Midjourney prompts work better when you describe material, wear, light, and background

tisch_eins · x · 2026-07-21

This post argues that vague adjectives like “beautiful” are too underspecified for image models.

Using Midjourney as the example, it shows that stronger prompts should describe:

the material
surface wear and age
craft marks
lighting
background

The point is not to add more words, but to reduce ambiguity so the model has fewer decisions to guess. The post includes a concrete rewrite from “beautiful mask” to a much more specific ceremonial-mask prompt.

Original post →

Midjourney prompts work better when you describe material, wear, light, and background

More from Multimodal