A Look at Adobe’s New Super Resolution Feature
Often, when editing images, we are required to crop the photo. This is certainly the case with bird and wildlife photography. We crop the image in order to fill the frame with our subject or to achieve a desired composition. Cropping the image tosses out valuable pixels, often leaving our image too small to display or print at the desired size. This is where Image Resampling becomes quite valuable. Specifically Adobe’s new Super Resolution feature, introduced in Camera Raw 13.2.
Let’s begin by defining the difference of Image Resizing and Image Resampling.
Image Resizing
With image resizing, the size of the image is changed without changing the actual number of pixels in the image. Simply select the size of the image we want, the total number of pixels in the image remains the same. The resolution of the image however, does not. What changes is the number of pixels per inch. The larger you make the print the less pixels per inch are available. The smaller you make the print the more pixels per inch.
Image Resampling
Resampling physically changes the number of pixels in an image. This can be used to increase the pixel count to get the resolution required to print or display an image of a certain size. Traditionally, resampling was performed using a technique called Interpolation. It uses the data from the pixels in the image to estimate the values of the new pixels to be created. When an image is upscaled (upsampling), more pixels are added to an image, while downscaled (downsampling) images have less pixels.
Photoshop and Lightroom employs multiple interpolation algorithms to complete the resampling process. Photoshop enables the ability to choose between the various resampling options, whilst in Lightroom the resampling method is chosen automatically.
Adobe are always looking ways to enhance the quality of images. As a result they recently introduced a new algorithm in Photoshop to improve the resampling process, called Preserve Details 2.0.
Here is a fantastic video from Matt Kloskowski on how to use these resampling methods to resize (enlarge) your photo.
Around the same time Photoshop introduced Preserve Details 2.0, Lightroom introduced a feature called Enhance Details to improve the resolution of images with texture, fine details, repeating patterns and sharp edges. What both of these new features had in common is that they are based on Adobe Sensei which employs artificial intelligence and machine learning to make it editing easier.
The Next Evolution in Machine Learning
Building on Preserve Details 2.0 and Enhanced Details, Adobe has introduced Super Resolution in Camera Raw 13.2. It will also be available in Lightroom and Lightroom Classic soon.
Super Resolution uses machine learning to increase the size of images while attempting to maintain sharp details. Details that can otherwise be softened from the simple upscaling process. Super resolution doubles the linear resolution of the photo which means the result will have double the pixels in the width and the height of the original photo, resulting in four times the original pixel count.
The Super Resolution process creates a new RAW file using the Adobe Digital Negative (DNG) format and Haitians the original RAW file. In addition to the benefits of the increase pixel count, the Super Resolution process also incorporates the Enhanced Detail feature for the better rendering of the edges and finer details of the image.
The technology itself is quite impressive. Super Resolution uses machine learning that interprets patterns in your image and compares it against millions if not billions of images that were used in training the artificial intelligence. It then uses that informations to reconstruct the image at double the resolution by creating detail that was not present in the original image.
Here are two great videos from Anthony Morganti that not only provides a demonstration of the technology, but also how to use it on images in Lightroom.
Summary
Welcome to the AI rEvolution! We are already seeing artificial intelligence being used in post processing software such as Luminar AI, Topaz Sharpen, Topaz DeNoise and Topaz Gigapixel. Now with the introduction of features like Preserve Details 2.0, Enhanced Details, Sky Replacement and Super Resolution, we are seeing the benefits of machine learning in Photoshop and Lightroom. I am really pleased to see the introduction of Super Resolution in Camera RAW, and coming soon to Lightroom. I am now able to produce larger, beautiful prints from my tightly cropped Bird images.
Check out our recent review of Luminar AI Post Processing Software. Download a FREE 7-Day Trial of Luminar AI.
