Skill issue.... dont read!
Every prompt i give is same outcome:
"oh...yes...here is corridor in the center of image..."
"oh you want few flowers?"
-"here is image with ONLY THAT FLOWER repreaded 50 times..."
No matter if my prompt is 100words or 200words..or just mere 50words.. same outcome..always and undeniable.
I can write prompt on my own, i can use llm, makes no difference.
I want river, path of forest, cabin...few animals?
OK... here is straight line river in the center with no cabin, no animals except some random critter... ALMOST PIXEL PERFECT SYMETRTY. lol... why? who wants that?
any indoor location will be a long corridor hallway...any outdoor will be a straight line from "first person view"...
And if write "lets make birds eye view" then it gives an image that is from upper stratosphere where every tree is like an ant.
What is this model made for? humans only? objects - to render one banana or something? Do i need to write an essay describing where everything is? I did it... same outcomes.. i want small house on left, small path, bushes, few fallen trees, small river that is snaking its way....
Nope..you try it.
My tools.. official comfyUI template. All settings unchanged. I never found any use for this model and its frustrating, as you can already notice.
For fake helpers who are pretend to be smart and mock me:
Dont ask for my prompts.. you give me the prompts. You show me, give me seed and prompt and i will try to replicate it.