Hey there thanks for reaching out. Its my first time playing around with an image generator AI. I want to compile super realistic images of humans in different scenes. So basically the same person but in different scenes (bar, gym, car, office, cinema…).

Here is the prompt i used + the result:

A full body portrait of a man, Rugged, muscular build, medium complexion, stern expression, short curly black hair, dark brown eyes, pronounced cheekbones, strong square jawline, slight beard, wisely serious face, confident stance, wearing a dark leather jacket, black t-shirt, sturdy dark jeans, combat boots, hint of brown undertones on the face, powerful and determined look, wearing a dog collar bracelet, realistic human form, sitting in a bar & enjoying a drink, wide angle, portrait orientation, head to toe, head and face fully visible, framed properly, proportional, high resolution, body fully visible, framed properly, high resolution

This is pretty much the ideal result.

Now the only thing i updated for the next image was instead of “sitting in a bar & enjoying a drink” i said “driving a supercar”

And this is what i got. Which is absolutely not what i wanted!

I used the same seed value (100). Also i used the same negative prompt which you can see below:

Cut-off face, partial face, cropped face, missing head, obstructed face, close-up, half-body, portrait only, upper body, bust shot, cropped, blurred, obstructed view, partial view, out of frame, cut-off face, obstructed view, blurred, close-up, shadows, no body

Every tip is appreciated a lot! Thanks!