Most image problems on the web come down to one decision made too quickly: which format to save in. A photograph saved as a PNG can be five times larger than it needs to be. A logo saved as a JPEG can grow fuzzy edges around its text. A WebP file can be beautifully small and still fail in one older tool in your workflow. None of these formats is bad. Each one is built for a different kind of picture, and the trick is matching the picture to the format before you upload anything.
Image Optimizer lets you convert between JPEG, PNG and WebP in a batch, so you can test the same picture in all three and compare the results side by side. That comparison is worth doing once, because it replaces guesswork with numbers you can see. This post walks through the same fictional photograph exported three ways, explains what each format does well and badly, and ends with a simple rule you can reuse for every image you publish.
What each format actually does
JPEG is designed for photographs. It compresses by discarding fine detail the eye is least likely to notice, especially in smooth colour gradients like skies and skin tones. That is why a detailed photo can shrink dramatically as a JPEG and still look fine at a glance. The trade-off is that JPEG compression is lossy and it struggles with sharp edges: text, line art and flat graphics come out with visible smudging around the edges, and every re-save loses a little more detail.
PNG is designed for exactness. It compresses without throwing any pixel data away, so a logo, diagram or screenshot comes back bit-for-bit identical. It also supports transparency, which is why it is the natural choice for a graphic that must sit on different background colours. The cost is size: for a photograph, PNG has no clever way to discard the detail nobody will miss, so the file stays large.
WebP was designed to cover both jobs. It offers lossy compression for photographs that is typically smaller than JPEG at similar quality, lossless compression for graphics, and transparency in both modes. The catch is support: modern browsers handle WebP well, but some older editors, email clients and content systems still do not, so a WebP file can occasionally trip up a step in your publishing chain.
The same photo, three ways
Imagine a fictional food blogger, Priya, who runs a small site at example.org and photographs a bowl of soup for a recipe post. Her camera produces a large JPEG, roughly 2400 pixels wide. She exports it three ways at a size suited to her blog layout, about 1200 pixels wide.
As a JPEG at a quality setting of 80, the file comes to around 210 KB. The soup looks warm and detailed, the steam is soft, and nothing in the image draws attention to the compression. As a PNG, the same picture comes to roughly 1.4 MB: identical to the eye, but nearly seven times the size, with no benefit the reader can see. As a WebP at a comparable quality setting, it comes to around 150 KB, visibly the same as the JPEG and noticeably smaller.
These are illustrative figures for one picture, not promises about yours — a busy, high-contrast scene compresses less well than a soft one. But the pattern holds across most photographs: PNG is the wrong tool for photos, and WebP usually beats JPEG by a useful margin. For Priya, the JPEG is the safe choice and the WebP is the efficient one. Which she picks depends on where the image will be used, which brings us to the decision rule.
A simple rule for choosing
Start by asking what the picture is. If it is a photograph or a realistic image with many colours and soft gradients, choose JPEG or WebP. If it is a logo, icon, diagram or screenshot with sharp edges, flat colours or text, choose PNG or lossless WebP. Then ask whether transparency matters. If the image must show the background through it, JPEG is out entirely, because JPEG has no transparency at all.
Next, ask where the image will live. For a modern website, WebP is usually the best default for both photographs and graphics. For an image that will be emailed, printed, or passed through an older system, JPEG for photos and PNG for graphics remain the safer pair, because every tool on earth reads them. When in doubt, keep the original as well as the export. Storage is cheap; re-shooting a photo or re-drawing a graphic is not.
Finally, check the result at the size it will actually be shown. Open the exported image at its display size, not zoomed in to 200 percent, and look at the edges and the smooth areas. Compression artefacts that are invisible at display size do not matter, however obvious they look under a magnifying glass. Judge the file the way your reader will see it.
Worked example: one blog post, three images
Priya's recipe post needs three images: a hero photograph of the soup, a small step-by-step photo of chopped vegetables, and her site logo in the corner of a printable recipe card. She processes all three in one batch.
The hero photograph becomes a WebP at 1200 pixels wide, because her blog platform supports WebP and the file is the smallest of the three options. The step photo becomes a JPEG at 900 pixels wide, because it will also be attached to an email newsletter, and she wants a format every inbox can display. The logo stays a PNG, because it has transparent edges and sharp lettering, and a JPEG version would blur the text and fill the transparent corners with white.
Before uploading, she checks the before-and-after sizes the tool reports. The two photographs together drop from several megabytes to a few hundred kilobytes, while the logo stays exactly as drawn. She uploads the WebP hero, the JPEG step photo and the PNG logo, and each one does the job its format was built for. If her newsletter system ever adds WebP support, she can switch the step photo over without re-shooting anything, because the original is still on her computer.
For a deeper explanation of how formats affect loading, see how to optimize images for the web.
Honest limitations
Format choice cannot rescue a bad source image. A blurry, badly lit photograph will still be blurry and badly lit as a WebP, just smaller. Compression settings are also a compromise: pushing quality too low saves bytes but introduces blocky patches in skies and gradients, and no format avoids that entirely. Browser-based conversion works on the files you give it, so very large originals can be slow in a tab with little memory, and converting to WebP does not help if the next system in your chain cannot read WebP. Check support at each step before committing to a single format.
FAQs
Should I convert all my old JPEGs to WebP?
Not necessarily. Converting a JPEG to WebP does not recover detail the JPEG already discarded, so the saving is smaller than converting from an original. For new images, starting from the original and exporting WebP makes sense; for an archive of JPEGs, the gain may not justify reprocessing everything.
Why does my logo look fuzzy as a JPEG?
JPEG compression blurs sharp edges, and a logo is almost entirely sharp edges and flat colour. Use PNG or lossless WebP for logos, icons and text graphics, and the edges will stay clean.
Does WebP support transparency like PNG?
Yes. WebP supports transparency in both its lossy and lossless modes, so it can replace PNG for many graphics as well as photographs, provided the tools in your workflow support it.
What size should a blog image be?
Match the widest size it will be displayed at, and no larger. If your layout shows images at 1200 pixels wide, uploading a 4000-pixel original wastes bandwidth without improving what readers see. Resize first, then choose the format.
If you later need that photograph inside a document, JPG to PDF can combine images into a single PDF, and if you need a still from a document, PDF to JPG renders pages as images.
human review pending client designation