name

ai-video-prompting-skill

description

Hollywood-grade AI video generation prompting (Runway, Pika, Kling, Luma, Stable Video). Use for text-to-video prompting, image-to-video workflows, camera controls, motion dynamics, character animation, scene design, lighting, cinematic storytelling, multi-platform workflows, and production-quality AI videography. Also use for Thai keywords "สร้างวิดีโอด้วย AI", "เขียนพรอมต์วิดีโอ", "Runway", "Sora", "Pika", "AI วิดีโอ", "วิดีโอ AI", "text to video", "ภาพเป็นวิดีโอ", "กล้อง AI", "เคลื่อนไหว AI", "แอนิเมชั่น AI", "แสงสว่างวิดีโอ", "ถ่ายหนัง AI", "สร้างหนัง AI", "ผู้กำกับ AI", "วิดีโอคุณภาพสูง".

🎬 AI Video Prompting Skill - Hollywood Production Grade

Version: 1.0 Last Updated: 2025-10-24 Expertise Level: Professional Director/Cinematographer Platforms: Runway Gen-3, Pika Labs, Kling AI, Luma Dream Machine, Stable Video Diffusion

🎯 Core Prompt Structure

Universal Prompt Formula (Works All Platforms)

[SHOT TYPE] of [SUBJECT] [ACTION] in [ENVIRONMENT],
[CAMERA MOVEMENT], [LIGHTING], [TIME/WEATHER],
[STYLE], [MOOD], [TECHNICAL SPECS]

Component Breakdown

1. Shot Type (Required)

Defines framing and composition.

Categories:

Character Shots:

Extreme Close-Up (ECU) - Eyes, lips, hands, small details
Close-Up (CU) - Face fills frame, emotional intimacy
Medium Close-Up (MCU) - Head and shoulders
Medium Shot (MS) - Waist up, conversational
Medium Long Shot (MLS) - Knees up, action visible
Full Shot (FS) - Entire body, head to toe
Long Shot (LS) - Character in environment context
Extreme Long Shot (ELS) - Wide landscape, tiny character

Specialty Shots:

Over-The-Shoulder (OTS) - Looking past character
Point of View (POV) - Camera as character's eyes
Bird's Eye View - Directly overhead (90° down)
Worm's Eye View - Ground level looking up
Dutch Angle (Canted) - Tilted frame (tension, unease)
Two-Shot - Two characters in frame
Three-Shot - Three characters in frame
Group Shot - Multiple characters

Environmental Shots:

Establishing Shot - Location introduction, wide context
Cutaway - Detail shot (clock, hands, object)
Insert Shot - Close detail for emphasis

2. Subject (Required)

Who/what is the focus.

Character Description Template:

[AGE] [GENDER] [ETHNICITY] [DISTINCTIVE FEATURES]
wearing [CLOTHING DETAILS],
[EXPRESSION/EMOTION]

Examples:

young Asian woman with long black hair wearing red dress, smiling warmly
elderly Black man with gray beard wearing blue suit, looking thoughtful
teenage boy with glasses wearing hoodie, surprised expression

Object Description:

[ADJECTIVES] [OBJECT] [MATERIAL] [CONDITION] [DETAILS]

Examples:

vintage red Ferrari 250 GTO, polished chrome, gleaming in sunlight
ancient leather-bound book, worn pages, golden embossed cover
steaming cup of coffee, ceramic white mug, wisps of vapor rising

3. Action (Required)

What is happening.

Character Actions:

Movement:

walking slowly / running fast / sprinting
dancing gracefully / spinning around
turning head left / looking over shoulder
reaching out hand / waving
jumping / falling / climbing
sitting down / standing up

Gestures:

smiling / laughing / crying
nodding / shaking head
pointing / gesturing
covering face / rubbing eyes
crossing arms / hands in pockets

Interactions:

talking to camera / speaking to friend
holding object / picking up item
opening door / closing window
drinking coffee / eating food
typing on laptop / writing on paper

Camera-Relative Actions:

walking towards camera / walking away from camera
turning to face camera / looking directly at camera
moving from left to right across frame
approaching from background

4. Environment (Required)

Where action takes place.

Indoor Locations:

modern minimalist apartment, white walls, large windows
cozy coffee shop, wooden tables, warm lighting
industrial warehouse, exposed brick, metal beams
luxury hotel lobby, marble floors, crystal chandeliers
dark nightclub, neon lights, smoke machines
art gallery, white walls, spotlit paintings
recording studio, soundproof walls, equipment racks

Outdoor Locations:

busy Tokyo street at night, neon signs, rain-slicked pavement
California beach at sunset, golden sand, crashing waves
New York City rooftop, skyscrapers background, twilight sky
misty forest, towering pines, dappled sunlight
desert landscape, sand dunes, clear blue sky
mountain peak, snow-covered, dramatic clouds
urban alley, graffiti walls, fire escape stairs

Environmental Details:

