{ "S1sh1": { "input_fingerprint": "dac8195923fe710c", "geometry": { "camera": { "look_target": { "x": 0.55, "y": 0.46 }, "origin": { "x": 0.53, "y": 0.4 } }, "entity_placements": [ { "slot": "E1", "subject_en": "terrified fleeing man", "x": 0.55, "y": 0.46 } ], "rationale_ko": "피해 남자를 U1 밀집 수풀 내부의 이동 경로에 배치했다. 카메라는 남자의 진행 방향 바로 앞에서 가슴 높이로 후진하며, 약간 비껴 선 정면 3/4 상반신 구도가 되도록 남자를 바라본다. 주변의 빽빽한 초목이 프레임 가장자리에 걸리면서도 뒤쪽의 위협은 드러나지 않는 배치다." }, "leakage": { "has_marker_leakage": false, "details_ko": "화살표나 기호 등 마커 유출이 없습니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.53, y=0.40) and looks toward (x=0.55, y=0.46); subjects outside the camera's horizontal field of view are out of frame\n- terrified fleeing man stands at (x=0.55, y=0.46)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): Track backward at the fleeing man's chest height in a slightly tightened frontal three-quarter upper-body frame, inheriting the path directly ahead of him. Keep his wide eyes and open mouth near the upper center while his shoulders drive toward the lens and dense foliage brushes the frame edges, preserving forward momentum without revealing the threat behind him.\n\nTHE LOCATION: 금월산 숲속·수풀 사건 현장\n\nSHOT TEXT (authoritative, Korean): 우거진 수풀을 헤치며 달려가는 중간 순간, 눈을 크게 뜨고 입을 벌린 채 겁에 질린 금월산 피해 남자의 상체 정면 구도.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S1sh4": { "input_fingerprint": "ae0b383af4b2bb75", "geometry": { "camera": { "look_target": { "x": 0.548, "y": 0.474 }, "origin": { "x": 0.474, "y": 0.492 } }, "entity_placements": [ { "slot": "E1", "subject_en": "airborne shadowy male attacker", "x": 0.546, "y": 0.462 }, { "slot": "E2", "subject_en": "injured fleeing man", "x": 0.552, "y": 0.488 } ], "rationale_ko": "U1 dense forest undergrowth 내부의 추격 동선에 두 인물을 배치했다. 피해 남자는 진행 중인 낮은 자세로 두고, 습격자는 지면 투영상 바로 뒤이자 거의 머리 위에 겹치도록 근접 배치해 도약 정점의 순간을 나타냈다. 카메라는 피해 남자의 옆쪽에 가깝게 두고 두 인물 사이를 향하게 하여, 실제 촬영에서는 허리 높이에서 급격히 올려다보며 하단의 피해 남자와 상단 중앙의 공중 습격자를 와이드 프레임에 함께 담도록 했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "스케치 내 마커나 기호 등의 노출 없음." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.47, y=0.49) and looks toward (x=0.55, y=0.47); subjects outside the camera's horizontal field of view are out of frame\n- airborne shadowy male attacker stands at (x=0.55, y=0.46)\n- injured fleeing man stands at (x=0.55, y=0.49)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): After the prescribed cut, place the camera at the man's side near waist height and tilt sharply upward, holding both bodies in a wide frame at the leap's apex. The fleeing man remains compressed into the lower portion while the airborne attacker occupies the upper center, making vertical character position the single dominant change from the preceding chase coverage.\n\nTHE LOCATION: 금월산 숲속·수풀 사건 현장\n\nSHOT TEXT (authoritative, Korean): 정체불명의 습격자(검은 인영·붉은 눈의 남자)가 공중으로 뛰어올라 금월산 피해 남자 위로 짙은 그림자를 드리운 찰나.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S2sh1": { "input_fingerprint": "51486374174a318b", "geometry": { "camera": { "look_target": { "x": 0.633, "y": 0.681 }, "origin": { "x": 0.514, "y": 0.085 } }, "entity_placements": [ { "slot": "E1", "subject_en": "shadowy male attacker with red eyes", "x": 0.636, "y": 0.681 }, { "slot": "E2", "subject_en": "female victim", "x": 0.626, "y": 0.686 } ], "rationale_ko": "카메라는 골목 통로(A1)의 북쪽 끝에 두고 전봇대 구역(P1)을 향해 길게 내려다보도록 배치했다. 두 인물은 희미한 가로등(A2) 아래 전봇대 바로 옆에 거의 같은 위치로 밀착시키되, 전신 실루엣이 구분되도록 미세하게 어긋나게 두었다. 인물과 전봇대는 화면 중심에서 약간 오른쪽에 모여 골목의 깊이가 유지된다." }, "leakage": { "has_marker_leakage": false, "details_ko": "도면 기호나 텍스트 누출 없이 장면이 그려짐." