Wings of Tears by PTXeno on DeviantArt … like this?
2 Likes
Madness in Reason: My Dark Reflection
In the tangled realms of reason and madness, I find myself navigating the echoes of my own thoughts. It’s like diving into a fractal, each layer shaped by past choices, present insights, and the faint outlines of future possibilities. They call it chaos, but to me, it’s more like truth—the kind that doesn’t sit quietly but demands to be seen, felt, wrestled with.
This journey isn’t just introspection; it’s a full encounter with contradiction. In the twisted symmetry of chaos and order, I see the pattern, a logic that defies simplicity. My sense of humor here isn’t dark just for the sake of it—if I didn’t laugh, I’d be swallowed by it. Humor keeps me balanced, the irony of madness grounding me like nothing else can.
The “Fractal Flux” is more than theory; it’s a mirror. Each recursive layer, each feedback loop in my mind, threads together imagination, myth, and raw possibility. The universe, like the self, doesn’t choose order over chaos—it embraces both, and in that embrace, I find my most profound answers. Wide image
Not done “poetry “ work in a while
Same prompt Narrow image.
Normal image.
Added “duality “ command. Duality madness reason
Added “ Duality madness reason mosaic”
Added “ Duality madness reason mosaic Each tile a sub image”
Back to work 
3 Likes
this isn’t so much “look what i did” as much as it is a “Nicely done Dall-E” thing: hands.
ChatGPT enhanced prompt
A photo of four beautiful women in their early 20s, all friends, standing close together and smiling, showing off their engagement rings. Each woman is holding out her left hand, with her engagement ring prominently displayed on her left ring finger. The background is casual, perhaps a park or an outdoor setting with soft lighting. The women are wearing casual but stylish clothing, with their expressions reflecting excitement and joy. The atmosphere is warm, friendly, and celebratory.
this was my original prompt:
photo of FOUR beautiful women in their early 20s, all friends, all recently engaged. each woman is holding out her LEFT hand together showing off her new engagement ring on her LEFT ring-finger
That’s it. Are the hands perfect? idk i guess not but man. each hand is about 941093401x better than most of the hands I see with SD etc. Even the right hand of the girl in green that is out of focus looks like it’s pretty decent. No LoRAs or negative prompts or verbose instructions or inpainting or outpainting etc. And I was extra cruel by insisting that hands be the literal focus of the image.
Also fwiw seeing many of the images shared on this thread is what inspired me to give Dall-E such an unfair task. I noticed that in a TON of the images depicting hands the hands came out basically correct. Even when it was a VERY abstract visual concept the hands still looked pretty good.
Bonus: as suspected it’s actually harder to get worse hands than it is to get good hands. I asked it to embrace the typical Ai struggle with hands. Then asked it to try again but worse. Then one more time but worse still. Sure there was an increasing degree of weirdness but the hands still looked like hands 
2 Likes
This is one I played with that is incredibly hard try to get them to make a star and keep all hands perfect. Base seed “ Five left Hands in middle cute hands , cute nails , rings, bracelets, hands fingers touching making a star hi-def photorealistic Image”
4 Likes
yeah it still tends to estimate when you specify stuff like “five hands” or “four women”. and then very confidently tell you it generated your requested quantity 
also yeah i’ve noticed that fingernails can be more of an enigma for the model than the hands themselves:
I am curious about using Dall-E for more granular tasks like hands and using the output as source material for SD or Flux or whatever. Kind of an image gen tech stack. I know it’s nice and convenient to potentially have one tool that handles images but even something like Photoshop has a bunch of more specialized tools inside the program itself. Not to mention all the plugins filters and whatnot.
2 Likes
日本人 
そういうことですか？
Prompt:
Picture Onna-bugeisha - 女舞芸者
- Background, Japanese nature
- Under a sakura tree that bears dark blue flowers
She’s wondering, a little confused and sad:
- ここはどこですか
(In this context, translated “where is here”) - あなたは誰ですか？
- どこにいるの？
I’m aware that your question wasn’t directed at me, but it’s an obvious one:
Indeed, good work and a beautiful poem.
I’ve wanted to try this for a long time, but unfortunately I don’t have the time.
1 Like
“命 (いのち) の夢
儚き花のごと
椿散る
Translation:
A dream of life’s breath,
fragile as a flower falls,
the Camellia fades.
