Comments

Elf out in her natural habitat...

The *_(cosplay) tags aren't character tags. Keep them as general tags.

Kaguya hime!

Based on: https://twitter.com/kagerou_honey/status/1224579516216930306

was edited by me, the original had an extra 4 eye and senseless veins.

tremolo_measure said:

was edited by me, the original had an extra 4 eye and senseless veins.

Wow this style is clean! What'd you do to nail it down?

I'm having a hard time recreating this prompt, do you generate at 1024x1920 or upscale there?
Any other ideas for settings or variables I might be missing?

ksdc said:

I'm having a hard time recreating this prompt, do you generate at 1024x1920 or upscale there?
Any other ideas for settings or variables I might be missing?

I'm generating at 512x960 and then upscaling. I used CLIP skip: 1, no hypernetwork and latent scaling. Also check if you don't have "Use old emphasis implementation" selected. Nothing else comes to mind.
I checked the prompt just in case and it gives me the exact same image.

antlers_anon said:

I'm generating at 512x960 and then upscaling. I used CLIP skip: 1, no hypernetwork and latent scaling. Also check if you don't have "Use old emphasis implementation" selected. Nothing else comes to mind.
I checked the prompt just in case and it gives me the exact same image.

I got it, thank you!

It was CLIP:1 and Latent: Off.

Great output and prompts.
I' ve got a lot of very attractive output with img2img.

For some reason the colours change when viewing the smaller version
looks better when viewed at 100%

merocle said:

For some reason the colours change when viewing the smaller version
looks better when viewed at 100%

That's odd. I've never seen that happen with any other image uploaded here before.

Fujiwara_no_Mokou said:

That's odd. I've never seen that happen with any other image uploaded here before.

I noticed the same happening with my posts post #6224 and post #6318

antlers_anon said:

I noticed the same happening with my posts post #6224 and post #6318

Huh. My guess would be that it has something to do with how image samples are loaded. Maybe certain colors appear differently when scaled down, but when shown in original resolution, it just appears normally or something.

Looks like it's heavily based on iesupa's RWBY art

(CONTEXTUAL PROMPTS BEGIN HERE:0.0001) looking at viewer, perfect face, floral background, black background, silver background, padding, border, wings, standing on ledge, gradient background, patterned background,, 16-year-old, black hair, (grey:1.3) frilly (robe:1.2) with floral stitching, (french braid:1.4), demon horns, pink eyes, black hair ornament, (black:1.3) tips, (black eyeliner:1.2), feather hair ornament, hair over left eye, full body, ribbon choker, black thighhighs, thighhighs, (gladiator sandals:1.2), twintails, (low wings:1.3), black wings, solo, solo focus, long hair, floating hair, (fine fabric emphasis:1.5), long legs, very long legs, wide hips, frilly skirt, thigh cutout, thread, thigh strap, (pink trim:1), detached sleeves, (very long sleeves:1.2),, masterpiece, best quality,
Negative prompt: detached wings, baggy pants, asymmetrical horns, hoodie, gold trim, multicolored hair, 2girls, multiple girls, (demon tail:1.2), tail, green eyes, liquid, paint splatter, succubus, dakimakura \(medium\), huge ass, leotard, (pink robes:0.7),, lowres, bad anatomy, bad hands, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, watermark, username, blurry,
Steps: 40, Sampler: Euler a, CFG scale: 7, Seed: 3137355910, Size: 1024x1344, Model hash: 925997e9, Model: nai, Denoising strength: 0.7, Clip skip: 2, ENSD: 31337, First pass size: 512x704

(that first tag was a test with placeholders)

she's sitting on a much more comfortable-looking throne than the shows depiction. Never understood why one would risk cutting themselves like that

perfect. ai missed 1 mistake on the ear. The picture flew to my desktop

Huh, that's a name I haven't heard in years.

Rika outpizzas the hut any day...

Megumini

I'm curious, what GPU are you running to be able to generate 4096x4096?

What a cute form of imminent death.

Probably the best hands I've got in 50+ generations.

is this generated with novelai or naidiffusion (leaked model)?