Image Compressor

Good to know

• Compressing re-encodes the picture, so camera metadata (EXIF: date, camera settings, GPS location) is not copied to the new file. That is often welcome before a photo goes public, but it also drops your copyright field. To see what a photo carries first, open it in the EXIF viewer.
• PNG is lossless, and a browser has no way to re-compress a PNG losslessly. This tool does not pretend to: PNG files are saved as JPG or WebP, and the list shows the real result.
• A photo that is already well compressed can come out larger when saved again. When that happens the original is kept unchanged, and the list says so.
• Large photos, 48 MP and up, take a few seconds each, longer with a target size because several qualities are tried. They are processed one at a time so that a phone does not run out of memory.

When photographers need smaller files

• Sending photos to clients by email. Mail services limit the size of attachments, and a few full-size camera files reach that quickly. A target size lets you fit a selection into one message; for a whole gallery, a download link is still the better way.
• Web galleries and portfolios. Every kilobyte is downloaded by every visitor, often on a phone. Smaller files make portfolio pages load faster, which keeps visitors looking at the work instead of a blank frame. Resize to the display size first with the image resizer, then compress: fewer pixels save more than any quality setting.
• Upload limits on marketplaces and forms. Competition entries, stock and print marketplaces, job applications and government forms often cap the file size. Check the limit on the site itself, then choose it as the target here.

JPG or WebP?

JPG is the safe choice: every browser, phone, mail program, print lab and upload form opens it. WebP usually makes a smaller file at the same visible quality and can keep transparency, and every current browser shows it, which makes it a good choice for your own website. For anything a client or another service has to open, JPG is still the format nobody has to ask about. To change format without compressing, use the image converter.

Why 80% quality usually looks the same

JPG and WebP compression throws away first the fine detail the eye is least sensitive to: tiny variations in colour and texture rather than edges and brightness. At around 80% what is removed is mostly below what you can see at normal viewing size, while the file shrinks a great deal compared with 100%. Lower down, the losses start to show, first as blocks or banding in smooth areas such as skies, studio backdrops and skin, and as halos along sharp edges. The percentage is not a share of the picture kept, and the same number gives different results in different programs, so judge by eye: that is what the comparison is for.

What “target size” does

You give the largest file you can use; the compressor finds the highest quality that fits under it. It saves the photo at a high quality, checks the size, and halves the gap between a quality that fits and one that does not until the two are close, so the result lands just under your limit rather than far below it. It will not go below 40% quality. If a photo is still too large there, and you allow it, the longest edge is reduced in steps instead, with quality kept at 70% or more, because a smaller clean photo looks better than a full-size blocky one. Every change is listed next to the file, and if a target cannot be met at all, the list says so rather than handing you a ruined photo.