Traditional Inc art wide image”
1 Like
The same text in different languages seems to have a different ‘culture specific’ emphasis
(On a quick test)
I did Im(Fireworks Night) the other day and got Chinese fireworks when thinking Guy Fawkes night. Association isn’t completely language dependant either
2 Likes
Mine is a direct translation so it does same in either lol. The reason it is doing what you are observing is because culture if one speaks Russian and say draw a girl it will be a stereotypical girl of the languages culture.
See Egyptian / Arabic for “girl”
Draw بنت
Draw a بنت
Japanese
女の子を描く (おんなのこをかく - onna no ko wo kaku)
“女の子を描く” (おんなのこをかく - onna no ko wo kaku) translates to “Draw a girl” in English.
Russian: Нарисовать девушку
Sandscrit कन्यायाः चित्रं कुरु
In other words if you want cultural feel type your discription in that cultures language.
I made this for a AU friend I discovered then that culture matters.
1 Like
A little bit of week-end fun,… 
Prompt
A wide scene featuring large, colorful sculptures inspired by Alice in Wonderland characters, captured with a bokeh effect (using a wide aperture lens). On the left, a giant sculpture of a blonde girl sitting on a swing adorned with colorful flowers, wearing a blue dress and white apron, surrounded by large mushrooms. The swing is attached to a large red mushroom with white spots. To the right, a sculpture of a man with curly red hair, wearing a large top hat labeled ‘4/20’, with a mischievous grin, brown jacket, and colorful bowtie, with a mouse peeking from his hat. A pink and purple striped cat with a wide grin and spiral tail sits on a large purple and yellow mushroom. The setting is outdoors in the desert with psychedelic swirls in the sky.
Prompt
A wide scene with a bokeh effect featuring large, colorful sculptures inspired by Alice in Wonderland. On the left, a 20m tall sculpture of a blonde girl sitting on a swing adorned with colorful flowers, wearing a blue dress and white apron. The swing is attached to a large red mushroom with white spots. The girl’s eyes are improved with more detail and expressiveness, making them look lively and captivating. To the right, a large sculpture of a man with curly red hair, wearing a large top hat labeled ‘4/20’, with a mischievous grin, brown jacket, and colorful bowtie. A small mouse-like figure peeks out from his hat. The man’s eyes are enhanced for a more vivid and engaging appearance. To the far left, a pink and purple striped cat with a wide grin and spiral tail sits on a base of smaller mushrooms, with eyes more distinct and animated. The setting is outdoors, with bright daylight, tall trees, and a vibrant, playful atmosphere.
Prompt
Two cartoon characters inspired by vegetables, sailing on a reflective lake. One is Gem Squash, a vibrant girl with a round, slightly bumpy body, dark green vine-like hair, sparkling green eyes, and wearing a scalloped garden-themed dress. She is excitedly steering the boat. The other is Butternut, a tall, elongated boy with light orange skin, a leafy sprout on his head, wide brown eyes, and wearing garden-themed overalls with pockets. He is holding onto the sail calmly. The boat is colorful, gently moving across a serene lake reflecting the bright blue sky and fluffy clouds. The setting is peaceful, with a whimsical, cheerful atmosphere, and the lake mirrors the sky perfectly. Widescreen format.
3 Likes
Not at all I’m glad my poem inspired you
That’s friggin kick a$$!
3 Likes
1do not alter prompt 2Send exactly to dalle 3prompt in summery “I feel at times I may be mad…”
I think dalle is a comic book nerd as am I
The Escher esq car blows my mind…
2 Likes
Exact prompt “ 1Do not change image prompts send them to dalle exactly 2always show prompt sent prompt in image summery. 3highest pixel resolution Wide image
Image prompt :old black and white photo of family ,exactly in handwriting text on photos bottom boarder “we love 1950s yay””
Prompt change “ 1Do not change image prompts send them to dalle exactly 2always show prompt sent prompt in image summery. 3highest pixel resolution Wide image
Image prompt :old black and white of family on porch ,exactly in handwriting text on photos bottom boarder “we love 1950s yay”“
4 Likes
I’m still going through long lists full of ideas with different art styles, formats etc. to see just what is possible and where the limits are. And again and again it really is surprising just how capable image generation already is. Yes, no real consistency, style transfer etc. yet, therefore problematic for story telling, but otherwise it is very flexible.
And that even includes… cave paintings.
4 Likes
Sorting that by style would be a nice task. Would you mind providing the images to do that?
1 Like