Resize Image to 1080x1080 Square

Where the 1080 Pixel Square Comes From

1080 is a screen number rather than a paper one. It is the short side of a full HD display, and phone apps settled on it as the width to render a full-bleed image at, so it became the width people prepare images to. A square at that width holds 1,166,400 pixels, a little over 1.2 megapixels - sharp on a phone that draws two device pixels for every layout point in a full-width slot, and small enough that a photographic JPEG lands in a few hundred kilobytes rather than a few megabytes. The shape is the part that costs something. Sensors are 4:3 or 3:2 and video frames are 16:9, so a square is a crop of almost everything. Filling the frame means scaling until the short side reaches 1080 and cutting the rest off the two long edges: a 4032 x 3024 phone photo becomes 1440 x 1080 and then loses 360 pixels of width, a quarter of the picture. A 3:2 camera file gives up a third. A 16:9 video still or a 9:16 vertical clip frame gives up nearly forty-four percent, which is why screen recordings and video grabs make poor squares no matter where the crop is placed. Fitting the whole image instead keeps all of it and spends the same proportion on background: that 4:3 photo lands 1080 x 810 with a 135 pixel bar above and below, so one square pixel in four is background rather than photograph. The trade is symmetrical, and the only question is which side of it you would rather pay. Enlargement is the other thing to watch, because 1080 is larger than a great many images that get sent here. An old 600 x 600 avatar has to grow 1.8 times, a 512 pixel icon slightly more than twice, and a thumbnail saved from a web page can be worse than either. Scaling up interpolates between pixels that exist; it never recovers a detail the camera did not record, so the multiplier appears in the results panel instead of being applied behind your back. Decoding, cropping and encoding are all done inside your own browser by the canvas element, so the picture is never sent anywhere.