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.51, y=0.09) and looks toward (x=0.63, y=0.68); subjects outside the camera's horizontal field of view are out of frame\n- shadowy male attacker with red eyes stands at (x=0.64, y=0.68)\n- female victim stands at (x=0.63, y=0.69)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): From one end of the alley at adult eye height, hold a static long-distance wide frame with the utility pole and the two intertwined figures grouped slightly off center. Preserve their full bodies and the alley depth, allowing their disturbing physical closeness—not a closer camera distance—to carry the opening tension.\n\nTHE LOCATION: 도시의 어두운 콘크리트 뒷골목\n\nSHOT TEXT (authoritative, Korean): 희미한 가로등 불빛 아래 어두운 골목, 콘크리트 전봇대에 정체불명의 습격자(검은 인영·붉은 눈의 남자)와 효장동 피해 여자의 몸이 밀착된 전신 구도.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S2sh4": { "input_fingerprint": "9ce3491366128b85", "geometry": { "camera": { "look_target": { "x": 0.55, "y": 0.7 }, "origin": { "x": 0.5, "y": 0.84 } }, "entity_placements": [ { "slot": "E1", "subject_en": "collapsed woman", "x": 0.55, "y": 0.7 } ], "rationale_ko": "여자는 골목길의 전봇대(P1) 바로 옆 아스팔트에 힘없이 쓰러진 위치로 배치했다. 카메라는 골목 남쪽으로 물러나 지면 가까이에서 여자를 바라보며, 전신이 대각선으로 넓게 들어오는 돌리아웃 종결 구도를 반영했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "다이어그램 기호가 없습니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.50, y=0.84) and looks toward (x=0.55, y=0.70); subjects outside the camera's horizontal field of view are out of frame\n- collapsed woman stands at (x=0.55, y=0.70)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): Retreat at near-ground height until the woman's entire body fits diagonally across the wide frame, keeping the lens almost level with the asphalt. Her closed eyes and slack limbs remain the focus while the increased camera distance is the sole emphasized change at the end of the dolly-out.\n\nTHE LOCATION: 도시의 어두운 콘크리트 뒷골목\n\nSHOT TEXT (authoritative, Korean): 효장동 피해 여자가 두 눈을 감은 채 아스팔트 바닥에 축 늘어져 쓰러진 전신 구도.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S3sh3": { "input_fingerprint": "ec7da22c345f5242", "geometry": { "camera": { "look_target": { "x": 0.54, "y": 0.43 }, "origin": { "x": 0.54, "y": 0.51 } }, "entity_placements": [ { "slot": "E1", "subject_en": "female lead investigator", "x": 0.54, "y": 0.43 }, { "slot": "E2", "subject_en": "Korean detective", "x": 0.515, "y": 0.432 }, { "slot": "E3", "subject_en": "Korean detective", "x": 0.565, "y": 0.431 }, { "slot": "E4", "subject_en": "Korean detective", "x": 0.5, "y": 0.416 }, { "slot": "E5", "subject_en": "Korean detective", "x": 0.58, "y": 0.414 } ], "rationale_ko": "U1 밀집 수풀의 조사 지점에 여성 지휘 형사를 중앙 배치하고 한국인 형사들을 좌우 가장자리에 모았다. 카메라는 남쪽의 가까운 위치에서 중앙 인물의 상체를 정면으로 담아, 한쪽을 가리키는 손동작이 군집 속 유일한 강조 축이 되도록 설정했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "기호, 텍스트, 평면도 등 배치도 데이터 유출 없이 장면만 그려졌습니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.54, y=0.51) and looks toward (x=0.54, y=0.43); subjects outside the camera's horizontal field of view are out of frame\n- female lead investigator stands at (x=0.54, y=0.43)\n- Korean detective stands at (x=0.52, y=0.43)\n- Korean detective stands at (x=0.56, y=0.43)\n- Korean detective stands at (x=0.50, y=0.42)\n- Korean detective stands at (x=0.58, y=0.41)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): Within the local-news camera image, hold a static chest-to-eye-height medium frame centered on 혜수 among the detective group. Her upper body anchors the middle while her extended index finger cuts toward one side of the frame; surrounding detectives remain at the lateral edges so her directing gesture is the only emphasized axis.