Why Plug-Ins
Why Plug-Ins: Enlarging or Resizing an Image
Resizing. Up-rezzing. Scaling. Upsampling. Interpolating. All of these terms are used to describe the act of change the size and resolution of a digital image. Why would you want to up-rez an image? You would want to do this if the resolution of your original image was too small for your desired output or if you wanted to crop a photo and then get a larger print out of it.
For example, maybe you have some digital photographs taken with a 3 megapixel camera and you want to print them on one of the latest wide format printers at 16x24 or 24x36 inches. The native resolution of that 3 megapixel file won't get you anywhere near close enough to print at that size so you will need to increase the file size and resolution.
There are many ways to resize an image in Photoshop. The oldest method is to use the Image Size dialog box in Photoshop. Until recently the only method available here was called Bicubic interpolation which simply doubles (or triples, quadruples, etc.) pixels to the size requested. This is an extremely fast method of resizing because it is simple in its approach. The downside to this method is that you are simply doubling pixels and when that happens, edge detail in your picture becomes soft and blurry because the contrasting colors at these edges are being averaged together.
There are some variations on Bicubic interolation that try to minimize the blurring and color averaging that occurs and to some degree, they are successful. The first variant is often referred to as the 10% stair step method. The strategy here is to increase the size of your image using Bicubic interpolation in 10% increments up to your desired new size. This does minimize the blurring to some degree, but the blurring and color averaging are still present and noticable because at the end of the day, it is still using the simple Bicubic algorithms. An additional downside to this method is that you have to repeat the process over and over again to get to your new desired file size. Doubling the resolution of your image requires you do this method 10 times.
To alleviate this repetitive task, Photoshop introduced a method of resizing called Bicubic Smoother for resizing your digital photos. The goal was to using the quick method of Bicubic interpolation but to deliver similar results that the 10% method would but in a single step. Bicubic Smoother has pretty much eliminated the need to increase the size of your file in 10% increments, but still, the base of the algorithm is the doubling nature of Bicubic and thus the results are still soft, blurry and as the name implies, smoother.
The method of resizing in Photoshop that onOne Software provides in a plug-in is called Genuine Fractals. Many people are familar with the term "fractals" from the Mandelbrot Fractal Set. Simply put, fractal theory goes something like this: all things in nature can be boiled down to a single pattern that repeats itself. Examples of fractals that can be observed in nature include the patterns of coastlines, rivers, clouds, mountains, lightning, trees, flowers and more. The key is that these patterns exist everywhere on both small and extremely large scales and it is just a matter of finding the patterns to repeat.
As a Photoshop plug-in, Genuine Fractals uses this theory to enlarge a digital photo to maintain edge detail as well as detail in non-edge, or continuous tone regions of your image. There's a whole lot of complicated math (or algorithms) going on in Genuine Fractals to ensure the highest-quality enlargement, but simply put, when determining the color values of the new pixels to be created for the enlarged file, Genuine Fractals examines a larger area around each pixel to get the context of how that pixel fits in with the others around it. The benefit is that when the new pixels are created, Genuine Fractals makes sure that these new pixels are in the same color range as the originals and thus avoids color averaging which leads to soft and blurry edges.
Learn more about Genuine Fractals 5.