Sharpen AI by Topaz Labs – a Winner!

Not long ago, I had the opportunity to photograph Kiane, a lovely Minneapolis-based model. One of my favorite shots of her was spoiled because I missed the focus. (Note to self: avoid using manual focus lenses in situations that are risky.)

Enter Sharpen AI, a software product created by Topaz Labs. Their claim was sharpening “repair jobs” that border on the miraculous, so I thought I’d give them a try. I’ll cut to the chase and present the before and after. The left ‘before” side represents my best effort to sharpen the image in Lightroom, the right “after” image is with Sharpen AI by Topaz Labs.

I have seen some “knock your socks off” examples, but this is not one of them. On a mobile device you won’t be able to see the difference, but on a desktop computer, particularly around her eyes and mouth, the difference is obvious. And it is exactly the difference between a shot that doesn’t quite make it, and one that does.

A side note or two. The image was shot on a Fuji X-E2. Results with Sharpen AI seemed to be better if I did not try to sharpen in Lightroom first, but this should be considered an early result. Also, Sharpen AI has three different sharpening “specialty” modes and I would not have considered the softness in this image, (exposed at 1/250s) to be a result of motion blur. But Sharpen AI’s auto-detect said that stabilization mode was the best way to go, and indeed it was.

[twenty20 img1=”22753″ img2=”22757″ offset=”0.5″ before=”Before” after=”After”]

On the right is another example. Viewed from a desktop computer, the difference might be marginally noticeable, and on a phone is invisible. Now click on the image for a blow-up. The difference is obvious. (The left image is a little over-sharpened – I didn’t take the time to fix it.)  This helps illustrate an important point: sharpness is largely dependent on resolution, which is a function of viewing size and/or viewing distance.

This principle illustrates the “danger” of pixel-peeping; you can waste a lot of time and effort working to sharpen an image to use at a size or (less often) distance that renders the additional sharpening unnoticeable. Learning when and where it matters is key.