Generating sbsulute different images with the same generation data

For an example:

I tried to generate this one:

modelshoot style, (extremely detailed CG unity 8k wallpaper), full shot body photo of the most beautiful artwork in the world, english medieval witch, white vale, pearl skin,golden crown, diamonds, medieval architecture, professional majestic oil painting by Ed Blinkey, Atey Ghailan, Studio Ghibli, by Jeremy Mann, Greg Manchess, Antonio Moro, trending on ArtStation, trending on CGSociety, Intricate, High Detail, Sharp focus, dramatic, photorealistic painting art by midjourney and greg rutkowski, sexy,

Negative prompt: canvas frame, cartoon, 3d, ((disfigured)), ((bad art)), ((deformed)),((extra limbs)),((close up)),((b&w)), wierd colors, blurry, (((duplicate))), ((morbid)), ((mutilated)), [out of frame], extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), ((ugly)), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), cloned face, (((disfigured))), out of frame, ugly, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck))), Photoshop, video game, ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, mutation, mutated, extra limbs, extra legs, extra arms, disfigured, deformed, cross-eye, body out of frame, blurry, bad art, bad anatomy, 3d render

Size: 696x696, Seed: 1839638814, Model: protogenX34Photorealism_1, Steps: 30, Sampler: DPM++ SDE Karras, CFG scale: 10, Model hash: 44f90a0972

and below is what i get:

modelshoot style, (extremely detailed CG unity 8k wallpaper), full shot body photo of the most beautiful artwork in the world, english medieval witch, white vale, pearl skin,golden crown, diamonds, medieval architecture, professional majestic oil painting by Ed Blinkey, Atey Ghailan, Studio Ghibli, by Jeremy Mann, Greg Manchess, Antonio Moro, trending on ArtStation, trending on CGSociety, Intricate, High Detail, Sharp focus, dramatic, photorealistic painting art by midjourney and greg rutkowski, sexy,
Negative prompt: canvas frame, cartoon, 3d, ((disfigured)), ((bad art)), ((deformed)),((extra limbs)),((close up)),((b&w)), wierd colors, blurry, (((duplicate))), ((morbid)), ((mutilated)), [out of frame], extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), ((ugly)), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), cloned face, (((disfigured))), out of frame, ugly, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck))), Photoshop, video game, ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, mutation, mutated, extra limbs, extra legs, extra arms, disfigured, deformed, cross-eye, body out of frame, blurry, bad art, bad anatomy, 3d render
Steps: 30, Sampler: DPM++ SDE Karras, CFG scale: 10, Seed: 1839638814, Size: 696x696, Model hash: 44f90a0972, Model: protogenX34Photorealism_1, Clip skip: 2

Time taken: 2m 25.48s

Torch active/reserved: 3888/4768 MiB, Sys VRAM: 6144/6144 MiB (100.0%)

any idea?

I finally figured out it is related to the Clip skip. though there is a little bit difference, they at least looks similar.

modelshoot style, (extremely detailed CG unity 8k wallpaper), full shot body photo of the most beautiful artwork in the world, english medieval witch, green vale, pearl skin,golden crown, diamonds, medieval architecture, professional majestic oil painting by Ed Blinkey, Atey Ghailan, Studio Ghibli, by Jeremy Mann, Greg Manchess, Antonio Moro, trending on ArtStation, trending on CGSociety, Intricate, High Detail, Sharp focus, dramatic, photorealistic painting art by midjourney and greg rutkowski

Negative prompt: canvas frame, cartoon, 3d, ((disfigured)), ((bad art)), ((deformed)),((extra limbs)),((close up)),((b&w)), wierd colors, blurry, (((duplicate))), ((morbid)), ((mutilated)), [out of frame], extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), ((ugly)), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), cloned face, (((disfigured))), out of frame, ugly, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck))), Photoshop, video game, ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, mutation, mutated, extra limbs, extra legs, extra arms, disfigured, deformed, cross-eye, body out of frame, blurry, bad art, bad anatomy, 3d render

ENSD: 31337, Size: 512x690, Seed: 1053066257, Steps: 30, Sampler: Euler, CFG scale: 10, Model hash: c88e730a

modelshoot style, (extremely detailed CG unity 8k wallpaper), full shot body photo of the most beautiful artwork in the world, english medieval witch, green vale, pearl skin,golden crown, diamonds, medieval architecture, professional majestic oil painting by Ed Blinkey, Atey Ghailan, Studio Ghibli, by Jeremy Mann, Greg Manchess, Antonio Moro, trending on ArtStation, trending on CGSociety, Intricate, High Detail, Sharp focus, dramatic, photorealistic painting art by midjourney and greg rutkowski
Negative prompt: canvas frame, cartoon, 3d, ((disfigured)), ((bad art)), ((deformed)),((extra limbs)),((close up)),((b&w)), wierd colors, blurry, (((duplicate))), ((morbid)), ((mutilated)), [out of frame], extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), ((ugly)), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), cloned face, (((disfigured))), out of frame, ugly, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck))), Photoshop, video game, ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, mutation, mutated, extra limbs, extra legs, extra arms, disfigured, deformed, cross-eye, body out of frame, blurry, bad art, bad anatomy, 3d render
Steps: 30, Sampler: Euler, CFG scale: 10, Seed: 1053066257, Size: 512x690, Model hash: 44f90a0972, Model: protogenX34Photorealism_1, ENSD: 31337

Time taken: 48.81s

Torch active/reserved: 3046/3068 MiB, Sys VRAM: 4561/6144 MiB (74.24%)

Hi!

Getting the prompt, negative prompt, sampler, steps, model name and CFG is 98% of what's required to replicate an image, but unfortunately, there are MANY settings which aren't stored in a png header chunk, including;

ETA (noise multiplier) for ancestral samplers, sigma noise, sigma churn, always discard next-to-last sigma, whether the generating user was using the old Karras scheduler sigmas, or the old emphasis implementation - there are just so many combinations of settings in Web UI that it's very difficult to actually get 1:1 images from the prompt generation data you copy from Civitai sample images; you can get close - as you have - but it's very rare you can actually replicate an image exactly.

Hope that sheds some light on the situation for you!