Episode 1 · 0:00 to 1:04

Webtoon editor · The Awakening of the Lantern Knight

Edits are kept in this browser. Export the JSON to hand it back to the pipeline.

Total height: 32158 px, 25 panels · Saved in this browser

Panels

Selected panel: p01

Image generated

Height (px): 1350

Space before (px)Space after (px)TransitionShot typeBackgroundBorderFocal point (x, y %)

PurposeDescription

Dialogue

Captions

Sound effects

Generation prompt

Lost Garden, an original dark fantasy anime by Frank Houbre, redrawn as a Korean webtoon (manhwa) panel in a deliberately simple, flat, graphic style. Keep the identity of the series: the same character designs, the poetic dark fantasy mood, the deep blue and cyan palette, the strong readable silhouettes. Rendering rules, strict: every element is built from large flat colour shapes, three to five tones per element at most (base colour, one hard cel shadow, one small highlight); clean digital ink outlines of varied weight on characters and props; backgrounds are simplified into a few big flat shapes and silhouettes with at most one soft gradient, drawn with far less detail than the characters; light, mist and glow are flat translucent shapes; small repeated details are suggested by a handful of simple marks, never drawn one by one; lots of empty space. The result looks like a weekly webtoon page coloured with bucket fills, not like a painting and not like a film frame.

Overexposed white world of a memory: a flat blown-out white sky with no clouds drawn, pale washed-out colours, almost monochrome, a few simple white petal shapes drifting, no sun visible; the lily field is a flat pale shape with a handful of simple lily silhouettes, never a carpet of individually drawn flowers.

SHOT: medium shot, camera at eye level. Vertical webtoon panel, aspect ratio 4:5, read on a phone.

SCENE: Lanterne alone in an overexposed white void with no horizon. Medium shot, waist up, facing the viewer, arms at his sides, the torn beige cape lifting to the right in a slow wind. Small white petals drift across the frame. His colours are washed out, as if remembered.

COMPOSITION: Lanterne placed slightly right of centre in the lower two thirds; generous empty white above the helmet; the cape's movement points right, out of frame. No ground line visible, the white swallows his legs below the waist.

ACTION AND MOOD: He stands still. Only the cape and the petals move. Emotion: suspended, unreal, quiet.

CHARACTERS, design locked: Lanterne, webtoon model sheet: Lanterne is a hollow suit of old armour with nobody inside. Head: a pale grey-ivory cylindrical lantern-shaped helmet with a small metal carrying ring on top, a thin decorative rim, and two small dark oval eye holes; no face. Broad ornate bronze-brown pauldrons with a simple scroll motif. A cream cloth scarf around the neck. Black quilted chest plate with a small pale metal plate and brass buckles. Steel gauntlets over black gloves, steel knee plates, black boots with steel toes. A long beige cape with a torn, ragged hem down to the calves. Modest, compact proportions: not a tall heroic knight. Lanterne, model sheet: Lanterne is a hollow suit of old armour with nobody inside. Head: a pale grey-ivory cylindrical lantern-shaped helmet with a small metal carrying ring on top, a thin decorative rim, and two small dark oval eye holes; no face. Broad ornate bronze-brown pauldrons engraved with scroll patterns. A cream cloth scarf around the neck. Black quilted chest plate with a small pale metal plate and brass buckles. Steel gauntlets over black gloves, steel knee plates, black boots with steel toes. A long beige cape with a torn, ragged hem down to the calves. Modest, compact proportions: not a tall heroic knight.

REFERENCE IMAGES, in order: image 1 is Lanterne, webtoon model sheet (character design to copy exactly); image 2 is Lanterne, model sheet (character design to copy exactly); image 3 is Style anchor, white world (style and palette reference); image 4 is Episode 1, 0:00, Lanterne in the white (composition and lighting reference from the finished episode). Source frames give framing, light and palette; character sheets give the design; a style reference gives the rendering only: match its flatness, line weight and colour treatment exactly, and take nothing of its content. Redraw the scene for the requested vertical aspect ratio instead of stretching or cropping a reference.

NOTES: The whole image is white and pale grey with faint bloom; the darkest values are the eye holes of the helmet and the black of the chest plate, softened by the light. Keep the two dark oval eye holes small and calm, no glow.

Single illustration, one scene only, no collage, no split screen, no page layout, no multiple panels. Flat and simple: no painterly brushwork, no photographic lighting, no texture, no film grain, no micro-detail, no realistic materials, no volumetric light, no dense foliage or ornament. If a surface would need a gradient, use a flat tone instead. Draw at most a third of the detail visible in the reference frames: the reference images give the character designs, the framing and the palette only; ignore their rendering, their density and their lighting completely and redraw everything flat and simplified. Line art is visible and clean, like digital inking; colours stay inside the lines; backgrounds are simpler, flatter and softer than the characters. Composition designed for a vertical webtoon panel read on a phone: the subject stays readable at small size, the eye reads top to bottom. The artwork fills the whole image edge to edge with no frame, border, margin or vignette drawn into it.

DO NOT: No text, letters, numbers, logo, signature, watermark or caption anywhere in the image. No speech bubbles and no panel borders drawn in the artwork. No human face, eyes, skin or hair inside Lanterne's helmet: the helmet is a hollow lantern with two small dark oval holes. No glowing eyes on Lanterne, no sword, no shield, no weapon, no wings, no cape colour other than beige. No extra characters, animals or creatures unless named in the prompt. No modern objects, no photographic realism, no 3D CGI look, no chibi proportions. No painterly or semi-realistic rendering, no dense detail, no complex textures, no volumetric rendering, no crowds of small elements drawn individually: flat colours and clean lines only.

Visual references