Prompt structure
As another user found out, the prompt structure seems to be a bit different in the sense that the subject description is not just one block and we can not correctly use styles to work in different art styles (styles are an older thing from a1111 or also used in fooocus, where the main prompt, the subject, is in the middle and a short style prompt is put in front and a longer style part behind that main part) anymore.
In other models I could easily use dynamic prompts for my main subject and keep the styles mostly static or be able to easily switch them out on just one node.
In Chroma, there seems to be a break in that workflow and it results in not being able to set proper art styles anymore and the subject controls the art style as keywords or sentences with styles are ignored. Which kind of breaks my workflows and the way I work with ai imagery...
Is there an official statement to the question if the prompt structure is different enough or are we just tumbling around the darkness and find things that are not really there? (updated as my text was a bit strange before and some things did not need to be part of it)