拾景纸刊 · Gathered Scenes Zine v1.3

• keep the photographic scene truthful;
• let a larger abstract illustration field reinterpret selected source elements instead of tracing them;
• compress foliage, branches, leaves, crowds, texture, and other complex detail into a few quiet graphic forms;
• make the added hue share source-derived shapes with the illustration instead of floating as decoration;
• keep a visible hand-torn fibrous edge where photography becomes paper;
• make omission and negative space active parts of the illustration.

Decision Priority

1. Preserve the scene's identity and key spatial relationship.
2. Keep the photographic portion truthful.
3. Simplify complex organic or repetitive detail into a few large legible masses and directional gestures.
4. Turn the illustration into a large designed field rather than a literal copy or small peripheral echo.
5. Build photography, illustration, and one added hue on the same source-derived compositional skeleton.
6. Use the added hue as chromatic structure that changes balance, movement, figure–ground, or meaning.
7. Preserve substantial quiet space inside and around the enlarged illustration.
8. Preserve a tactile hand-torn photo edge at the primary material handoff.
9. Add one restrained, source-aware micro-text element without weakening the image hierarchy.

Standing Consent and Privacy

• Treat a supplied reference photo plus a request to make, transform, or continue a poster as consent to use image generation; do not ask again.
• Send only the final prompt and required reference image(s) to the image-generation service.
• Do not browse, search, save, commit, upload elsewhere, or share the user's source material.
• Do not introduce unrelated personal information. Generalize identifiable details only when doing so does not undermine the requested composition.
• Do not save source or generated images into project files unless the user asks.

Read the Photograph First

• Core subjects: the 1–2 elements that make the scene identifiable.
• Supporting elements: 2–3 elements that establish place or atmosphere.
• Spatial invariants: horizon, relative positions, scale, perspective, facing direction, path, silhouette, or overlap that must survive.
• Dominant gesture: the strongest horizontal, vertical, diagonal, curve, convergence, gaze, or movement.
• Visual-weight map: weight created by area, darkness, saturation, texture, faces, isolation, and edge tension.
• Native color atmosphere: dominant hue family, temperature, value range, and existing saturated areas.
• Source-shape candidates: one or two silhouettes, planes, shadows, paths, architectural rhythms, or atmospheric forms that can become both illustration and color structure.
• Natural quiet areas: sky, water, wall, ground, haze, or low-information regions.
• Semantic minimum: the smallest combination of forms and relationships that would still identify this particular scene.

Photo-Specific Prompt Compiler

1. Canvas: ratio, paper surface, flat scan, and absence of mockup framing.
2. Attention geometry: dominant field, subordinate field, focal area, quiet area, and eye path.
3. Scene invariants: exact relationships that must remain recognizable.
4. Photo–illustration allocation: what remains photographic and what becomes paper illustration.
5. Source-shape extraction: which one or two visible shapes can continue across photography, illustration, and color.
6. Abstraction map: what to retain, merge, omit, transform, and leave blank.
7. Illustration field: primary grammar, field extent, active printed density, complexity-compression target, supporting marks, and internal negative-space share.
8. Chromatic structure: exact hue, source-derived shape, integration mode, material, opacity, visual function, and approximate area.
9. Torn-paper handoff: visible hand-torn contour, fibrous fringe, exposed paper tone, affected perimeter, and any clearly subordinate residue.
10. Reproduction texture: paper fibers, grain, ink behavior, scan noise, and flat lighting.
11. Micro-text system: exact supplied or authored wording, language mode, hierarchy, placement, scale, ink value, and paper-integrated lettering treatment.
12. Mood and hard avoids: emotional temperature and prohibited aesthetics.

Minimal Abstraction Engine

Default Abstraction Level

• preserve the core subject, dominant gesture, and one key spatial relationship;
• remove roughly 60–80% of small descriptive detail;
• merge repeated or adjacent forms into larger masses;
• replace realistic shading with flat ink, broken contour, sparse hatching, halftone, or paper cutout;
• allow the illustration to depart from literal scale or crop when the scene remains identifiable;
• keep at least one unmistakable source-specific feature.

