers wanted a free, technically better lossless format to replace it. PNG was never meant to compete with JPEG for photographs — it was built to out-do GIF for everything else.

That’s why the two formats have coexisted peacefully for thirty years rather than one displacing the other: they were solving different problems from the start, and the content each is good at hasn’t changed. Photographs still benefit from throwing away imperceptible detail; screenshots, logos and diagrams still benefit from never losing a pixel.

Frequently asked questions

Which is better quality, JPG or PNG?

PNG is technically higher fidelity because it’s lossless, but that rarely matters for photographs, since JPEG’s loss is invisible at a sensible quality setting. For graphics and text, PNG’s lossless compression avoids the smudging artefacts JPEG introduces around sharp edges.

Does converting a JPG to PNG improve its quality?

No. PNG can only store, exactly, the pixels it’s given. It can’t restore detail a JPEG’s compression already discarded. The PNG will look identical to the JPEG, just larger.

Can a JPG have a transparent background?

No, JPEG has never supported transparency. If you need a see-through background, save the image as PNG or WebP instead.

Why is my PNG file so much bigger than the same picture as a JPG?

PNG is lossless, and photographs, with millions of subtly different pixels, don’t compress well without throwing some detail away, which is exactly what JPEG does and PNG doesn’t.