horrible, terrible, but possible - only prompt

chroma style: DSLR-quality, SLR-saturated, SLR-faded and B&W

each style has been configured for specific lighting:

• DSLR - bright natural daylight

• SLR-saturated - daylight brightness

• SLR-faded - daylight diffuse

• B&W - direct sunlight and overexposed white sky

Experiment with cfg settings; cfg=3.5 can generate more interesting results.

DSLR style - negative prompt: low quality, ugly, out of focus, blurry, smudged, flat colors

SLR style - negative prompt (no)

B&W style - negative prompt (no)

Why did I use slight blur and out-of-focus in the prompt? AI models tend to generate high-quality faces, which kills style. Using these phrases allows the model to focus on style rather than facial quality.

I didn't define the color of the clothes in the prompt. This is intentional. The model will generate the color based on associations.

z-Image Turbo model. DSLR-quality (work but it doesn't generate 70s hairstyles.), B&W (work, but it doesn't generate 70s hairstyles.). SLR (not work - style description)