Pixelmator Pro, an increasingly popular macOS photo-editing program, has a new tool its creators say will let you blow up an image and maintain detail and sharpness “like they do in all those cheesy police dramas.”
ML Super Resolution lets users increase the resolution up to three times without the muddy, pixelated mess normally associated with upscaled images.
The ML, of course, is shorthand for machine learning, and the science is revolutionizing photography and post-production.
In the case of Pixelmator’s new feature, the algorithm recognizes edges, texture and contours. It creates like-pixels to interpolate. While it can’t recreate tiny details, ML Super Resolution recreates enough to reduce the bitmap look when a tiny image is enlarged.
Many photo editing apps have basic mathematical image-upscaling features. ML Super Resolution uses a “convolutional neural network” that scans the image and creates several versions containing a wide range of details.
The Pixelmator blog shares it super res tech in great detail and shows off the magic with before-and-after images.
Pixelmator Pro 1.5.4, featuring ML Super Resolution, can be downloaded from Mac App Store today for $39.99.