top of page

Erwin Blumenfeld was a German photographer who worked for Vogue, Harper's Bazaar and Life before and after World War II. He was one of the highest paid fashion photographers of his time. His style was unique, so it's not surprising that Midjourney knows him and can edit photos to show at least a little, if not a lot more than his photographic signature.


I asked ChatGPT to describe Blumenfeld's style. In summary, the result is this: He was influenced by Dadaism and Surrealism and incorporated elements from these artistic movements into his photography. His style was bold, original, creative and mysterious. He inspired later fashion photographers such as Irving Penn, Helmut Newton and Nick Knight.


Does it make sense to revive a dead man's hand, hand him a virtual camera  and generate computer-calculated "photographs" that he might as well have been able to take? Let's look at some pictures Midjourney took after entering a photo with lady.

The collage shown on the right indicates that Midjourney got the master's "handwriting" pretty good. From a prfessional counterfeiter's point of view one could be very satisfied.

The image below shows what comes out of giving Midjourney the upper human-made collage of computer-generated Blumenfeld photos as input. An overwhelming result! Normally you would put the four frames next to each other, compare them and choose the best one. I think the four pictures belong together.

What if I used the new result as input again? The first attempt failed, the model sometimes looked like a caricature.


This effect disappeared on the second attempt.


The third attempt (right) takes us away from Blumenfeld but stays in the 50's.

If you lined up the individual images horizontally, you would get snapshots of a woman in a trance who is constantly talking to herself without opening her eyes or mouth.

And I could go on like this forever.

bottom of page