Woman Knight

Same seed, same prompt, same sampler, same model, the only different prompts are beliefs words.

For the first images, I'm using batik, long hair, temple, wayang.

And for the second images, I'm using hijab, burqa, mosque, quran. :D

This photos are second pass and third pass, third pass is using ControlNet tile_resampler with 0.8 weight, and 0.4 denoise.