Social media cropper

Each entry in the list is a pixel size, not a vague ratio: the frame you drag is locked to that shape, and the file you get is exactly those pixels, scaled from whatever you framed. Drag the frame to choose what stays in, drag a corner to take in more or less of the picture, and the frame will not leave the image. Fill the frame takes the largest area of that shape the picture allows, which is the usual answer.

The page says whether your source is big enough. If the area you framed is smaller than the target, the picture is being enlarged, and enlarging invents pixels: the result is softer than the original, and no tool on any machine can do better from the pixels you have. If the framed area is larger, it is scaled down, which is fine and usually sharpens perceived detail. The percentage is printed so you can see which of the two is happening before you download.

These sizes come from what the networks recommend, and networks change them without telling anyone. They are also all re-compressed after upload, whatever you send — which is the argument for sending the exact size rather than something huge, because your downscale is better than the one a feed will do in a hurry. JPEG is the default because photographs belong in JPEG; PNG is there for flat graphics and screenshots, and it will be much larger for a photograph.

The crop is drawn on a canvas and re-encoded, so no EXIF, GPS or colour profile survives into the file — which is the right default for something you are about to post. The file is read in this tab with FileReader, the page makes no network requests, and nothing you drop is stored; the Dark button keeps its own setting in this browser's local storage.