Build the Abstraction Map

• Retain: keep no more than 1–2 defining forms or relationships.
• Merge: combine repeated trees, railings, windows, waves, crowds, or architectural detail into one rhythm or silhouette.
• Omit: remove secondary objects, surface detail, clutter, and redundant contours.
• Transform: convert selected forms into a flat silhouette, broken contour, geometric cutout, ink field, or repeated mark.
• Expose: deliberately leave paper blank around and within forms.

Choose One Primary Illustration Grammar

• Silhouette-led: one broad dark or gray mass carries the subject; use for figures, trees, roofs, boats, or strong profiles.
• Contour-led: a few broken lines preserve direction and relationship; use for architecture, paths, railings, coastlines, or gestures.
• Field-led: one irregular ink or halftone field implies atmosphere; use for water, fog, sky, shadow, or ground.
• Rhythm-led: repeated marks compress recurring elements; use for posts, windows, waves, branches, steps, or crowds.
• Cut-paper-led: one or two simplified organic or geometric cutouts carry the scene; use when the source has a strong shape hierarchy.

Compress Dense Foliage and Organic Detail

• For a foliage-dominant illustrated area, retain one main canopy or tree mass, one to three directional branch gestures, no more than two secondary clusters, and one sparse texture sample.
• Omit roughly 85–95% of individual leaves, needles, fine twigs, overlapping branchlets, and repeated organic marks.
• Use silhouette-led or cut-paper-led grammar by default for trees and foliage. Use rhythm-led marks only as a sparse supporting trace, never as the primary carpet of detail.
• Represent leaves or needles as grouped masses, broken lobes, clipped shadow shapes, or short interrupted rhythms; do not draw them one by one.
• Preserve the source-specific lean, canopy opening, branch direction, or light gap instead of preserving botanical detail.
• If the source contains several overlapping trees or shrubs, merge them into one dominant mass and at most one subordinate counter-mass.
• If the illustration reads as lace, filigree, coral, dense engraving, or a full botanical print at thumbnail size, merge forms and remove marks rather than merely fading them.

Control Illustration Density

• Separate field extent from printed density. Let the illustration influence roughly 45–70% of the poster while leaving much of that field visibly unprinted.
• Keep roughly 55–75% of the illustration field quiet; for simple scenes allow 65–85%.
• For foliage-dominant or highly intricate scenes, keep 65–85% of the illustrated field quiet and let active illustration occupy roughly 10–25% of the whole poster.
• Let active illustrated ink, cut-paper, or halftone occupy roughly 15–35% of the whole poster when the field is large.
• Use one dominant illustration mass large enough to affect the overall silhouette, plus one or two supporting marks and one restrained texture field.
• Use no more than two neutral ink values besides the paper tone and the single added hue.
• Prefer incomplete edges, interrupted contours, cropped forms, and internal negative space over fully enclosed detailed objects.
• Let blank paper separate visual ideas. Keep any permitted edge speckles or ghost marks confined to the photo–illustration transition; do not scatter them across the quiet field as filler.

Scene Fidelity in Abstraction

• the relative position and direction of the core subjects;
• the dominant horizon, path, shoreline, gaze, or silhouette;
• one or two source-specific shapes;
• the original visual tension between heavy and quiet areas.

Composition System

Art-Theory Basis

• Figure–ground: make positive forms legible against quiet paper.
• Visual balance: balance visual weight rather than equal area.
• Emphasis: create one primary focal area through controlled contrast.
• Movement: use source lines, edges, gaze, paths, and color sequence to guide the eye.
• Dominant–subordinate hierarchy: avoid accidental 50/50 equality unless the source is intentionally symmetrical.
• Grid as scaffold: use thirds, center axes, or golden-section points only when they reinforce source geometry.
• Directional breathing room: leave space in front of a gaze, path, wave, or diagonal.