\n\nTHE LOCATION: 금월산 숲속·수풀 사건 현장\n\nSHOT TEXT (authoritative, Korean): 혜수가 한국인 형사들 한가운데 서서 집게손가락으로 특정 방향을 가리키며 입을 벌린 상체 정지 구도.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S6sh1": { "input_fingerprint": "8c9100e91bc26e00", "geometry": { "camera": { "look_target": { "x": 0.548, "y": 0.565 }, "origin": { "x": 0.563, "y": 0.601 } }, "entity_placements": [ { "slot": "E1", "subject_en": "smiling young woman riding bicycle", "x": 0.548, "y": 0.565 } ], "rationale_ko": "A1 해안도로 중앙에서 자전거를 타고 남동쪽으로 진행하도록 배치했다. 카메라는 자전거의 진행 방향 바로 앞에서 같은 속도로 이동하며 라이더를 정면으로 바라보는 근거리 위치로 잡아 상체 중심의 헤드온 구도를 만든다. A2 바다는 도로 오른쪽에 유지된다." }, "leakage": { "has_marker_leakage": false, "details_ko": "기호나 글자가 없음" }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.56, y=0.60) and looks toward (x=0.55, y=0.56); subjects outside the camera's horizontal field of view are out of frame\n- smiling young woman riding bicycle stands at (x=0.55, y=0.56)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): Move directly ahead of the bicycle at 수리영's chest height while matching her speed, holding a head-on medium upper-body frame. Her hands and the handlebar sit across the lower third, her smiling face remains centered, and her hair trails backward to preserve the forward travel that will later arc toward the sea-facing side.\n\nTHE LOCATION: 서해안 해안도로\n\nSHOT TEXT (authoritative, Korean): 밝은 해안가 도로 위, 수리영이 양쪽 귀에 하얀 이어폰을 꽂은 채 자전거 핸들을 쥐고 정면을 향해 미소 짓는 상체 구도, 주행 바람에 머리카락이 뒤로 날리는 찰나.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S8sh1": { "input_fingerprint": "5c6f191d8b297db1", "geometry": { "camera": { "look_target": { "x": 0.535, "y": 0.72 }, "origin": { "x": 0.505, "y": 0.655 } }, "entity_placements": [ { "slot": "E1", "subject_en": "young woman riding a bicycle", "x": 0.535, "y": 0.72 } ], "rationale_ko": "자전거를 탄 여성을 마트 입구 전면(B3)으로 이어지는 상업지구 도로 접근로(A1)의 열린 구간에 배치했다. 카메라는 진행 방향인 북쪽의 앞쪽이자 자전거 기준 약간 왼쪽에 두고, 뒤로 이동하며 전신을 안정적으로 담도록 라이더를 바라보게 했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "다이어그램 기호나 마커가 발견되지 않음" }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.51, y=0.66) and looks toward (x=0.54, y=0.72); subjects outside the camera's horizontal field of view are out of frame\n- young woman riding a bicycle stands at (x=0.54, y=0.72)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): From rider height ahead and slightly left of the bicycle, the camera tracks backward at matching speed to hold 수리영’s full body steadily against the moving street. She remains near the middle-right, one knee lowered over the pedal, while taxis, solicitors, stalls, and lit signs pass through the deeper frame without obscuring her path.\n\nTHE LOCATION: 대형 마트 외부·입구\n\nSHOT TEXT (authoritative, Korean): 어스름한 저녁 네온사인이 켜진 길거리 포장마차 사이 도로, 수리영이 자전거 안장에 앉아 한쪽 무릎을 아래로 굽혀 페달을 밟아 누른 채 앞을 예리하게 주시하는 전신 구도.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S9sh1": { "input_fingerprint": "0e9bb43b4302ec1c", "geometry": { "camera": { "look_target": { "x": 0.585, "y": 0.655 }, "origin": { "x": 0.608, "y": 0.675 } }, "entity_placements": [ { "slot": "E1", "subject_en": "young woman riding bicycle", "x": 0.585, "y": 0.655 } ], "rationale_ko": "E1을 A1 해안도로의 하단 직선 구간에 두고 남동쪽으로 빠르게 주행하는 순간으로 배치했다. 카메라는 자전거 진행 방향 기준 앞쪽이자 왼쪽인 바다 쪽에 가깝게 붙여 허리 높이에서 병렬 트래킹하며, 상체와 내려밟는 다리·페달을 함께 담는 클로즈 미드액션 구도를 만든다." }, "leakage": { "has_marker_leakage": false, "details_ko": "마커나 기호 없이 정상 렌더링됨." