E

AI is a tool. With the right technique it is capable of more than anyone expects. Quality over quantity Every piece I post tries to do something innovative. If you want me to try something leave a comment. Most Recent Tests: -Text (success) -Group shots of multiple recognizable characters Progress update: Purely Generated Text either works or doesn’t - Onomatopoeia works without issue - Text on clothing and objects is finicky and random. It may choose the wrong words or devolve into a blob. Best when seeded beforehand with a manual edit - Background Text works when seeded and large, preferably with a canny image control to maintain it -Speech bubbles need to be created manually at a significantly large size. Iteration can occur as long as the strength of the reference image is .4 or below If you want me to try something leave a comment on a post. Give me a challenge.