Many other details don't show much difference between A.S.S. The next examples are details of pictures coming directly from the cam: Ideally, upscaling is the first geometric transformation, which means that the input still has the resolution of the camera. Small versions of whole pictures are probably not the most typical use case for image upscaling. (I suppose it's also based on bicubic interpolation internally, but I may be wrong.) The function has a single adjustment 'Reduce Noise' which was set to zero for the test because the pictures have very little noise only. The 3rd alternative shown is a scaling method Adobe introduced in 2013 with Photoshop CC, called 'Preserve Details'. (The sharpening method used for the examples is 'unsharp masking'.) The interpolation results are a bit blurry, so another conventional action might be to sharpen them. Bicubic interpolation is what most image editors offer as the most appropriate method for image scaling. On a Windows PC or Notebook, just hold down the Ctrl key and spin the mouse wheel up or down until the left side pattern of this test stripe looks exactly like a miniature version of the right side pattern (chessboard-like).įor comparison, all examples show the results of two conventional methods as well. It's usually easy to get the original resolution shown.
