From prompt to LUT: how the workflow works

What the prompt parser reads

lutgen.app extracts your intended mood, tone, and constraints from plain English. It looks for descriptors such as color palette, contrast style, texture, grain, and camera profile cues.

The goal is to reduce ambiguity. Explicit direction like lighting references, destination software, and intensity is interpreted as stronger constraints.

• Mood and atmosphere (cinematic, pastel, moody, vintage, neon).
• Color direction (temperature, saturation, hue balance).
• Contrast and shadow detail preferences.
• Target app format and output dimensions.

How the model builds a graded profile

A conversion layer converts your prompt variables into a color transform profile with mathematically consistent channel behavior.

The system then generates the LUT values and applies format-specific post-processing constraints so files remain compatible with Resolve, Premiere, Lightroom, and Final Cut workflows.

1. Normalize prompt instructions into canonical grading variables.
2. Clamp ranges for clip-safe output and avoid harsh clipping.
3. Render candidate matrix-like LUT values for the requested cube size.
4. Export in the exact file format selected by your request.

What you should expect

Your output should reflect your prompt style while staying practical for production use. If it misses your target, small edits to prompt weight words and contrast direction usually produce a cleaner correction on the next version.

Tip: use concise, compositional prompts first, then add style refinements in a second pass.