Weather: raining, snowing, foggy, sunny, overcast
Time: dawn, morning, midday, afternoon, dusk, night, midnight
Season: spring blossoms, summer heat, autumn leaves, winter snow
Crowd: empty, few people, crowded, bustling

5. Camera Movement (Critical for AI Video)

Static Shots:

camera locked, no movement - Maximum stability
static shot - Simple, clean framing
fixed frame - No camera motion at all

Basic Movements:

Pan (Horizontal Rotation):

slow pan right - Reveal right side of scene
quick pan left - Fast reveal, energetic
continuous pan - Sweeping motion
whip pan - Very fast blur transition

Tilt (Vertical Rotation):

slow tilt up - Reveal height, grandeur
tilt down - Ground to subject
tilt up to sky - Opening up, freedom

Zoom:

slow zoom in - Draw attention, build tension
fast zoom out - Reveal context, surprise
dolly zoom (Vertigo effect) - Background stretches while subject stays same size

Dolly/Track (Camera Position Moves):

dolly in / push in - Move camera forward toward subject
dolly out / pull back - Move camera backward away from subject
tracking shot following subject - Camera moves alongside moving subject
truck left / truck right - Camera slides sideways parallel to subject

Advanced Movements:

orbit around subject / 360° rotation - Circle around subject
crane up / crane down - Smooth vertical elevation change
pedestal up / pedestal down - Camera rises/lowers vertically
handheld camera, slight shake - Documentary, realistic feel
Steadicam smooth glide - Floating smooth movement
FPV drone racing through - Fast, dynamic, first-person drone

Combined Movements:

dolly in while panning right - Complex reveal
crane up and pull back simultaneously - Epic opening
orbit while slowly zooming in - Dramatic focus
tracking shot with slight handheld shake - Realistic following

6. Lighting (Defines Mood & Quality)

Natural Light:

golden hour sunlight - Warm, soft, magical (1 hour after sunrise/before sunset)
blue hour twilight - Cool, ethereal (just before sunrise/after sunset)
harsh midday sun - Strong shadows, high contrast
overcast soft light - Even, flattering, no hard shadows
dappled sunlight through trees - Patchy, organic patterns
window light from left - Directional, soft, Rembrandt-style

Artificial Light:

studio lighting, 3-point setup - Professional, controlled
neon lighting - Cyberpunk, colorful, urban
warm tungsten interior - Cozy, nostalgic, orange tones
cool fluorescent - Clinical, modern, blue-green tones
candlelight - Intimate, flickering, warm
street lamps at night - Urban, pools of light in darkness
LED RGB color wash - Modern, customizable, vibrant

Lighting Qualities:

soft diffused light - Flattering, minimal shadows
hard directional light - Dramatic shadows, contrast
backlight / rim light - Subject outlined in light
side lighting - Texture emphasis, depth
silhouette lighting - Subject dark against bright background
volumetric lighting / god rays - Light beams visible in atmosphere
chiaroscuro lighting - Extreme light/dark contrast (Renaissance style)

Lighting Effects:

lens flare - Sun or light source in frame creates flare
light leaks - Vintage film aesthetic
bokeh in background - Out-of-focus light orbs
practical lights in scene - Visible lamps, candles, screens

7. Style Modifiers (Aesthetic Direction)

Cinematic Styles:

cinematic, film grain - Movie quality, organic texture
35mm film aesthetic - Classic cinema look
anamorphic lens, 2.39:1 aspect ratio - Wide cinematic, horizontal flares
IMAX quality - Maximum detail, epic scale
documentary style - Realistic, observational
music video aesthetic - Stylized, fast-paced, bold

Genre Styles:

film noir, high contrast black and white - Classic detective, shadows
horror atmosphere, desaturated colors - Eerie, unsettling
sci-fi futuristic - Clean, high-tech, blue tones
fantasy epic - Grand, saturated colors, sweeping
romantic comedy, bright and warm - Cheerful, inviting
thriller, moody and tense - Dark, suspenseful

Art Styles:

anime style - Japanese animation aesthetic
Studio Ghibli aesthetic - Whimsical, detailed, nostalgic
Pixar 3D animation - Polished, colorful, expressive
claymation stop-motion - Tactile, handmade feel
watercolor painting come to life - Soft, artistic
oil painting aesthetic - Textured, classical art
cyberpunk neon - Dark, neon colors, futuristic grit

Photography Styles:

National Geographic documentary - Natural, educational, beautiful
fashion photography, high contrast - Bold, stylized, editorial
street photography, candid - Authentic, moment-capture
architectural photography, symmetrical - Clean lines, geometric
portrait photography, shallow depth of field - Subject sharp, background blur

Era/Period Styles:

1920s silent film, sepia tone - Vintage, nostalgic
1980s VHS aesthetic, scan lines - Retro, lo-fi
1990s home video - Nostalgic, family memories
modern Instagram aesthetic - Clean, filtered, social media
futuristic 2050s - Advanced, sleek, imaginative

