Question
Two things I don't understand about generative fill
- Why does GF hew to the color of the central objects. For instance, today, I have an old photo of two people dressed largely in blue. I requested a fill behind of “white curtains,” but the curtains are generated as blue. Why does GF not allow me to choose what colors I want?
- Why does GF usually misspell words?
Why can’t I dictate what will be generated and have GF simply follow directions properly?
