Just realized my AI image prompts were way too vague

For months I'd been typing stuff like "a cat sitting on a chair" into Midjourney and getting weird results. Then last week I tried being super specific with lighting, camera angle, and mood like "overhead view, golden hour lighting, cozy living room, fluffy orange tabby" and the difference was night and day. It cut my retry rate from like 8 tries per image down to 2 or 3. Has anyone else found that adding more concrete details to prompts saves a ton of time?

2 comments

2 Comments

taylor_fox7d ago

Oh man, this hit me right in the gut because I had the exact same struggle. I spent like three months getting really weird art deco style stuff because I kept typing "vintage lamp" without saying anything about the era or vibe. The moment I started throwing in "60s table lamp, warm amber glow, dust motes in the light" it went from random junk to exactly what I pictured. Honestly it feels kind of dumb in hindsight that I was being so lazy with words when the whole point is to paint a picture for a literal machine.

margaretw417d ago

Right? Total game changer once you spell it out like that.