8. Mood & Atmosphere (Emotional Tone)

Positive Moods:

joyful and uplifting - Happy, celebratory
peaceful and serene - Calm, meditative
romantic and intimate - Loving, tender
energetic and dynamic - Exciting, active
whimsical and playful - Fun, light-hearted
inspiring and motivational - Empowering, hopeful

Negative Moods:

tense and suspenseful - Anxious, anticipatory
melancholic and sad - Sorrowful, reflective
ominous and foreboding - Threatening, dark
chaotic and frantic - Overwhelming, intense
lonely and isolated - Solitary, empty
mysterious and enigmatic - Unknown, intriguing

Neutral/Atmospheric:

contemplative and thoughtful - Reflective, introspective
nostalgic and bittersweet - Memories, longing
dreamlike and surreal - Unreal, ethereal
gritty and raw - Realistic, unpolished
epic and grand - Large-scale, impressive
intimate and personal - Close, private

9. Technical Specs (Optional but Powerful)

Focus & Depth:

shallow depth of field, f/1.4 - Subject sharp, background very blurred
deep depth of field, f/11 - Everything in focus
rack focus from foreground to background - Shift focus during shot
bokeh background - Creamy, soft out-of-focus areas
tilt-shift miniature effect - Toy-like, selective focus band

Frame Rate & Motion:

24fps cinematic - Standard film, natural motion blur
60fps smooth motion - Clear, video-like, sports
120fps slow motion - 5x slowdown, dramatic
240fps extreme slow-mo - 10x slowdown, ultra-detailed
time-lapse speed - Compressed time, fast clouds/traffic
hyper-lapse movement - Moving time-lapse

Camera & Lens:

wide angle 16mm lens - Exaggerated perspective, spacious
standard 50mm lens - Natural human perspective
telephoto 200mm lens - Compressed space, isolated subject
fisheye lens, distorted - Extreme wide, curved edges
macro lens, extreme close-up - Tiny details magnified
vintage anamorphic lens - Horizontal flares, wide aspect

Quality & Processing:

8K resolution, ultra sharp - Maximum detail
film grain texture - Organic, cinematic
clean digital, no grain - Modern, pristine
high dynamic range (HDR) - Enhanced contrast, vibrant
motion blur - Realistic movement, cinematic
no motion blur, crystal clear - Every frame sharp (video game style)

Color Grading:

warm color grade, orange and teal - Blockbuster look
cool blue color grade - Night, tech, cold
desaturated, muted colors - Gritty, realistic
vibrant saturated colors - Energetic, bold
black and white - Timeless, dramatic
sepia tone - Vintage, nostalgic

🎥 Camera Language - Professional Cinematography

Shot Angles (Camera Height & Orientation)

Vertical Angles:

Eye Level (Most Common)
- Camera at subject's eye height
- Neutral, conversational, relatable
- Example: eye level shot of woman talking
High Angle (Looking Down)
- Camera above subject, tilted down
- Makes subject appear smaller, vulnerable, weak
- Example: high angle shot of child sitting alone
Low Angle (Looking Up)
- Camera below subject, tilted up
- Makes subject appear powerful, dominant, heroic
- Example: low angle shot of businessman in suit, looking powerful
Bird's Eye View (Directly Overhead)
- 90° straight down
- God's perspective, abstract, mapping
- Example: bird's eye view of city intersection, cars moving
Worm's Eye View (Ground Level Up)
- Camera on ground looking up
- Extreme power to subject, towering
- Example: worm's eye view of skyscraper reaching into clouds

Horizontal Angles:

Front Angle (Face Camera)
- Subject facing camera directly
- Direct address, confrontational, intimate
- Example: front angle of actor speaking to camera
Profile / Side Angle (90° Perpendicular)
- Subject in pure profile
- Observational, subject unaware of camera
- Example: side angle profile of woman looking out window
3/4 Angle (45° Offset)
- Most common, natural conversation angle
- Dimensional, flattering, engaging
- Example: 3/4 angle of man at desk working
Over-The-Shoulder (OTS)
- Camera behind one subject, looking at another
- Conversation, POV context, relationship
- Example: over-the-shoulder shot during conversation

Specialty Angles:

Dutch Angle / Canted Angle
- Camera tilted on axis (horizon not level)
- Unease, tension, psychological disorientation
- Example: Dutch angle 15° tilt of villain entering room, unsettling
POV (Point of View)
- Camera becomes character's eyes
- Subjective, immersive, first-person
- Example: POV shot of hands opening door, entering room
Reverse Angle
- 180° opposite of previous shot
- Show other side of conversation/action
- Example: reverse angle revealing who character is talking to

Camera Movements - Deep Dive

1. Pan (Horizontal Rotation)

Description: Camera rotates left or right on horizontal axis while staying in same position.

