April 2023 - Artificial Intelligence in Photography
I have used photography as my sole artistic means of image creation since the mid '60's. A long time playing with different cameras, studio equipment, and editing software, one would say. And all of a sudden everyone is talking AI. All photo-editing software is now integrating AI into their new builds. As an image creator, I was struck with this new possibility of instantly seeing how my thoughts might take shape without using the cumbersome hardware that I was accustomed to carrying with me on trips or using in the studio. After having generated some 10k images in Midjourney, I might say I am starting to get the hang of it. Upscaling my keepers and further tweaking them in Photoshop has become my other workflow. Yes, it is addictive and yes, I am hooked. I have recently discovered the possibility of having another AI engine (Clip Interrogator) analyze my own photographs and return a prompt that I can use in Midjourney to shed new light on a real photograph. This is incredibly interesting. A bit like a painter's canvas having been Jessoed and repainted several times over with the same theme in mind. I agree that AI Art is on the table for discussion by all and I think it is a good thing and needs to be part of the digital discussion. There will always be photographers around who will prefer to use a Polaroid, or some type of film camera and that is fine. We will still need photojournalists to cover current events and feed the news cycles and photographers to document recent discoveries whether they be in space, climatology, medicine, or any other field where a particular photo doesn't yet exist and instantly as it is published this new image will be available to Midjourney and other AI generating engines. We have entered a "brave new world".
Since writing the original text above back in January, Midjourney has published v5 with lots of extra goodies.
I now no longer need to switch between ClipInterrogator and the Midjourney/Discord application. Indeed, a new command was just added: /describe.
This new command allows users to upload from their device any image and Midjourney will analyze it and suggest four different prompts that can be used to reconstruct that image. Here is an example:
The other new and exciting addition is the /blend command which allows one to combine two or more images together.
By making judicious choices as to aspect ratio, color scheme, and image composition, one can instantly obtain images that might have taken a considerable amount of time in a photo editor.
By making judicious choices as to aspect ratio, color scheme, and image composition, one can instantly obtain images that might have taken a considerable amount of time in a photo editor.
From left to right: image 1; image 2; the result of the first two blended with the /blend command
February 15, 2021 - PICTORIALISM
My interest in the Pictorialist movement of the latter part of the XIXe century is a natural segue from my previous Blog article in which I discuss my thoughts about those who believe in the strict use of the camera as a tool to only document and, oh God forbid, those who prefer to create. Obviously, you know where I made my bed ;-0
Read all about it on this page: "Pictorialism"
November 20, 2020 - TO POST-PROCESS OR NOT
I am often asked what the original image actually looked like before I decided to transform it into the image I really had in mind when I reviewed the Raw. My main photo editor is Adobe Photoshop CC and a limited number of plugins: DxO Viewpoint to address perspective problems; DxO/Nik Collection for B&W conversions, Denoising, & the occasional color effect, and finally ParticleShop for the occasional brush effect.
I therefore decided to devote a page to this exercise we'll call "From SOOC to My Vision"
Appropriately, I called the page simply "Before & After".