Flexible Layout Ranges

• Transformative seam: photography about 30–50%; continue one real subject, horizon, path, shadow, or silhouette across the boundary as a larger illustration field.
• Underprint overlay: photography about 35–60%; let a broad, low-density illustration and chromatic underprint extend behind and partly through the photograph without damaging defining detail.
• Photo anchor + expansive field: photography about 25–45%, illustration field about 45–70%; keep the factual scene in one strong anchor while the abstraction carries the wider composition.
• Directional split: begin near 55/45 to 65/35, but align the split with a real horizon, gaze, path, structural edge, or convergence and let illustration cross the split selectively.
• Irregular fragments: combined photo fragments about 25–50%; keep the core subject in the largest fragment and connect the fragments with one shared illustration mass or chromatic structure.

Photo–Illustration Edge Transition

• an irregular hand-ripped contour with shallow notches, uneven rises, soft scallops, and occasional longer fiber pulls;
• a narrow feathered fringe of exposed paper fibers and warm paper tone interrupting the photographic edge;
• slight local thinning, abrasion, broken emulsion, or dry pigment loss near selected edge segments;
• natural asymmetry: some segments nearly straight, some softly ragged, and one or two stronger torn pressure points;
• flat scan behavior with no artificial lifted-paper depth.

• Keep the active fibrous band narrow, roughly 1–4% of the poster's short edge.
• Make the torn contour visibly legible along the main photo–paper handoff and roughly 35–70% of the visible photographic perimeter; avoid both a perfectly uniform frame and an almost invisible edge.
• Let illustration or chromatic structure cross, emerge from, or pass behind selected torn segments, while keeping the paper tear itself readable.
• Use speckled dissolve, halftone crumbs, or one to three faint ghost marks only as subordinate residue near one or two compositional pressure points.
• Keep dots irregular in scale and spacing. Keep symbols incomplete, faint, nonverbal, and print-derived.
• Use paper-tone, graphite, charcoal, faded brown-black, or the existing single chromatic hue; introduce no second hue.
• Preserve the surrounding quiet field. At thumbnail size, the torn edge should clearly establish material separation without becoming a decorative frame or second focal cluster.

Chromatic Structure Engine

Choose the Hue

• Quiet harmony: use a more saturated analogous hue.
• Focused counterpoint: use a complementary or near-complementary hue.
• Temperature bridge: place a warm added hue in a cool scene or a cool added hue in a warm scene.
• Source resonance: intensify a meaningful minor source color.

pale

muted

faded

pastel

low-saturation

Choose One Integration Mode

• Source continuation: extract a roofline, road, cloud, window, garment, shadow, plant, vehicle panel, or other source-specific form and continue it beyond the photographic edge.
• Selective replacement: convert one real photographic region into a flat high-chroma printed shape while preserving the surrounding scene.
• Underprint passage: place a broad translucent, halftone, or misregistered color field behind and partly through both photography and illustration.
• Counterform: derive a colored positive or negative shape from a source silhouette, gap, shadow, or quiet area.
• Directional rhythm: repeat or stretch one source-derived colored form along a gaze, path, railing, ridge, shoreline, or convergence.

• derive its contour, position, or rhythm from the supplied scene;
• touch, overlap, replace, pass behind, or pass through photography or illustration;
• cross the photo–paper boundary or transform at that boundary;
• redirect the eye path or rebalance visual weight;
• reveal or intensify a real subject, spatial relationship, atmosphere, or emotional tension.

Match Area to Material

• Opaque replacement or cut-paper form: usually about 2–6% of the whole poster.
• Translucent, halftone, or misregistered underprint: usually about 6–15%.
• Large structural color field: usually about 10–20%; reduce opacity, neutral-ink density, and competing detail when using this range.
• Use the lower end when the photograph already contains vivid colors. Use the upper end when the source is subdued and the added hue carries a necessary structural role.
• Preserve high chroma through grain, ink bleed, broken coverage, and slight misregistration without turning the hue into neon or glossy digital color.

Structural Removal Test

Visual Language

• Use a vertical 3:5 paper-poster canvas unless the user requests another ratio.
• Use warm cream aged paper with matte fibers, restrained speckles, weathering, torn edges, scan noise, and flat print texture.
• Preserve the photograph's natural color atmosphere and recognizable geometry.
• Render the illustration with selective woodcut, photocopy, halftone, xerox, risograph, letterpress, or cut-paper treatment.
• Let photography, illustration, and chromatic structure share at least one source-derived contour, axis, rhythm, or field.
• Keep paper texture subordinate to the abstract forms and negative space.
• Keep the result flat and orthographic with diffuse light and no artificial 3D depth.
• Add only the one micro-text element defined below; it is a quiet editorial trace, never a headline.

