Convert images between JPG, PNG and WebP

Choosing between JPEG, PNG and WebP

JPEG is still the most compatible format for photographs. It uses lossy compression, so fine detail is discarded to keep files small, and it has no alpha channel. PNG is lossless and supports transparency, which makes it the right choice for screenshots, charts, logos and images with large flat areas or crisp text. WebP supports both lossy and lossless compression, handles transparency, and usually produces a smaller file than JPEG at a comparable visual quality. Current browsers can display it; older desktop software may not open it.

The file extension is not always a reliable clue. A file named .jpg can contain a PNG, and some downloads have no useful extension at all. This tool reads the magic bytes at the start of the file to identify the real source format. That matters for the “no unnecessary re-encode” rule and for the GIF warning.

Transparency and the JPEG background color

A transparent PNG dropped onto a JPEG canvas has to become something. By default the empty pixels become white, but a white box around a dark logo is often worse than keeping the PNG. Set the background to the color of the page, email or document where the image will be used. The fill is applied before the image is drawn, so semi-transparent edges blend into that color rather than producing a halo.

If the image will sit on several backgrounds, keep it as PNG or WebP. There is no single JPEG background that works everywhere. The background field has no effect on PNG or WebP output, where transparent pixels remain transparent.

Quality, repeated conversions, and file size

Quality is a lossy encoder setting, not a percentage of visual accuracy. At 0.92 most photographs are hard to distinguish from the source; 0.82 is a practical default for the web; 0.70 starts to show blocking in gradients and fine texture. PNG ignores the setting because it is lossless. Every JPEG or lossy WebP encode loses a little information permanently, so convert from the best original you have rather than from a file that has already been through several rounds.

Re-encoding an image that is already in the target format costs quality and may not even reduce the file size. When the detected source matches the target and quality is left on automatic, the tool returns the original bytes with a new filename. If you need a smaller WebP, choose an explicit quality value; if you need a specific byte limit, use the image compressor instead.

GIFs, metadata and what the output keeps

Animated GIFs are decoded as a still image by the browser's canvas API. Only the first frame is converted, and the output is not animated. If the animation matters, keep the GIF or use a video format. This limitation applies to every target here, including WebP, which can hold animation in principle but is not produced as an animation by this pipeline.

Conversion redraws the pixels, so EXIF orientation is applied and most camera metadata is dropped. That is usually good for privacy: location, device model and original timestamps do not travel with the converted copy. It also means color profiles and other embedded data may not survive. Keep the original file if any of that is part of your workflow. Output names use the source base name plus the correct extension, which makes a mixed batch easy to sort after download.