Use Cases:

Following moving subject horizontally
Revealing wide landscape or environment
Connecting two subjects in conversation
Showing cause and effect (look left, pan right to reveal what they see)

Speeds:

slow pan - Contemplative, revealing (5-10 seconds)
medium pan - Natural observation (2-4 seconds)
fast pan / whip pan - Energetic transition, surprise (0.5-1 second)

Examples:

slow pan right revealing New York skyline at sunset
quick whip pan left from woman's face to man entering room
continuous 180° pan across crowded marketplace

2. Tilt (Vertical Rotation)

Description: Camera rotates up or down on vertical axis while staying in same position.

Use Cases:

Revealing height or depth
Character's upward/downward gaze following
Establishing scale (building, mountain)
Emotional rise or fall

Directions:

tilt up - Reveal height, grandeur, hope, freedom
tilt down - Ground, reveal bottom, falling, despair

Examples:

slow tilt up from shoes to face, revealing character
tilt down from skyscraper top to street level
tilt up following bird flying into sky

3. Dolly / Track (Camera Position Moves)

Description: Entire camera moves forward, backward, or sideways on track/dolly.

Dolly In / Push In:

Camera moves toward subject
Intensifies focus, builds tension, draws viewer in
More intimate as it progresses

slow dolly in on woman's face as she realizes truth
fast dolly in on ticking clock, building suspense

Dolly Out / Pull Back:

Camera moves away from subject
Reveals context, isolation, bigger picture
Can create surprise by revealing unexpected environment

dolly out from close-up revealing character is in cage
slow pull back from couple kissing to reveal they're at airport

Truck / Lateral Dolly:

Camera slides left or right parallel to subject
Following action, revealing environment alongside
Maintains distance while showing movement

truck right following runner along beach
lateral dolly left past row of soldiers standing at attention

4. Zoom (Lens Focal Length Change)

Description: Lens magnification changes, NOT camera position. Background compresses/expands.

Zoom In:

Magnify subject, narrow field of view
Artificial feeling (lens, not camera moving)
Direct attention, isolate subject

slow zoom in on suspicious object on table
fast crash zoom on character's shocked face

Zoom Out:

Widen field of view, reveal more
Show context, de-emphasize subject

zoom out revealing character is surrounded

Dolly Zoom (Vertigo Effect / Zolly):

Dolly in while zooming out (or vice versa)
Subject stays same size but background stretches/compresses
Disorientation, realization, psychological shift
Made famous in Hitchcock's "Vertigo" and Spielberg's "Jaws"

dolly zoom effect, dolly in while zooming out, vertigo feeling, subject stays centered but background warps

5. Orbit / Arc (Circular Around Subject)

Description: Camera moves in circular path around subject.

Use Cases:

Showcasing product from all angles
Dramatic character reveal
Time passage or transformation
"Matrix" bullet-time style

Variations:

slow 360° orbit - Complete circle, full reveal
partial arc clockwise - Smooth angle change
tight orbit close to subject - Intimate, detail
wide orbit with subject centered - Epic, showcase

Examples:

slow 360° orbit around sports car, showcasing design
clockwise orbit around couple dancing in spotlight
orbital shot rising while circling monument

6. Crane / Jib (Vertical Sweeping)

Description: Camera on arm that can move up, down, and sweep.

Use Cases:

Establishing shots (start high, descend into scene)
Epic reveals
Following vertical action (climbing, falling)
Transitioning between levels

Movements:

crane up - Rise majestically, opening up world
crane down - Descend into action, focus inward
sweeping crane - Combine vertical with horizontal motion

Examples:

crane up from ground level revealing vast battlefield
sweeping crane shot starting low, rising over crowd to stage
crane down from building top to street level entrance

7. Pedestal (Camera Raises/Lowers Vertically)

Description: Camera moves straight up or down without tilting.

Difference from Crane: Pedestal is pure vertical (elevator), Crane is sweeping arc.

Examples:

pedestal up following character standing from chair
pedestal down as character sits

8. Handheld (Camera Operator Holds Camera)

Description: Camera held by operator, natural human shake and movement.

Use Cases:

Documentary realism
Action intensity, chaos
Immersive, "you are there" feeling
Low-budget or guerrilla style

Qualities:

slight handheld shake - Subtle realism
intense handheld, shaky - Action, urgency, panic
smooth handheld - Practiced operator, gentle

Examples:

handheld camera following character running through crowd, chaotic
documentary style handheld, slight natural shake

9. Steadicam (Stabilized Handheld)

Description: Camera on stabilizing rig worn by operator, smooth gliding motion.

Use Cases:

Following characters through spaces
Long continuous takes
Smooth but dynamic motion
Tracking shots without tracks

Famous Examples:

"The Shining" hallway shots
"Goodfellas" Copacabana shot

Examples:

Steadicam smooth glide following character walking through mansion
floating Steadicam shot through crowded party, weaving between people

10. Drone / Aerial (Flying Camera)

Description: Camera mounted on drone, unrestricted 3D movement.

