The surest way to fail at a full set of visuals from a generative model is to carefully rewrite the prompt for every image.<\/p>\n
However good each prompt is, the images drift from one to the next. You’re describing the same thing again every time, and no description comes out the same twice.<\/p>\n
\u2460 Lock the style block.<\/strong> Lighting, color, material, contrast and camera language go into one fixed paragraph that every image shares, word for word. Only the line about the subject changes from image to image.<\/strong><\/p>\n \u2461 Condition on reference images.<\/strong> Feed the palette and existing assets in as part of the input instead of describing colors in words. Color described in text is the least stable part of any prompt.<\/p>\n With those two changes, the drift went from “every image is different” to “now and then one gets thrown out.”<\/p>\n Our first version wrote the rules for light very specifically. Out of a few dozen images, half ended up using the same trick.<\/p>\n It didn’t look like a set of assets. It looked like a filter.<\/p>\n The fix was to split the rule into two layers. What the light is like<\/strong> (cold, dim, always leaving a visible physical effect) stays locked. What carries the light<\/strong> (leaking through a gap, falling on a wall, passing through frosted material, reflecting off metal) rotates from image to image.<\/p>\n \u2b50 Identity comes from repetition, but the thing to repeat is the quality. Fix the shape and by the fifth image people read it as a template.<\/p>\n \u26d4 We don’t generate faces to stand in for our team or clients, and we don’t generate interface screenshots to pass off as real work.<\/strong><\/p>\n Generated images handle mood and material. Evidence is always the real thing.<\/strong> This is about credibility more than taste. One “client case study” that someone spots as generated drags down everything real sitting next to it.<\/p>\n","protected":false},"excerpt":{"rendered":" We generated several hundred images and used about a dozen. What decided whether the style held wasn’t the prompt. It was two other things.<\/p>\n","protected":false},"author":1,"featured_media":1021,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_post_was_ever_published":false},"categories":[30],"tags":[],"class_list":["post-2142","post","type-post","status-publish","format-standard","has-post-thumbnail","category-craft-notes-en-us"],"yoast_head":"\nRepeat the qualities, not the shapes<\/h2>\n
One more hard line<\/h2>\n