I confess that I barely, if at all, remember making this photograph, nor can I say precisely where in London I made it. I grabbed it while in a particular mode of street photography that is the antithesis of how I might photograph, say, a landscape. I shoot handheld. I carry a small camera with a single small prime lens. I work very quickly, often not even slowing down to frame the subject, and continue on. In this case, I shot “blind,” holding the camera down at my side. (That created a little problem. More below.)
There are a few things about this photograph that could make some viewers a bit uneasy. First, I did not ask for permission and I doubt the subject was even aware that I made the photograph. There was no time, and if I had asked, the photograph would have been a very different thing. (BTW, sometimes I’m in public and I see that I might end up in someone’s photograph like this — and I remind myself that it is OK since I do it, too!) Second, the original image had a serious problem. It was badly tilted, perhaps as much as 20-degrees. I liked the image, but I had to crop radically to get the framing I wanted, and there were still problems. How did I fix it? For the first time, I allowed an AI tool to generate some content in the corners of the frame. Still not sure how I feel about that, but it seems worth experimenting with at least.
G Dan Mitchell is a California photographer and visual opportunist. His book, “California’s Fall Color: A Photographer’s Guide to Autumn in the Sierra” is available from Heyday Books, Amazon, and directly from G Dan Mitchell.
Blog | About | Instagram | Flickr | Facebook | Threads | Post | Email
Links: Articles, Sales and Licensing, my Sierra Nevada Fall Color book, Contact Info.
Scroll down to share comments or questions. (Click post title first if viewing on the home page.)
All media © Copyright G Dan Mitchell and others as indicated. Any use requires advance permission from G Dan Mitchell.
That’s a neat picture and I love the title. Re: the AI – I have no qualms about using software to fix problems with a picture IF I am not changing the major elements/components of the picture. Fixing the corners of this picture is no problem to me. Taking the 2nd person out of the picture. substantially changes the pictures and should be disclosed. I don’t use AI yet, I do it manually with software, so I guess I’ll have to start to learn what AI can do. I’m not a big fan of AI because of it’s potential risks to the world. Some good movies have been made about that.
So, all that said, I did have one picture I did not adhere to the above. We made a big hike to Lone Tree (and Weavers Needle) and I really wanted to get a picture of the Lone Tree. But our big group (~25 people) and others were always around the base of the tree as it was the only shade on a hot day. I finally took some pictures with as few people as I could and then Photoshopped all those people out. So now I do have a picture of just the Lone Tree that is just for my use/memories.
Hi Ernie: we are probably on the same page. I don’t think AI per se is a bad thing, but it has both good and bad potential. Would I let it completely produce a photograph? Other than as an experiment, no. Would I let it fix small issues under my control? This picture answers that question!