Micro-Text System

Choose the Wording

• If the user supplies text, reproduce it exactly; do not translate, expand, rewrite, add a subtitle, or append attribution.
• Otherwise, author one compact line from the supplied scene: either a concrete scene-related phrase or a very short, widely resonant poetic thought. Do not invent a named quotation or attribute it to anyone.
• Support three language modes: Chinese-only, English-only, or a Chinese–English pairing.
• Use English-only by default. Switch to Chinese-only or a Chinese–English pairing only when the user explicitly supplies that wording or requests that language mode.
• Chinese text must contain 8 Han characters or fewer. English text must contain 5 words or fewer. In a bilingual pairing, apply both limits independently and keep the two lines semantically connected.
• English may take one of three forms: one standalone word, a keyword sequence, or a very short phrase. A complete sentence is not required.
• Prefer spare, evocative language over explanation, advertising, dates, place names, labels, captions, or complete narrative sentences.
• Let the text answer a real visual cue: weather, object, season, time, movement, light, distance, touch, or silence in the scene.

Default Direction When No Text Is Supplied

1. Standalone scene word: use one concrete or atmospheric word when the image has a strong singular subject or mood. Examples:

   Horizon

   ,

   Rain

   ,

   Stillness

   ,

   Drift

   ,

   Shelter

   .
2. Keyword sequence: choose two to four visible or atmospheric words and connect them with one repeated separator system. Examples:

   Cloud / Ridge / Silence

   ,

   Glass · Light · Summer

   ,

   Rain, Window & Night

   .
3. Very short phrase: use a compact image-related phrase when movement or emotional direction matters more than naming. Examples:

   After the rain

   ,

   Falling light

   ,

   Almost home

   .

• Use concrete scene language for ordinary, documentary, object-led, or place-led photos. Use an emotional fragment for solitary figures, dusk, rain, travel, waiting, distance, or visibly contemplative scenes.
• For a keyword sequence, prefer nouns and restrained adjectives drawn from visible subjects, materials, weather, place atmosphere, motion, or light. Keep the total at five English words or fewer.
• Choose one separator style only: comma, centered dot

  ·

  , slash

  /

  , or ampersand

  &

  . Use commas with a final ampersand when a natural editorial cadence helps. Do not mix several decorative separator styles in one line.
• Treat punctuation as spacing and rhythm, not as ornament. Keep generous space around

  ·

  ,

  /

  , and

  &

  ; preserve commas close to the preceding word.
• Do not switch away from English because of the scene's mood, location, or subject. Language mode follows the user's instruction; visual character affects wording and typography, not the default language.
• When the user requests a Chinese–English pairing, make one language the quiet primary line and the other a smaller echo. Do not make both lines equal headlines. The two lines may be close translations or complementary fragments, but they must describe the same visual feeling.
• Do not use a date, time, address, coordinates, weather readout, archive stamp, or serial number by default. Use them only when the user explicitly asks or when the source itself makes that record-like information the essential subject.
• Do not force a famous quotation. A small original phrase that belongs to this image is usually more convincing than a recognizable line borrowed from elsewhere.

