Two Formats, Two Completely Different Jobs
You right-click an image, hit "Save As," and a dropdown asks: JPG or PNG? Most people pick whichever one is on top and move on. Then the photo they emailed is somehow 9 MB, or the logo they uploaded has a grimy gray box behind it, or the screenshot they sent looks like it was faxed through a sandstorm.
Here's the thing nobody tells you: JPG and PNG aren't two flavors of the same thing. They're built for opposite jobs, and each one is genuinely *bad* at the other one's job. Use the right one and your files are small and crisp. Use the wrong one and you either bloat your file 10x or smear your edges into mush.
This is the rule you'll wish you'd known years ago, plus exactly what to do when you've already saved the wrong one.
The One-Sentence Version
If you only remember one thing: JPG is for photographs. PNG is for graphics with sharp edges, text, or transparency.
Photos — anything with smooth gradients, skin tones, sky, foliage — go in JPG. Logos, icons, screenshots of text, diagrams, and anything that needs a see-through background go in PNG. That single distinction solves about 95% of "which format?" decisions. The rest of this post is just *why*, so you can handle the edge cases yourself.
Why JPG Wins for Photos (and Loses for Logos)
JPG is lossy. When it compresses a photo, it throws away detail your eye won't miss — subtle color shifts, fine gradients, tiny variations between neighboring pixels. A 12-megapixel photo that would be 30 MB uncompressed becomes a 2 MB JPG that looks identical to you. That's a miracle of engineering, and it's exactly why every camera and phone shoots JPG by default.
But lossy compression has a tell: it hates hard edges. JPG works in 8×8 pixel blocks, and where a crisp black line meets a white background — the edge of a letter, the border of a logo, a clean line in a diagram — it can't decide which pixels are black and which are white, so it smears a halo of muddy gray pixels around the transition. This is called "ringing" or "JPG artifacts," and once you see it you can't unsee it.
So:
JPG also can't store transparency at all. If your image needs a see-through background, JPG is disqualified before the conversation even starts.
Why PNG Wins for Graphics (and Loses for Photos)
PNG is lossless. Every pixel is preserved exactly — no approximation, no smearing. That's why a logo, a screenshot of text, or a line-art diagram looks razor-sharp as a PNG: the hard edge between black and white stays a hard edge. PNG also supports a full transparency channel, so your logo can sit cleanly on any background color without a box around it.
The catch is the flip side of the same coin. Because PNG refuses to throw anything away, it's terrible at compressing photos. A photograph has millions of subtly different pixels and almost no repeated patterns, so PNG has nothing to optimize — it just stores all of them, faithfully, enormously. That gorgeous sunset photo that's 2 MB as a JPG can balloon to 15 MB as a PNG, and it won't look one bit better. You paid 7x the file size for detail your eye literally cannot perceive.
That's the trap that fills people's inboxes with attachments that won't send: saving photos as PNG because "lossless sounds higher quality." For photos, it isn't higher quality you can see — it's just bigger.
A Side-by-Side Cheat Sheet
| | JPG | PNG |
|---|---|---|
| Compression | Lossy (throws away detail) | Lossless (keeps everything) |
| Best for | Photos, complex images | Logos, text, screenshots, line art |
| Transparency | ❌ No | ✅ Yes |
| File size on photos | Small | Huge |
| File size on graphics | Small but fuzzy edges | Small and crisp |
| Repeated saves | Degrades each time | Never degrades |
That last row is worth a beat: every time you re-save a JPG, it re-compresses and loses a little more. Edit-save-edit-save a JPG ten times and it visibly rots. PNG never does — re-save it a thousand times and it's pixel-identical. For working files you'll keep editing, that matters.
What to Do When You Guessed Wrong
You don't have to live with the wrong format. Converting takes about ten seconds.
You saved a photo as PNG and now it won't email. Convert it to JPG. A 15 MB PNG photo drops to 1–2 MB with quality you can't tell apart from the original. Run it through PNG to JPG and the attachment sends instantly.
You saved a logo or screenshot as JPG and the edges look fuzzy. Converting back to JPG to PNG won't magically un-smear edges JPG already destroyed — once detail is gone, it's gone — but it stops *further* degradation, and if you can re-export the graphic from the original source as PNG, do that instead. The lesson sticks: graphics should never have been a JPG in the first place.
You want the best of both for the web. There's a third option that quietly beats both: WebP. It does lossy *and* lossless, supports transparency, and comes out 25–50% smaller than JPG at the same quality. For anything headed to a website, PNG to WebP and JPG to WebP are almost always the smallest, sharpest result. Every modern browser supports it.
Your file is still too big after converting. Drop it into the image compressor to squeeze it further without a visible quality hit, or use the image resizer if the pixel dimensions are larger than you actually need.
The Edge Cases Worth Knowing
Bottom Line
JPG and PNG aren't competitors — they're specialists. JPG throws away invisible detail to make photos tiny, and pays for it with fuzzy edges and no transparency. PNG keeps every pixel perfect for graphics and text, and pays for it with enormous photo files. Match the format to the content — photos to JPG, graphics and transparency to PNG — and your files come out small *and* sharp every time.
And when you've already saved the wrong one, you're never stuck. PNG to JPG rescues bloated photos, JPG to PNG stops a graphic from degrading further, and a quick hop to WebP gets you the smallest file of all for the web. Ten seconds, and the format finally fits the job.