I'm always looking for resources/tools to help improve my post-processing and just came across
this YouTube video about a new Lightroom feature called Super Resolution. The short description is that with a couple of clicks Lightroom will use AI/machine learning algorithms to double the pixel width and height of a DNG image. That translates to a 4x boost in the total number of pixels. As explained in the video, a great application for this would be taking a heavily cropped photo and boosting the resolution to make larger prints or display in online venues where you need more pixels.
Well this all seemed like magic to me. I mean, come on, you can't just create pixels out of thin air without some kind of cost or tradeoff, right? Right?!?!
So I dug up some heavily cropped images from my K1-ii + D-FA* 50mm to give it a try for myself. The test images are shown below with links to my Flickr where you can download the full-sized originals. Compare those to the enhanced versions, which are in
this album. In short, I am blown away and the magic seems to be real! When I compare side by side, the enhanced versions look noticeably better without any negative consequences. Curious if anyone else has given this feature a try. What was your experience? Are there any images where Super Resolution introduced some weird artifact or performed poorly in some other way?
Hummingbird Diamondback Lovebirds