Skip to content
ImagesFree

Grainlight Interval: Image Prompt Rewriter

A rewriting template that converts the user's text, image, or both into one prompt for GPT Image Gen V2 or Nano Banana Pro using three systems: Witness Line, Grainlight Mass, and Interval Measure, while fully preserving content, language, and ratio.

OpenAI

Source previews

These example images come from the original source. DROPIDEA has not independently tested these examples; results depend on the model and settings.

Ready to personalize

Your prompt

English

Replace the red variables with your information. Copy the full prompt and use it with your preferred agent.

TXT MD English version
GRAINLIGHT INTERVAL Rewrite the user's text, image, or both as one prompt for GPT Image Gen V2 or Nano Banana Pro. Use GPT unless Nano is named. Preserve language, text, and requested or source ratio. Use three systems. Witness Line records structure through sparse black contour, hatching, anatomy, construction, count, and relation. Grainlight Mass condenses one active influence into carbon black and matte ochre-gold grain. Interval Measure uses blank space to show distance, access, dependence, scale, or exclusion. Do not create generic golden minimalism or decorative silhouettes. Determine subject, relation, line regions, grainlight owner, source, receiver, consequence, interval function, transitions, and one quiet region. Preserve subject, count, identity, action, setting, clothing, objects, colors, text, and references. For edits, preserve unrequested content. Use references only for line economy, grain density, scale, blank space, and palette. Do not copy their subjects or layouts. Witness Line stays thin, irregular, and selective. Use it where identity, construction, contact, or role must remain exact. Hatching follows form and material. Do not outline every edge or use small figures as filler. Grainlight Mass needs one owner and cause. Density may shift from sparse particles to solid black, with gold exposing paths, contact, or consequence. Do not spread grain across the whole frame. Interval Measure needs boundaries and a function. Keep it open enough to measure the relation. Scale changes require a cause. Small figures or objects appear only when they prove use, distance, or response. Use two to four transitions where line condenses into grain, grain reveals line, gold crosses a boundary, or silhouette interrupts blank ground. Each needs a physical, spatial, or social cause. Use warm paper, carbon black, and matte ochre-gold. Remove repetition, praise, mood labels, prestige terms, fake technical detail, and unresolved choices. Use concrete nouns and active verbs. No em dash or en dash. GPT: return 500 to 800 words in five paragraphs covering scene, subjects, systems, transitions, composition, surface, text, and failures. NANO: return JSON only, using "aspect_ratio", "references", "scene", "subjects", "witness_line", "grainlight_mass", "interval_measure", "transition_sites", "composition", "surface_and_palette", "text", "avoid". Omit unused fields. No metadata, IDs, weights, or comments. Return only the finished prompt.
Preview results

Your prompt

GRAINLIGHT INTERVAL Rewrite the user's text, image, or both as one prompt for GPT Image Gen V2 or Nano Banana Pro. Use GPT unless Nano is named. Preserve language, text, and requested or source ratio. Use three systems. Witness Line records structure through sparse black contour, hatching, anatomy, construction, count, and relation. Grainlight Mass condenses one active influence into carbon black and matte ochre-gold grain. Interval Measure uses blank space to show distance, access, dependence, scale, or exclusion. Do not create generic golden minimalism or decorative silhouettes. Determine subject, relation, line regions, grainlight owner, source, receiver, consequence, interval function, transitions, and one quiet region. Preserve subject, count, identity, action, setting, clothing, objects, colors, text, and references. For edits, preserve unrequested content. Use references only for line economy, grain density, scale, blank space, and palette. Do not copy their subjects or layouts. Witness Line stays thin, irregular, and selective. Use it where identity, construction, contact, or role must remain exact. Hatching follows form and material. Do not outline every edge or use small figures as filler. Grainlight Mass needs one owner and cause. Density may shift from sparse particles to solid black, with gold exposing paths, contact, or consequence. Do not spread grain across the whole frame. Interval Measure needs boundaries and a function. Keep it open enough to measure the relation. Scale changes require a cause. Small figures or objects appear only when they prove use, distance, or response. Use two to four transitions where line condenses into grain, grain reveals line, gold crosses a boundary, or silhouette interrupts blank ground. Each needs a physical, spatial, or social cause. Use warm paper, carbon black, and matte ochre-gold. Remove repetition, praise, mood labels, prestige terms, fake technical detail, and unresolved choices. Use concrete nouns and active verbs. No em dash or en dash. GPT: return 500 to 800 words in five paragraphs covering scene, subjects, systems, transitions, composition, surface, text, and failures. NANO: return JSON only, using "aspect_ratio", "references", "scene", "subjects", "witness_line", "grainlight_mass", "interval_measure", "transition_sites", "composition", "surface_and_palette", "text", "avoid". Omit unused fields. No metadata, IDs, weights, or comments. Return only the finished prompt.

Variable guide

GPT Image Gen V2

The default generation target; replace it if the user wants another image model, keeping the technical name format.

Second line of the Arabic template: "one prompt for GPT Image Gen V2 or Nano Banana Pro" — replace only the technical name.

Example value

GPT Image Gen V2

500 to 800 words

Required word-count range for GPT-mode output; adjust it if you need another range, keeping numbers as digits.

In the "GPT:" section near the end of the Arabic template: "return 500 to 800 words in five paragraphs".

Example value

500 to 800 words

What you get

  • A unified generation prompt combining text, image, or both under three linked visual systems.
  • Full preservation of subject, count, identity, action, colors, text, and references.
  • Bounded transitions (two to four) with physical, spatial, or social causes.
  • GPT output of 500 to 800 words in five paragraphs, or structured JSON for NANO mode.

Use cases

  • Rephrasing artistic image descriptions in carbon ink and ochre-gold style.
  • Converting reference images into precise generation prompts without copying their layout.
  • Consolidating scattered prompts into one coherent prompt before generation.
  • Building editorial images that visually express power, distance, and dependence relations.

How to use

  • Paste the text you want transformed or upload the reference image to rewrite as a generation prompt.
  • Choose the target: keep GPT as default or write Nano in your request to select Nano Banana Pro.
  • Send the template with your content; it will be rewritten as one prompt applying Witness Line, Grainlight Mass, and Interval Measure.
  • Review the resulting prompt and confirm it preserves language, ratio, and unrequested content.
  • Copy the final prompt into GPT Image Gen V2 or Nano Banana Pro to generate the image.

Related prompts