Headwear
Do hat tags get the shape right?
Props reproduce more accurately than anything else measured here. A thing with a definite shape leaves the model little room to wander.
A hat covers the top of the face, so it moves the light as well. The brim shadow differs in each of the four.
Every cut below shares one seed, one base prompt and one sampler setting, and differs only by the tag marked in amber.
What each cut shows
witch hat- Wide brim and pointed crown, exactly as named — and the brim throws a shadow across half the face.
beret- Even the soft tilt is reproduced — the most natural of the hats.
baseball cap- Brim-to-crown proportion is right, and it invented a front logo as well.
straw hat- Woven texture and a ribbon band — a tag that carries material, not just shape.
How these were made
Shared prompt
masterpiece, best quality, amazing quality, very awa, 1girl, solo, mature female, adult woman, short black hair, brown eyes, white blouse, portrait, plain grey background, soft lighting, <tag slot>
Tap a tag to carry it over to the builder
Negative prompt
nsfw, nude, topless, bikini, swimsuit, lingerie, underwear, cleavage, suggestive, school uniform, pleated skirt, child, loli, teenage, bad quality, worst quality, worst detail, censor, jpeg artifacts, watermark, signature, text, extra limbs, bad hands, bad anatomy
Model
rinFlanimeIllustrious_v30 · euler_ancestral/normal · steps 28 · CFG 5 · 896×1152 · seed 777032
Takeaway
Props with a definite silhouette come out as written. Unlike style or expression tags, they rarely miss.
How far this result reaches
A hat hides the top of the head and can change the hairstyle itself. All four prompts say short black hair: three cuts keep the short hair visible under the hat, while the witch hat cut came out with hair well past the shoulders.
Measured on a single Stable Diffusion family checkpoint. The same tags can behave differently on NovelAI, Midjourney or another checkpoint. This is a reference point, not a verdict.