Use Cases:

Sweeping landscapes
Impossible angles
FPV (first-person view) racing through spaces
Establishing shots, reveals

Movements:

slow aerial rise - Epic reveal, opening up
FPV drone racing through - Fast, dynamic, immersive
sweeping aerial over landscape - Beauty, scale
orbital drone around building - 360° architectural

Examples:

slow aerial drone rise revealing coastline at sunrise
FPV drone racing through forest, dodging trees, fast and dynamic
sweeping aerial shot over mountain range, epic scale
drone descending from clouds to street level city

Combining Camera Movements

Advanced directors combine movements for complex shots:

dolly in while panning right, following character entering room and turning
crane up while pulling back simultaneously, revealing character's isolation in vast space
orbit around subject while slowly zooming in, building dramatic focus
tracking shot with slight handheld shake, following character running, intense realism
tilt up while dollying forward, emphasizing building height as approaching

Pro Tip: Start with single movement, master it, then combine for complexity.

🏃 Motion Control - Directing Movement

Subject Motion Speed

Human Movement:

standing still / stationary
slow walk - Contemplative, casual
normal walk - Natural pace
fast walk / power walk - Purpose, urgency
jogging - Exercise, moderate speed
running - Urgency, escape, chase
sprinting at full speed - Maximum urgency, action
slow motion walk, dramatic - Heroic, emphasis

Object Motion:

drifting slowly - Gentle, floating
smooth continuous motion - Steady, mechanical
stuttering motion - Stop-motion, glitchy
rapid movement - Quick, energetic
explosive motion - Sudden, dramatic

Motion Blur & Shutter

Motion Blur (Natural):

natural motion blur, 24fps - Cinematic, realistic
180° shutter angle - Standard film motion blur
heavy motion blur, fast action - Emphasize speed

No Motion Blur:

no motion blur, crystal clear - Video game, unnatural
fast shutter, frozen motion - "Saving Private Ryan" battle scenes
360° shutter, maximum blur - Dreamy, smooth

Speed Ramping

Time Manipulation:

real-time normal speed
slow motion, 2x slower - 60fps played at 24fps
slow motion, 5x slower - 120fps → 24fps, dramatic action
extreme slow motion, 10x slower - 240fps → 24fps, cinematic beauty
time-lapse, 10x faster - Clouds, traffic, day to night
hyper-lapse, moving time-lapse - Camera moves during time-lapse

Speed Ramping (Variable Speed in Single Shot):

starts in slow motion, ramps to real-time, then back to slow motion
begins real-time, suddenly shifts to 5x slow motion for emphasis

Parallax & Depth

Creating 3D Depth:

foreground elements moving faster than background - Parallax effect
subject moving through depth, toward and away from camera - Z-axis motion
layers of motion at different speeds - Depth cues

Examples:

woman walking toward camera, foreground flowers blowing fast, background mountains still, strong parallax depth

👤 Character Animation - Directing Actors

Facial Expressions

Basic Emotions:

smiling warmly / broad genuine smile
subtle smirk / mischievous grin
laughing heartily / giggling
crying tears / wiping away tears
frowning sadly / lips trembling
angry scowl / gritted teeth
shocked, eyes wide, mouth open
confused, furrowed brow
fearful, trembling
disgusted, nose wrinkled

Subtle Expressions:

eyes narrowing suspiciously
raising one eyebrow skeptically
biting lower lip nervously
pursed lips in concentration
soft exhale of relief
barely perceptible nod

Body Language

Posture:

standing confidently, shoulders back
slouching, defeated posture
leaning casually against wall
sitting upright, attentive
curled up, defensive
pacing back and forth anxiously

Gestures:

crossing arms defensively
hands in pockets, casual
hands on hips, assertive
wringing hands nervously
pointing accusingly
beckoning with finger
waving goodbye
thumbs up gesture
covering face with hands
running fingers through hair
touching neck nervously

Walking Styles:

confident stride - Head high, purposeful
nervous shuffle - Small steps, hesitant
casual stroll - Relaxed, unhurried
power walk - Fast, determined
tiptoeing quietly - Sneaking, careful
limping - Injured, struggling
stumbling - Drunk, disoriented

Character Interactions

Two-Person Interactions:

shaking hands formally
embracing warmly, hugging
high-five celebration
argument, pointing at each other
whispering secret in ear
dancing together, close
pushing away angrily
reaching out to comfort

Object Interactions:

picking up coffee cup, sipping
typing on laptop, focused
writing on paper with pen
answering phone, bringing to ear
opening door slowly
throwing object forcefully
catching object mid-air
reading book, turning pages

Gaze & Eye Direction

Eye Lines:

looking directly at camera - Direct address, breaking 4th wall
looking off-screen left - Watching something/someone not visible
gazing upward thoughtfully - Contemplation, remembering
staring down at ground - Shame, sadness, introspection
eyes following moving object - Tracking, curiosity
making eye contact with another person - Connection, conversation