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.61, y=0.68) and looks toward (x=0.58, y=0.66); subjects outside the camera's horizontal field of view are out of frame\n- young woman riding bicycle stands at (x=0.58, y=0.66)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): At waist height off 수리영’s front-left side, the camera tracks parallel with the bicycle in a close mid-action composition. Her pitched torso occupies the upper center while the forceful descending leg and pedal remain visible in the lower-right, with the coastal road slipping laterally behind her and her clothing trailing opposite the travel direction.\n\nTHE LOCATION: 서해안 해안도로\n\nSHOT TEXT (authoritative, Korean): 어둠이 짙게 깔린 해안도로 위, 상체를 굽힌 채 한쪽 자전거 페달에 체중을 실어 내리밟는 수리영의 mid-action, 옷자락이 주행 반대 방향으로 휘날리는 찰나.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S11sh4": { "input_fingerprint": "2a542a07b00de530", "geometry": { "camera": { "look_target": { "x": 0.639, "y": 0.326 }, "origin": { "x": 0.619, "y": 0.308 } }, "entity_placements": [ { "slot": "E1", "subject_en": "disgusted young woman", "x": 0.641, "y": 0.328 } ], "rationale_ko": "인물은 외부 계단(S1)을 막 올라와 옥상방 철문(R4)으로 향하는 옥상 개방 접근부에 두었다. 카메라는 인물의 진행 방향 앞쪽에서 정면 축을 살짝 벗어난 매우 가까운 위치에 두고 얼굴을 바라보게 하여, 코와 미간의 불쾌한 표정을 담는 타이트 클로즈업과 철문 쪽으로 이어질 아크 동선을 반영했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "도면 기호나 문자 같은 유출이 없습니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.62, y=0.31) and looks toward (x=0.64, y=0.33); subjects outside the camera's horizontal field of view are out of frame\n- disgusted young woman stands at (x=0.64, y=0.33)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): At 수리영’s eye level just off her front axis, the dolly-in ends on a tight facial close-up that isolates her wrinkled nose and compressed brow. Her face sits slightly right of center, with her recoil turning one shoulder away as the camera prepares to arc past it toward the entrance.\n\nTHE LOCATION: 옥탑방으로 이어지는 가파른 외부 시멘트 계단\n\nSHOT TEXT (authoritative, Korean): 옥상 바닥에 두 발을 딛고 선 수리영이 불쾌감에 코끝을 찡그리고 미간을 잔뜩 좁힌 얼굴 클로즈업.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S13sh4": { "input_fingerprint": "e7de311e622aaac6", "geometry": { "camera": { "look_target": { "x": 0.66, "y": 0.7 }, "origin": { "x": 0.66, "y": 0.635 } }, "entity_placements": [ { "slot": "E1", "subject_en": "middle-aged female police officer", "x": 0.66, "y": 0.7 }, { "slot": "E2", "subject_en": "distressed young woman", "x": 0.66, "y": 0.665 } ], "rationale_ko": "외부 계단(S1) 하단과 뒤뜰(C3)이 만나는 지점에 두 사람을 매우 가깝게 마주 보도록 배치했다. 젊은 여성은 계단을 막 내려온 위치에, 여성 경찰은 뒤뜰 쪽에서 접근해 양어깨를 붙잡을 수 있는 정면 거리에 둔다. 카메라는 젊은 여성의 바로 뒤 계단 하단에 놓고, 그녀의 머리와 어깨 너머로 중앙의 여성 경찰을 바라보게 했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "기호나 평면도 유출 없이 정상적인 투시도로 그려졌습니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.66, y=0.64) and looks toward (x=0.66, y=0.70); subjects outside the camera's horizontal field of view are out of frame\n- middle-aged female police officer stands at (x=0.66, y=0.70)\n- distressed young woman stands at (x=0.66, y=0.67)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): From just behind 수리영 at shoulder height, finish the dolly at close conversational distance and look firmly across her shoulders toward 혜수. 수리영's head and seized shoulders form the lower foreground, while 혜수 occupies the central midground with both hands visibly locked onto 수리영's shoulders.\n\nTHE LOCATION: 낡은 다세대 빌라 마당·계단 아래 외부\n\nSHOT TEXT (authoritative, Korean): 혜수가 수리영의 양어깨를 두 손으로 강하게 움켜쥔 채 굳은 표정으로 마주 보는 정면 구도.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S15sh2": { "input_fingerprint": "18b8d775f17e5b63", "geometry": { "camera": { "look_target": { "x": 0.486, "y": 0.715 }, "origin": { "x": 0.443, "y": 0.739 } }, "entity_placements": [ { "slot": "E1", "subject_en": "adult woman offering a canned drink", "x": 0.466, "y": 0.704 }, { "slot": "E2", "subject_en": "seated young woman", "x": 0.497, "y": 0.719 } ], "rationale_ko": "B1 해안 버스정류장 부근의 해안 좌석 구역에 두 인물을 나란히 가깝게 배치했다. 