The more you do it, the better you get at text to image - Prompt Crafting. I find, no matter what model I'm utilizing, to be direct with the parameters and the theme/s.; PROMPT >> professional artist rendered t-shirt design, vector, realism, glitchy, cyberpunk,; the Grim Reaper, in metallic cyber armor - rowing a boat across a lake towards a castle Garbage in, garbage out. Is my input quality input? Am I constantly running a routine scan for areas where the prompt could improve? I could muck about with different words; - poster, instead of t-shirt - contour instead of vector, etc - I could rewrite the Garb the Reaper is wearing - differing outfits. - I could replace "the Grim Reaper" with "the Accidental Mutant" or "an Astronaut" or whatever persona. I could spend an hour brainstorming how many different roles could the Grim Reaper be employed in? (Write the roles down on an size A4 lined pad, cross off each role as you go through the labor of e...