Eye Movements:

blinking rapidly - Nervousness, processing
slow blink - Relaxation, agreement
widening eyes in realization - Sudden understanding
eyes darting around suspiciously - Paranoia, searching
rolling eyes - Annoyance, dismissal

Character Movement in Space

Relative to Camera:

walking toward camera - Approaching, confrontational
walking away from camera - Leaving, rejection, ending
walking left to right across frame - Lateral progression, journey
walking from background to foreground - Emergence, arrival
circling camera, maintaining eye contact - Predatory, suspicious

Relative to Environment:

entering through door into room
exiting through window
climbing stairs slowly
descending escalator
weaving through crowd
pushing through dense foliage

Performance Intensity

Subtle (Realistic):

subtle smile, slight upturn of lips
micro-expression of doubt flickering across face
barely noticeable tensing of shoulders

Moderate (Theatrical):

expressive gesturing while speaking
animated conversation, hands moving
visible emotional reactions

High (Dramatic):

over-the-top dramatic gasp, hand to chest
falling to knees in despair
arms spread wide, shouting
operatic levels of emotion

💡 Lighting & Atmosphere - Painting with Light

Natural Lighting Scenarios

Golden Hour (Magic Hour):

Time: 1 hour after sunrise / 1 hour before sunset
Quality: Warm, soft, directional, long shadows
Color Temp: 3000K-4000K (orange-amber)
Mood: Romantic, nostalgic, beautiful, cinematic
Best For: Portraits, landscapes, romance, beauty

Examples:

golden hour sunlight streaming through window, warm amber glow
sunset golden hour, subject backlit with rim of gold light

Blue Hour:

Time: 20-40 min before sunrise / after sunset
Quality: Cool, even, ethereal, soft
Color Temp: 12000K-20000K (deep blue)
Mood: Mysterious, dreamy, tranquil, surreal
Best For: Cityscapes, architecture, moody portraits

Examples:

blue hour twilight, city lights just turning on, deep blue sky
pre-dawn blue hour, silhouettes against gradient sky

Midday (Harsh Sun):

Time: 10am-2pm
Quality: Hard, direct, strong shadows, high contrast
Color Temp: 5500K (neutral)
Mood: Realistic, documentary, harsh, unforgiving
Best For: Vibrant colors, silhouettes, drama
Challenge: Unflattering for faces (top shadows under eyes/nose)

Examples:

harsh midday sun, deep black shadows, high contrast
noon sun directly overhead, minimal shadows

Overcast (Soft Diffused):

Quality: Soft, even, wraparound, no harsh shadows
Color Temp: 6500K (slightly cool)
Mood: Calm, melancholic, gentle, flattering
Best For: Portraits, even skin tones, subtle

Examples:

overcast soft light, no shadows, even illumination
cloudy day diffused light, gentle and flattering

Window Light:

Quality: Directional but soft (if diffused), Rembrandt-style
Best For: Indoor portraits, interviews, intimate scenes
Variations: Side window (dramatic), front window (flattering)

Examples:

soft window light from left, Rembrandt lighting on face
morning window light, gentle and warm

Artificial Lighting Setups

3-Point Lighting (Studio Standard):

Components:

Key Light - Main light, strongest, defines shape (45° angle)
Fill Light - Softens shadows from key (opposite side, less intense)
Back/Rim Light - Separates subject from background (behind subject)

studio lighting, 3-point setup, professional, clean

Rembrandt Lighting:

Key light 45° to side and above
Creates triangle of light on shadowed cheek under eye
Classic portrait style, dramatic but flattering
Named after painter Rembrandt

Rembrandt lighting, triangle of light on cheek, dramatic portrait

Butterfly Lighting (Paramount):

Key light directly in front and above subject
Creates butterfly-shaped shadow under nose
Glamorous, beauty style
Used in classic Hollywood portraits

butterfly lighting, glamorous beauty style, shadow under nose

Split Lighting:

Light from directly left or right
Half face in light, half in shadow
Dramatic, mysterious, edgy

split lighting, half face in shadow, dramatic contrast

Rim/Edge Lighting:

Light from behind subject
Outlines subject with bright edge
Separates from background, dramatic
Common in music videos, commercials

rim lighting from behind, subject outlined in light, dramatic separation

Practical Lights (In-Scene):

Visible light sources within frame
Lamps, candles, monitors, signs, windows
Adds realism and production value

practical lamp on desk illuminating scene, warm glow
candlelight as only light source, flickering shadows

Lighting Moods & Scenarios

Neon Lighting (Cyberpunk/Urban):

neon lighting, pink and blue neon signs, wet pavement reflections, night
Tokyo alley, red neon overhead, green neon storefront, atmospheric haze
cyberpunk city, multi-color neon, rain-slicked streets, purple and orange glow

Volumetric Lighting (God Rays):

volumetric lighting, dust particles visible in light beams, atmospheric
god rays streaming through forest canopy, misty atmosphere
light beams cutting through smoke in dark room