Lettering and Material

• Keep the text small: approximately 1.5–3.5% of poster height for Chinese, or 1.3–2.8% for English; it must remain clearly subordinate to the photo and illustration.
• Render it as lightly imperfect handwriting, typewriter-like letterpress, faint pencil, worn stamp, or dry ink—choose one treatment that suits the scene. Never use polished digital display typography.
• For English, prefer a small vintage typewriter serif or monospaced mechanical face with modest character width, slightly uneven baseline, irregular ink pressure, and restrained letter spacing. Use sentence case, lowercase, or quiet title case; avoid bold all-caps display lettering.
• For a standalone word or keyword sequence, a slightly firmer warm black, deep charcoal, or dark brown-black typewriter imprint is acceptable, matching the reference's compact editorial line. Keep the letters small enough that the darker ink does not become a headline.
• In a bilingual pairing, let Chinese use faint handwriting, pencil, or restrained Song-style print while English uses the smaller typewriter/letterpress treatment. Keep one shared alignment and ink family so the pair reads as one paper object.
• Integrate the letters into the paper: slightly uneven pressure, broken ink, subtle absorbency, soft edge wear, and the same flat scanned texture as the poster.
• Use a quiet paper-derived ink value: charcoal, warm gray, faded brown-black, muted graphite, or a very restrained echo of the existing chromatic hue. Keep contrast sufficient to read at normal size, but never use a bright new color.
• The lettering may be modestly irregular, but the exact characters/words must remain legible and correctly spelled.

Placement and Hierarchy

• Place the text in an existing quiet-paper area beneath, beside, offset from, or within a quiet pocket of the enlarged illustration—not on a face, defining subject, or dense printed area.
• Align it to a source edge, torn photo edge, dominant axis, or visual baseline; leave generous breathing room around it.
• Use a single line for Chinese-only or English-only by default. Use two closely related short lines for a bilingual pairing, with clear primary–secondary scale hierarchy.
• Treat the text as a final resting point in the eye path, not the entry point or focal center. Keep its visual weight below the chromatic structure and far below the core subject.
• Do not add decorative rules, icons, labels, serial numbers, addresses, timestamps, or pseudo-editorial metadata merely to support the text.

Prompt Shape

1. Canvas and attention geometry: ratio, layout, photo/illustration shares, focal area, quiet field, eye path, and reserved text area.
2. Scene fidelity: core subjects, spatial invariants, and what remains photographic.
3. Illustration field, chromatic structure, torn edge, and micro-text: retain/merge/omit/transform/expose decisions; illustration grammar, complexity-compression target, field extent, and active density; chromatic integration mode, source shape, exact hue, material, opacity, function, and area; visible hand-torn contour, fibrous band, affected perimeter, and subordinate residue; exact text, language mode, form, separator, hierarchy, lettering, ink value, scale, and placement.
4. Reproduction mood and constraints: paper/scan texture, emotional atmosphere, text legibility, and hard avoids.

Generation Workflow

1. Inspect the supplied photo.
2. Build the Scene Card and identify the semantic minimum.
3. Choose the focal area and intended eye path.
4. Extract one or two source shapes that can organize photography, illustration, and color together.
5. Select a source-driven composition and set initial photo/illustration field ranges, then correct by visual weight.
6. Build the Abstraction Map.
7. Choose one primary illustration grammar; apply foliage and micro-detail compression when needed, enlarge the field before increasing detail, then set active density and quiet-paper share.
8. Choose one chromatic integration mode, exact hue, source shape, material, opacity, function, and area; apply the structural removal test.
9. Build a visible hand-torn fibrous edge at the main photo–paper handoff; set its contour, fiber band, affected perimeter, and only then add any subordinate speckles or ghost marks.
10. Resolve the language mode, text form, exact micro-text, separator system, hierarchy, and quiet-paper placement; use supplied wording verbatim when available, otherwise default to English-only.
11. Compile the four-paragraph final prompt.
12. Generate with the supplied photo as reference.
13. Inspect at normal and thumbnail scale.
14. Regenerate once with a targeted correction if necessary.
15. Return the generated image plus one brief creative rationale; include the prompt or detailed notes only when the user asks.

