By skyforbes 
Or how to blend and mix images and tools
We all know that creating an image with any generative AI app starts with the prompt. After all, that’s what text-to-image means. But, as it happens, there are cases when a prompt is not necessary.
And there are cases when, to finish the job, it is easier to use multiple generative AI apps than to struggle with one, even one as good as Midjourney.
I was working on some concept art for the story and was looking for ways to create images in a more controlled way than prompts allow.
The problem with Midjourney and Midjourney prompts is that Midjourney puts aesthetics above all, and while it is a great thing, it comes at the price of prompt adherence. That is why artists like me use personalization profiles and moodboards to steer Midjourney in the right direction.
Still, there are limitations even when using all those tools. It is a rare case when the artist’s vision and Midjourney’s interpretation of the prompt match.
So for my project, I had world-building images and characters, and I wanted to create images containing both. Of course, I could use the same prompt as the original worldbuilding images and character pictures as “oref,” but that did not create the…