Chiaroscuro (High Contrast Light/Dark):

chiaroscuro lighting, extreme contrast, Renaissance painting style
deep blacks and bright highlights, dramatic shadows, film noir

Silhouette Lighting:

silhouette, subject backlit, bright background, subject in shadow
sunset silhouette, person against orange sky, no detail on subject

Candlelight:

candlelight illumination, warm orange glow, flickering shadows, intimate
single candle lighting face, dramatic shadows dancing

Firelight:

firelight, warm dancing orange light, flickering, cozy
campfire glow on faces, night scene, warm atmosphere

Moonlight:

moonlight, cool blue tone, soft silvery light, night scene
full moon illumination, low light, mysterious blue glow

Street Lamps:

street lamp pools of light, night, warm yellow circles in darkness
urban street lighting, sodium vapor orange glow

Color Temperature & Grading

Warm (Orange/Yellow) - 2000K-4000K:

Mood: Cozy, nostalgic, romantic, inviting
Sources: Candles, tungsten bulbs, golden hour, firelight
Use For: Intimate scenes, comfort, warmth

Neutral (White) - 5000K-6000K:

Mood: Natural, realistic, balanced
Sources: Midday sun, LED daylight
Use For: Documentary, natural look

Cool (Blue) - 6500K-20000K:

Mood: Clinical, technological, cold, eerie, twilight
Sources: Overcast sky, blue hour, fluorescent, moonlight
Use For: Sci-fi, tension, isolation

Color Grading Styles:

orange and teal color grade, blockbuster cinematic look
desaturated muted colors, gritty realistic
vibrant saturated colors, music video energetic
monochromatic blue, cold and technological
warm sepia tone, nostalgic vintage

🎨 Scene Design - Building Worlds

Environment Categories

Urban Settings:

Modern City:

modern city, glass skyscrapers, clean streets, minimalist architecture
bustling downtown, LED billboards, crowds, contemporary urban
high-tech smart city, holographic displays, futuristic clean

Cyberpunk City:

cyberpunk megacity, neon signs, rain-soaked streets, steam vents, crowded and dense
dystopian urban, towering megastructures, smog, multi-level highways, dark and oppressive
blade runner aesthetic, flying cars in background, Asian-fusion architecture, perpetual night

Historic City:

medieval European city, cobblestone streets, stone buildings, lanterns
ancient Rome, marble columns, forum, togas, classical architecture
1920s New York, art deco, vintage cars, period costume

Urban Decay:

abandoned city, overgrown vegetation, broken windows, post-apocalyptic
industrial wasteland, rusted machinery, graffiti, gritty and worn
slum district, makeshift structures, crowded, poverty, realistic

Natural Environments:

Forest:

dense forest, towering pine trees, dappled sunlight through canopy, moss-covered ground
misty forest, fog rolling between trees, mysterious atmosphere, limited visibility
autumn forest, orange and red leaves, carpet of fallen foliage, warm tones
enchanted forest, glowing mushrooms, magical atmosphere, ethereal light

Beach/Ocean:

tropical beach, white sand, crystal clear turquoise water, palm trees, sunny
rocky coastline, crashing waves, dramatic cliffs, stormy sky, moody
sunset beach, golden sand, orange sky reflected in wet sand, romantic
underwater scene, coral reef, colorful fish, rays of light from surface

Mountains:

snow-capped mountain peak, dramatic rocky terrain, blue sky, epic scale
mountain valley, green meadows, wildflowers, distant peaks, peaceful
mountain cliff edge, vertiginous drop, clouds below, dangerous

Desert:

vast desert, rolling sand dunes, clear blue sky, harsh sun, minimalist
desert at sunset, purple and orange sky, long shadows from dunes, dramatic
desert oasis, palm trees around water, contrast with surrounding barren land

Arctic/Tundra:

arctic landscape, glaciers, icebergs, frozen sea, cold blue tones
snowy tundra, endless white, low clouds, desolate and isolated
northern lights, aurora borealis over snowy landscape, magical colors

Interior Spaces:

Residential:

modern minimalist apartment, white walls, large windows, natural light, clean
cozy cottage interior, fireplace, warm wood, comfortable furniture, inviting
luxury penthouse, marble floors, floor-to-ceiling windows, city view, opulent
cluttered bedroom, personal items, lived-in feel, realistic mess, intimate

Commercial:

trendy coffee shop, exposed brick, vintage furniture, warm lighting, hipster
upscale restaurant, dim mood lighting, white tablecloths, elegant
busy office, cubicles, fluorescent lighting, computers, corporate
abandoned warehouse, concrete floors, broken windows, graffiti, industrial

Public Spaces:

grand museum, marble floors, high ceilings, artwork on walls, echoing space
busy airport terminal, people walking, departure boards, modern architecture
ornate theater, red velvet seats, gold trim, chandelier, classical
subway station, tiled walls, platform, waiting passengers, urban transit

