Has anyone noticed any benefit from Raw Details and Super resolution tools in Lr, in comparison to Image Size tool in PS?
Based on my experience using the two tools, they have different advantages and disadvantages because their fundamental mechanisms are different.
Photoshop's
Image Size tool appears to insert extra pixels through a process of interpolation in order to increase the size of the image. In other words, it stuffs extra pixels throughout the image based on the neighboring pixels at each insertion point. The main advantage of this approach is that the tool doesn't need to perform a sophisticated analysis of the image contents in order to make it larger. The main disadvantage, as far as I've been able to determine, is that the image tends to lose some apparent sharpness because the extra pixels inevitably sometimes disrupt high-contrast edges.
The new
Super Resolution tool in Lightroom and Adobe Camera Raw uses a neural network to recreate the image based on the visual elements the neural network has learned to identify by ingesting many sample images during a machine-learning phase. In other words, it doesn't so much
enlarge the image as
rebuild it by assembling larger versions of the visual elements. The main advantage of this approach is that the new, larger image has high-contrast boundaries that are at least as sharp as those in the original. The main disadvantage is that if the neural network misidentifies some visual element in the source file, it may introduce a larger and more prominent version of that element as an undesirable artifact in the destination file.
If Adobe has addressed this authoritatively, I'm not aware of it. Perhaps some of the forum participants here who test software for Adobe could provide a better explanation. I don't have any inside information about the inner workings of the respective tools so I'm just inferring what is happening under the hood from my observations of their outputs.
As an aside: I've been using Super Resolution to increase the apparent sharpness of some photographs even when I don't need the enlarged pixel dimensions for printing or some other purpose. By doubling the size of the image before exporting or printing it at the original resolution or smaller, the high-contrast edges become more compact and therefore seem sharper. This technique doesn't always make a significant difference in the final appearance of the picture, but more often than not it seems to help. I've found it to be especially useful when working with scans of analog media.
