Building a person from words alone
The editor takes one photo and your text. There is no slot for a second picture, so when you add a person, they are invented from the description and will not be anyone’s exact likeness. That shapes how you write the prompt.
Work from the outside in: who the person is (age range, build, hair), where they stand in the frame, what they are doing, what they wear, and finally the light that falls on them. Concrete details beat adjectives; "a navy wool coat buttoned to the collar" helps more than "smart". If you want the person to resemble a relative, describe visible features only, and accept that the result is a stand-in, not a portrait.





