앉아 있는 젊은 여성의 무릎 바로 위로 성인 여성이 캔 음료를 내민 순간이며, 카메라는 좌석 옆 무릎 높이에서 두 사람 사이의 손과 무릎을 향하도록 근접 배치해 얼굴은 프레임 밖에 두는 클로즈업 구도다." }, "leakage": { "has_marker_leakage": false, "details_ko": "도해 기호나 텍스트 누출이 없는 장면입니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.44, y=0.74) and looks toward (x=0.49, y=0.71); subjects outside the camera's horizontal field of view are out of frame\n- adult woman offering a canned drink stands at (x=0.47, y=0.70)\n- seated young woman stands at (x=0.50, y=0.72)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): At lap height beside the bench, finish the dolly in on 혜수's hand holding the canned drink just above 수리영's knees. Keep the can below forty percent of the image, using 수리영's lap and the bench edge as scale references while both faces remain outside the frame.\n\nTHE LOCATION: 해안가 버스정류장과 주변 도로\n\nSHOT TEXT (authoritative, Korean): 혜수의 손에 들린 차가운 캔 음료가 수리영의 무릎 위 허공에 멈춰 있는 클로즈업.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S15sh5": { "input_fingerprint": "1d0a198cb4d5ee78", "geometry": { "camera": { "look_target": { "x": 0.65, "y": 0.69 }, "origin": { "x": 0.43, "y": 0.75 } }, "entity_placements": [ { "slot": "E1", "subject_en": "woman calling out with hand cupped at mouth", "x": 0.47, "y": 0.73 }, { "slot": "E2", "subject_en": "departing young woman seen from behind", "x": 0.65, "y": 0.69 } ], "rationale_ko": "해안 버스정류장 인근(B1)의 열린 육지에 두 인물을 배치했다. 혜수는 카메라 가까운 좌측 중앙 전경에 두고, 수리영은 같은 시선축을 따라 오른쪽 배경으로 거리를 벌려 떠나는 뒷모습이 작게 보이도록 했다. 카메라는 혜수의 바로 뒤쪽에서 수리영의 출발 경로를 향한다." }, "leakage": { "has_marker_leakage": false, "details_ko": "도면 기호나 문자 등의 유출이 보이지 않음." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.43, y=0.75) and looks toward (x=0.65, y=0.69); subjects outside the camera's horizontal field of view are out of frame\n- woman calling out with hand cupped at mouth stands at (x=0.47, y=0.73)\n- departing young woman seen from behind stands at (x=0.65, y=0.69)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): At chest height near 혜수, hold a slightly tightened static angle past her toward 수리영's departing path. 혜수 occupies the left-center foreground with one hand cupped at her mouth, while 수리영's back remains smaller in the right background and continues away along the same sightline.\n\nTHE LOCATION: 해안가 버스정류장과 주변 도로\n\nSHOT TEXT (authoritative, Korean): 혜수가 한 손을 입가에 모으고 수리영의 뒷모습을 향해 입을 크게 벌린 정지 상태.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S24sh3": { "input_fingerprint": "21cf1ef218dade66", "geometry": { "camera": { "look_target": { "x": 0.548, "y": 0.56 }, "origin": { "x": 0.572, "y": 0.56 } }, "entity_placements": [ { "slot": "E1", "subject_en": "rain-soaked young woman riding a bicycle", "x": 0.548, "y": 0.56 } ], "rationale_ko": "해안도로(A1) 중앙부에서 인물이 도로를 따라 남쪽으로 달리는 순간으로 배치했다. 카메라는 인물의 바다(A2) 쪽 바로 옆에서 같은 높이와 속도로 병행하며 얼굴을 향해, 전방을 노려보는 젖은 측면 얼굴을 타이트하게 포착한다." }, "leakage": { "has_marker_leakage": false, "details_ko": "마커나 평면도 기호 없이 측면 시점으로 잘 그려진 인물 샷입니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.57, y=0.56) and looks toward (x=0.55, y=0.56); subjects outside the camera's horizontal field of view are out of frame\n- rain-soaked young woman riding a bicycle stands at (x=0.55, y=0.56)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): The camera travels parallel to Suriyoung at face height, maintaining the flow's immediate lateral distance and locking her rain-soaked profile into a tight close-up. Her eye and set jaw occupy the forward half of the frame while a narrow strip of coastal movement trails behind her, keeping her gaze aimed into the open space ahead.