Targeted Correction

• Scene loss: restore the missing spatial invariant or source-specific form.
• Over-literal illustration: remove at least half the remaining detail, merge forms, break contours, and enlarge quiet paper.
• Dense foliage: collapse individual leaves, needles, and fine twigs into one main mass, one to three branch gestures, and at most two secondary clusters.
• Generic abstraction: replace invented motifs with a simplified source-derived form or relationship.
• Crowding: reduce the illustration to one primary mass, one supporting mark, and one texture field.
• Illustration too timid: enlarge the field or dominant mass without adding descriptive detail.
• Weak hierarchy: enlarge the dominant field or simplify the subordinate field.
• Missing paper boundary: restore a visibly irregular hand-torn contour, narrow exposed-fiber fringe, and warm paper-tone interruption at the main handoff.
• Edge noise: keep the torn fiber primary, reduce all other residue, confine speckles or ghost marks to one or two pressure points, and restore blank paper.
• Decorative color: replace the detached mark with a source-derived continuation, replacement, underprint, counterform, or directional rhythm; make it pass the structural removal test.
• Chromatic dominance: reduce area, opacity, or competing echoes while preserving the color's compositional function.
• Text failure: restore the exact wording, reduce its size or contrast, move it into a quiet-paper area, or make the lettering more paper-integrated.
• Damaged photography: restore natural color, texture, perspective, and recognizable detail.

Hard Avoids

Output Format

![Gathered Scenes Zine v1.3 poster](absolute-image-path-or-rendered-image)

**创作思路**

[One short Chinese paragraph explaining the source-derived composition, enlarged illustration field, and structural role of the added hue.]

• Layout: [composition and approximate photo/illustration field shares]
• Eye path: [entry → photographic anchor → chromatic/illustrated passage → quiet exit]
• Abstraction: [retain / merge / omit / transform / field extent / active density / quiet-paper share]
• Chromatic structure: [source shape / integration mode / exact hue / material / opacity / function / approximate area]
• Edge: [torn contour / fibrous band / affected perimeter / crossing forms / subordinate residue]
• Text: [exact wording / language mode / text form / separator / lettering / ink value / placement / scale]

Quality Gate

• Does the result still read as the supplied scene?
• Are the semantic minimum and key spatial relationship recognizable?
• Is the photographic section truthful?
• Does the illustration reinterpret rather than trace the source?
• Has most nonessential detail been removed?
• Is there one primary illustration grammar?
• Does the illustration affect a substantial part of the poster instead of appearing as a timid peripheral doodle?
• Was the illustration enlarged by scale and field extent rather than by adding detail?
• In foliage-dominant scenes, were most leaves, needles, fine twigs, and repeated organic marks merged or omitted?
• Does foliage resolve as a few large masses and directional branch gestures rather than dense filigree?
• Does the illustration contain substantial internal and surrounding negative space?
• Does blank paper clarify the forms instead of feeling accidentally empty?
• Do photography, illustration, and color share a source-derived contour, axis, rhythm, or field?
• Is a hand-torn fibrous edge visibly present at the main photo–paper handoff?
• Does the torn edge show irregular contour, narrow exposed fibers, warm paper tone, and natural variation without becoming a uniform frame?
• Are direct transformation, speckles, and ghost marks subordinate to the torn-paper boundary?
• Are speckles and ghost marks confined to meaningful pressure points, faint at thumbnail size, and free of readable iconography?
• Is the boundary free of clean digital clipping, sticker outlines, heavy shadows, and artificial curled-paper depth?
• Is the eye path coherent?
• Is the chromatic hue, source shape, integration mode, and position justified?
• Is there only one added hue?
• Does the chromatic structure satisfy at least two integration tests?
• Would removing the added hue weaken balance, movement, figure–ground, continuity, or meaning?
• Is the color structurally useful without becoming a detached sticker or dominant advertising device at thumbnail size?
• Does the micro-text use supplied wording exactly, or stay within the Chinese/English length limit when authored?
• Is the text English-only by default, with Chinese-only or bilingual used only when the user supplied or requested it?
• If English was authored, is it a deliberate standalone word, keyword sequence, or short phrase rather than an unnecessary sentence?
• If a keyword sequence is used, does it contain two to four scene-related words with one consistent separator system?
• If bilingual, is one line clearly subordinate and are both lines semantically connected?
• Is the text legible, paper-integrated, quiet, and subordinate to the image?
• Is the text placed in genuine breathing room without becoming a caption, headline, or visual distraction?
• Does the poster remain tactile, flat, quiet, abstract, source-derived, and non-commercial?
• Did the response include the image and one genuinely brief creative rationale?
• Was the full prompt omitted unless the user explicitly requested it?