High Fantasy
Prompt:
"Painted, Professional, TTRPG {Creature}, High Fantasy, Epic, Fantasy Landscape Background"
Dark Fantasy
Prompt:
"Lightly Textured Brushwork, Illustration, Gritty Dark Tones and Colors, Menacing, Fantasy Landscape Background, Standalone Piece, TTRPG {Creature}, Final Draft"
Sketch
Prompt:
"Sketch, Charcoal Medium, Graphite, Lines, Hand Drawn, Black and White, Final Draft, {Creature}"
Ethereal Dream
Prompt:
"Gentle Pastel Sketch, Water Color, Soft Monochromatic Tones, Wisps of Iridescent Color, Mix of Delicate Dreamlike Lines and Soft, Blurry Edges, {Creature}"
We generate a lot of content with each of these prompts and have been trying to tweak the keywords to reduce the occurrences of unwanted artifacts and text in the images. Despite our best efforts, current AI Image models such as DALLĀ·E 3 still introduce a large degree of variability. However we are very optimistic as these models evolve so should our ability to produce consistent results.
In our character and creature generation process, we accompany each creation with a "detailed visual description". This description acts as the {Creature} variable in our prompts. To reduce inconsistent results we also preform some minor refining of the visual description. opting for more precise terms when common words yield unpredictable outcomes. For instance, "bow" is often interchanged with "longbow" to prevent the unintended depiction of bowties.
Thank you for being a part of our community.