\n\nTHE LOCATION: 서해안 해안도로\n\nSHOT TEXT (authoritative, Korean): 수리영의 머리카락과 얼굴이 빗물에 흠뻑 젖은 채 앞을 매섭게 노려보는 측면 얼굴 클로즈업.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S26sh3": { "input_fingerprint": "5d183518cd8b10fc", "geometry": { "camera": { "look_target": { "x": 0.45, "y": 0.7 }, "origin": { "x": 0.45, "y": 0.67 } }, "entity_placements": [ { "slot": "E1", "subject_en": "resting young woman", "x": 0.43, "y": 0.7 }, { "slot": "E2", "subject_en": "ringing smartphone", "x": 0.45, "y": 0.7 } ], "rationale_ko": "수리영을 갯바위 휴식 구역(R1)에 앉히고 스마트폰을 바로 손이 닿는 오른쪽 바위 위에 배치했다. 카메라는 스마트폰을 향해 짧게 전진한 뒤 액정 바로 위에 도달하는 구도로, 평면상에서도 유효한 촬영 방향이 생기도록 원점과 주시점을 0.03만큼 분리했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "도면 마커나 기호 없이 렌더링된 장면입니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.45, y=0.67) and looks toward (x=0.45, y=0.70); subjects outside the camera's horizontal field of view are out of frame\n- resting young woman stands at (x=0.43, y=0.70)\n- ringing smartphone stands at (x=0.45, y=0.70)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): At rock level, the dolly-in finishes directly above the smartphone, framing its active screen as a close detail within the surrounding rock surface. The phone occupies roughly a third of the image, centered with enough context to establish that it lies within Suriyoung's reach rather than floating as an abstract screen.\n\nTHE LOCATION: 안개 낀 포구·시멘트 선착장·갯바위\n\nSHOT TEXT (authoritative, Korean): 갯바위에 놓인 수리영의 스마트폰 액정 화면에 '혜수이모'라는 발신자 이름이 떠 있는 클로즈업.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S26sh8": { "input_fingerprint": "cfc3f7a1f12a7c39", "geometry": { "camera": { "look_target": { "x": 0.624, "y": 0.642 }, "origin": { "x": 0.658, "y": 0.655 } }, "entity_placements": [ { "slot": "E1", "subject_en": "desperate young woman with clasped hands", "x": 0.624, "y": 0.642 }, { "slot": "E2", "subject_en": "young man facing her", "x": 0.653, "y": 0.649 } ], "rationale_ko": "두 인물을 H1 항구 선착장의 바다 쪽 출항 가장자리에 가깝게 배치하고 서로 마주 세웠다. 카메라는 남자의 가까운 어깨 바로 옆 가슴 높이 위치로 설정해 그의 어깨가 화면 오른쪽에 좁게 걸리면서, 두 손을 모은 여자의 상체와 올려다보는 시선이 중심부에 담기도록 했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "도면 기호나 문자 등의 유출 흔적이 보이지 않습니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.66, y=0.66) and looks toward (x=0.62, y=0.64); subjects outside the camera's horizontal field of view are out of frame\n- desperate young woman with clasped hands stands at (x=0.62, y=0.64)\n- young man facing her stands at (x=0.65, y=0.65)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): From just beside 인우's near shoulder at chest height, hold a static medium-close view of 수리영, whose upper torso occupies the center-left while her clasped hands remain visible along the lower edge. 인우's shoulder forms a narrow foreground frame on the right, and 수리영 lifts her gaze past it toward his face without changing the established confrontation axis.\n\nTHE LOCATION: 안개 낀 포구·시멘트 선착장·갯바위\n\nSHOT TEXT (authoritative, Korean): 수리영이 인우 앞에 마주 선 채 두 손을 모으고 절박한 눈빛으로 올려다보는 상체 구도.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S26sh12": { "input_fingerprint": "208b65399ca73fea", "geometry": { "camera": { "look_target": { "x": 0.625, "y": 0.635 }, "origin": { "x": 0.603, "y": 0.625 } }, "entity_placements": [ { "slot": "E1", "subject_en": "young woman with closed eyes and a faint smile", "x": 0.625, "y": 0.635 } ], "rationale_ko": "수리영을 H1 항구 선박 접안 가장자리의 바다 쪽에 배치했다. 