ToP: Topaz AI Sharpen Thing Whatever

Ctein appears to have returned to The Online Photographer, which is basically a good thing, I think. He's written a review of a piece of software, a machine-learning powered thing that sharpens photos for you.

The thing is clearly borderline magical in its results.

What it is doing, obviously, is re-painting a new, sharper, picture based on your picture. Whether it knows what a "stick" is or not is probably not even a meaningful question, but it can certainly replace a fuzzy stick with a sharp stick, and the sharp stick is credibly the same stick as if it had been photographed more sharply. It's not the same stick, but it's credible. It does this for every little detail of the frame.

You could do it too, if you were a competent painter, and had an incredible amount of time and patience.

Now, I could rattle on about "the index!" and how it's not a photograph any more, but a digitally rendered drawing based on a photograph and the potential cultural implications of that. Long time readers could probably write it for me at this point, so I shan't.

Instead I am going to talk about the vague sense of unease the whole thing gives me.

Ctein's photo is a good example. Why on earth is this thing something that needs to be saved? This is not a picture of a Yeti or an Alien. It's a vaguely dynamic but ultimately not very interesting picture of an eagle coming in for a landing (I think). These pictures exist already as technical exercises. The entire point of this exercise is that you nailed focus, because that's the hard part. Nobody gives a shit about the picture. Ctein fucked it up, and he lost the game. It's a stupid game, so who cares?

But going in afterwards to fix it, as if you had won the game, seems like a cheat. I know that he would argue that it's not about that at all, that this picture is Art and whatever means are necessary to restore the technical details that allow Ctein to oh god I can't go on, but it dribbles on for a few more lines without ever making the picture actually particularly good or noteworthy.

Even if it was a great picture, so what? Go shoot another one, and try not to screw the pooch this time. There are lots of great pictures out there. The problem is not that we don't have enough great pictures.

More generally, it is not at all clear to me that a software device which allows us to "salvage" more pictures is doing anything that needs to be done. We already have too damned many pictures. I consider a missed shot to be something of a mixed blessing. On the one hand, maybe it would have been great. On the other hand, it's a hell of a lot easier to edit out, now.