I’ve been finding it’s very difficult to get the generator to create faces for the main character of my novel. She a short, plus sized woman at 40. This is one of the few successful image generations that is proportional to her body. Most of the time, I end up with a skinny woman, who doesn’t look even close to 40 years old. Any suggestions?
Here’s her current prompt:
Concept art in a semi-realistic style, bust portrait of Amanda, short plus-size woman, full softly rounded face with fuller cheeks proportionate to curvy figure, subtle double-chin crease when smiling, warm fair skin with healthy glow and natural texture, softly arched eyebrows, large almond-shaped green eyes proportionate to fuller face, long dark lashes, soft eyeliner, petite straight nose, full natural lips in soft coral pink, gentle smile lines adding maturity and warmth, extremely long brunette hair with subtle caramel highlights in loose waves, hair tucked behind one ear, large gold hoop earrings, confident warm inviting expression, soft warm lighting, neutral blurred background
Asking for prompts can get you a ban
uh oh someone asked for a prompt in the main perchance forum.
i will pray for your soul.
also putting on headphones so as to not hear the screams that are about to happen
More like asking for a better structure, not a prompt. Prompting is easy. Not knowing about the different types of generators and what they do, makes it a bit intimidating, especially when you have on a rudimentary understanding of the subject.
Plus, Google led me here. Speaking of which, if the moderator could move the thread to the appropriate location, that would help!
The sequence of your prompt tokens does matter. If you really want something, it should go first. As a rule of thumb, you could split the prompt into three parts, as proposed in the perchance t2i beginner guide. Description (who), Composition (where, doing what), Style. In your example, you prioritize the style, followed by an “empty” token of the woman’s name, followed by the okay description. So, I’d suggest you use chat gpt or a similar LLM, ask it to “create a detailed descriptive prompt, using natural language, in the format (description>composition>style) for a Chroma Unlocked text to image generator of a …<insert your description>”. You’ll get much better results. As a side note, chroma unlocked as well as most flux models are trained on a large dataset, so the results are greatly diverse. You can only achieve character consistency locally on your machine using comfyUI or similar tools.
Good stuff to know, thank you!
Also, where is this perchance t2i beginner guide?
You’re overloading your prompt with too many things that could conflict one another. Try to reduce the filler and you’ll get a more accurate result.
Lol, why has someone downvoted this?