카메라는 그녀가 바라보는 H2 해상 방향 바로 앞, 눈높이에 해당하는 가까운 위치에서 얼굴을 정면으로 향하게 하여 타이트 클로즈업의 최종 돌리 지점을 표현했다." }, "leakage": { "has_marker_leakage": false, "details_ko": "마커나 도면 기호 등의 누출 흔적이 없습니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.60, y=0.62) and looks toward (x=0.62, y=0.64); subjects outside the camera's horizontal field of view are out of frame\n- young woman with closed eyes and a faint smile stands at (x=0.62, y=0.64)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): At 수리영's eye height on her sea-facing side, complete the slow dolly into a tight frontal close-up and hold at the closest point, with her face occupying most of the frame. Keep only a soft margin of sea and fog around her as her eyelids settle and the faint smile forms, making facial distance the sole final emphasis.\n\nTHE LOCATION: 안개 낀 포구·시멘트 선착장·갯바위\n\nSHOT TEXT (authoritative, Korean): 수리영이 눈을 지그시 감은 채 옅은 미소를 띤 얼굴 클로즈업.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S28sh5": { "input_fingerprint": "a4093628b6a67c1a", "geometry": { "camera": { "look_target": { "x": 0.462, "y": 0.35 }, "origin": { "x": 0.443, "y": 0.371 } }, "entity_placements": [ { "slot": "E1", "subject_en": "woman extending an old photograph", "x": 0.405, "y": 0.4 }, { "slot": "E2", "subject_en": "seated man studying the photograph", "x": 0.493, "y": 0.321 } ], "rationale_ko": "두 인물을 조타실 앞 A1 오픈 데크에 마주 보게 두고, 여성이 남성의 눈앞으로 사진과 손가락을 뻗은 순간으로 배치했다. 카메라는 두 사람 사이의 손 높이에서 사진 바로 뒤까지 접근해 남성을 향하며, 여성의 손은 화면 좌하단에서 들어오고 남성의 흔들리는 눈은 우상단 배경에 남도록 사선을 잡았다." }, "leakage": { "has_marker_leakage": false, "details_ko": "도면 마커나 기호 등의 누출 없이 장면이 잘 표현됨." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.44, y=0.37) and looks toward (x=0.46, y=0.35); subjects outside the camera's horizontal field of view are out of frame\n- woman extending an old photograph stands at (x=0.41, y=0.40)\n- seated man studying the photograph stands at (x=0.49, y=0.32)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): At hand height between 수리영 and 인우, finish the dolly inches from the old photograph and hold on 수리영's fingertip above the small writing, with the photograph taking less than a third of the frame. Her extended hand enters from the lower left while 인우's wavering eyes remain a scale reference in the upper-right background, placing the suspended gesture directly before his sightline.\n\nTHE LOCATION: 낡은 소형 어선 갑판\n\nSHOT TEXT (authoritative, Korean): 수리영의 손가락 끝이 낡은 사진 표면을 짚은 채 인우 눈앞 허공에 멈춰 있는 클로즈업.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S28sh6": { "input_fingerprint": "0bf595d16f6f81bc", "geometry": { "camera": { "look_target": { "x": 0.43, "y": 0.36 }, "origin": { "x": 0.37, "y": 0.4 } }, "entity_placements": [ { "slot": "E1", "subject_en": "startled seated man", "x": 0.43, "y": 0.36 } ], "rationale_ko": "인우를 소형어선의 open exterior deck(A1) 중앙부에 배치했다. 카메라는 수리영이 사진을 내미는 쪽에서 인우 바로 앞까지 접근해 얼굴을 정면에 가깝게 바라보며, 짧은 촬영 거리로 당황한 표정의 클로즈업과 좌하단 사진 조각의 전경 구도를 만든다." }, "leakage": { "has_marker_leakage": false, "details_ko": "도면 기호나 텍스트 누출이 발견되지 않음" }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.37, y=0.40) and looks toward (x=0.43, y=0.36); subjects outside the camera's horizontal field of view are out of frame\n- startled seated man stands at (x=0.43, y=0.36)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): Track just beyond the photograph's edge and rise to 인우's eye height, tightening immediately into a face close-up from 수리영's side. His widened eyes and slightly parted mouth occupy the center while the photograph remains a thin, soft foreground sliver at lower left, emphasizing only the shift of focus from evidence to reaction.\n\nTHE LOCATION: 낡은 소형 어선 갑판\n\nSHOT TEXT (authoritative, Korean): 인우가 두 눈을 크게 뜨고 입을 살짝 벌린 채 당황한 표정의 얼굴 클로즈업.