I've been adapting my prompting for Midjourney v5 to include more natural language and it's working and freaking me out how good the results are.
🧵Examples w/ prompts in ALT tags & notes at the end
🧵Examples w/ prompts in ALT tags & notes at the end
/imagine street style photo of an elderly french woman with deep wrinkles and a warm smile, walking down the streets of soho, wearing a white gucci blazer made of cotton & black eyeglasses, natural morning lighting, shot on Agfa Vista 200, 4k --ar 16:9 --stylize 1000
MJ --v5
MJ --v5
/imagine elderly french woman w/ deep wrinkles & a warm smile, sitting in a charming soho cafe filled w/ plants, looking out the window, wearing a bright pastel linen blazer & floral silk blouse, natural light shining through the window & reflecting off her eyeglasses...
MJ --v5
MJ --v5
Some notes:
-Right now it seems --stylize 1000 produces the sharpest images (at least it has in most of my tests so far)
Here is a side-by-side of same prompt + seed, with --stylize 0 vs --stylize 1000. Stylize 0 (left) is much blurrier.
-Right now it seems --stylize 1000 produces the sharpest images (at least it has in most of my tests so far)
Here is a side-by-side of same prompt + seed, with --stylize 0 vs --stylize 1000. Stylize 0 (left) is much blurrier.
More notes:
-I pretty much copy & pasted one of my old prompts and just turned each variable into a more complete sentence rather than a comma-separated term, then I rearranged the order of the variables so the sentences flowed better. v5 seems to like that
-I pretty much copy & pasted one of my old prompts and just turned each variable into a more complete sentence rather than a comma-separated term, then I rearranged the order of the variables so the sentences flowed better. v5 seems to like that
Follow me @nickfloats to follow these tests.
Here are some side-by-sides using the same v4 prompts in v5 👇
Here are some side-by-sides using the same v4 prompts in v5 👇
Some color, fabric, and material tests 👇
Loading suggestions...