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." }, "S30sh3": { "input_fingerprint": "9237fe5f3b3616ad", "geometry": { "camera": { "look_target": { "x": 0.57, "y": 0.56 }, "origin": { "x": 0.44, "y": 0.47 } }, "entity_placements": [ { "slot": "E1", "subject_en": "weathered small fishing boat", "x": 0.58, "y": 0.58 } ], "rationale_ko": "유일한 지속 구역인 O1 오픈 씨 수역 중앙부에 어선을 두고, 선수는 우하단에서 중앙 쪽으로 향하며 짙은 안개 경계 안에 전방 절반가량 들어간 순간으로 설정했다. 카메라는 선수 높이에서 진행 방향 앞쪽이자 약간 측면에 짧은 거리를 두고 위치해 전방 선체를 되돌아보는 정지 와이드 구도를 만든다." }, "leakage": { "has_marker_leakage": false, "details_ko": "기호나 텍스트 등 마커 유출이 없습니다." }, "prompt": "Draw a single CAMERA-VIEW storyboard frame (16:9) of the scene staged\non the attached TOP-DOWN SITE PLAN, seen from the camera position\ngiven in the STAGING GEOMETRY data below.\n\nJurisdiction of sources — follow all three, they do not conflict:\n- The attached SITE PLAN is the SOT for what exists on the site and\n how its zones and landmarks are arranged relative to each other.\n- The STAGING GEOMETRY data below is the SOT for XY POSITIONS and\n horizontal heading ONLY: where each subject stands on that plan,\n where the camera stands, and which way it looks. Distances,\n directions and relative depths between subjects come from these\n coordinates. Subjects outside the camera's horizontal field of\n view are out of frame.\n- The CAMERA DIRECTION text below (when present) is the SOT for the\n VERTICAL angle, elevation, tilt and framing (eye-level, high angle,\n overhead crane, low angle, ...). If it asks for a high or overhead\n angle, draw that angle — the site plan being top-down does NOT mean\n the frame must be eye-level.\n\nSTYLE — LIGHT STORYBOARD FRAME: thin uniform dark outlines on plain\npaper. No shading, no hatching, no texture, no facial features beyond\na minimal head shape, no costume detail. Human figures are correctly\nproportioned full-body outlines in their canonical poses. Environment\nis reduced to a few placement lines (ground line, key landmark\nsilhouettes) — just enough to read positions and depth.\n\nSTAGING GEOMETRY — PLANNING DATA ONLY, NEVER RENDER IT: the lines\nbelow are normalized map coordinates on the attached site plan\n(x runs left→right 0..1, y runs top→bottom 0..1). Use them solely to\nplace the camera and subjects. Do NOT draw any of this data in the\nimage — no numbers, no coordinates, no labels, no arrows, no cones,\nno dots, no diagram symbols of any kind.\n- the camera stands at (x=0.44, y=0.47) and looks toward (x=0.57, y=0.56); subjects outside the camera's horizontal field of view are out of frame\n- weathered small fishing boat stands at (x=0.58, y=0.58)\n\nCAMERA DIRECTION (vertical angle / elevation / framing SOT): At bow height a short distance ahead and slightly to one side, hold a static wide view facing back toward the forward section as the small fishing boat crosses the fog boundary. Keep the bow below forty percent of the frame, angled from lower right toward center, with roughly half of its forward section lost inside the dense fog and open sea retained as scale context.\n\nTHE LOCATION: 서해의 먼바다\n\nSHOT TEXT (authoritative, Korean): 낡은 소형 어선의 앞머리가 짙고 탁한 안개 덩어리 속으로 절반쯤 밀려 들어간 찰나의 멈춘 구도.\n\nCRITICAL — THE SITE PLAN IS A STAGING REFERENCE, NOT CONTENT: do NOT\ndraw the plan itself, and do NOT draw any diagram elements in the\nframe — no circles, dots, wedges, cones, arrows, callouts, codes or\nany other overlay symbols. The output is a clean storyboard drawing\nof the scene as seen by the camera only.\n\nNo text, no letters, no numbers, no captions anywhere in the image." } }