Special Venues:

nightclub interior, strobing lights, smoke machines, crowded dance floor, energetic
art gallery, white walls, spotlit paintings, minimalist, quiet
recording studio, soundproofed walls, equipment racks, mixing console, professional
gym, exercise equipment, mirrors, bright lighting, energetic

Atmospheric Elements

Weather:

clear sunny day, blue sky - Happy, optimistic
overcast clouds, grey sky - Melancholic, neutral
raining, wet pavement, puddles - Moody, introspective, cinematic
heavy rain, downpour, water streaming - Intense, dramatic
light drizzle, mist - Gentle, romantic
thunderstorm, lightning flashes, dark clouds - Ominous, powerful
snowing, falling snowflakes, winter - Peaceful, cold, beautiful
blizzard, heavy snow, low visibility - Harsh, survival
fog, misty, limited visibility - Mysterious, eerie
sandstorm, dust clouds, desert - Harsh, hostile

Time of Day:

dawn, first light, soft pastels - New beginning, hope
morning, bright and fresh - Energy, start
midday, bright sun overhead - Peak activity
afternoon, warm light - Winding down
sunset, golden and orange sky - Ending, beautiful
dusk, fading light, deep colors - Transition, contemplative
night, dark, artificial lights - Mystery, different world
midnight, deep darkness - Isolation, stillness

Atmospheric Effects:

dust particles visible in light - Texture, atmosphere
smoke/fog rolling along ground - Mysterious, dramatic
heat shimmer, heat waves rising - Hot, desert, summer
steam vents in urban scene - Industrial, city atmosphere
lens flare from sun/bright light - Natural, cinematic
light beams cutting through darkness - Dramatic, theatrical
falling leaves - Autumn, change, melancholic
floating cherry blossoms - Spring, Japanese aesthetic, gentle
embers/sparks floating - Fire, danger, energy

🎬 Platform-Specific Guides

Runway Gen-3 Alpha

Overview:

Text-to-Video: 5-10 second clips
Image-to-Video: Animate still images
1280×768 or 768×1280 resolution
Best for: Realistic, cinematic output

Prompt Structure (Runway-Optimized):

[CAMERA MOVEMENT]: [SHOT TYPE] of [DETAILED SUBJECT] [ACTION],
[ENVIRONMENT DETAILS], [LIGHTING], [STYLE], [MOOD]

Key Tips:

Camera movement FIRST - Runway prioritizes camera control
Be specific about subject - Detailed descriptions work better
Lighting is crucial - Specify clearly for best results
Keep actions simple - Complex multi-step actions may fail
Use reference images - Image-to-Video mode for consistency

Runway-Specific Parameters:

Motion Brush - Paint areas for selective movement
Camera Control - Pan, tilt, zoom, roll (in interface)
Seed - Consistency across generations
Interpolate - Smooth transitions between images

Example Runway Prompts:

Slow dolly forward: Close-up of woman's face, eyes opening slowly, morning light through window, soft focus, intimate and peaceful

FPV drone shot: Racing through neon-lit Tokyo alley at night, fast movement, rain-slicked ground, cyberpunk aesthetic, dynamic and energetic

Static camera: Medium shot of coffee cup on wooden table, steam rising, sunlight from left, shallow depth of field, cozy cafe atmosphere

Common Issues & Fixes:

Morphing/warping: Simplify action, reduce movement complexity
Inconsistent subject: Use Image-to-Video with reference image
Jittery motion: Specify "smooth" camera movement
Wrong lighting: Be very specific about light source and direction

Pika Labs 1.0

Overview:

Text-to-Video & Image-to-Video
3-second default (can extend)
1280×720 resolution
Unique "Modify Region" feature

Pika-Specific Features:

Modify Region:

Select area of image to animate
Keep rest static
Great for subtle animations (eyes blinking, hair blowing)

Parameters:

-camera - Camera motion control
-motion - Subject motion intensity (1-4)
-fps - Frame rate (8, 16, 24)
-gs - Guidance scale (8-24, higher = follow prompt more strictly)
-neg - Negative prompt (what to avoid)

Example Pika Prompts:

Portrait of woman, hair blowing gently in wind, eyes blinking, subtle smile
-camera zoom in -motion 2 -fps 24 -neg "morphing, distortion"

City street, cars driving past, pedestrians walking
-camera pan right -motion 3 -fps 24

Coffee being poured into cup, steam rising
-camera static -motion 2 -fps 24 -neg "liquid morphing"

Pika Best Practices:

Use negative prompts - Avoid common issues like morphing
Motion parameter - Start at 2, adjust if too still (1) or chaotic (4)
Extend feature - Generate 3s, extend for longer
Modify Region - Best for subtle, controlled motion

Kling AI (Kuaishou)

Overview:

Chinese platform, high quality
5-second clips
720p or 1080p
Strong at realistic human motion

Kling Strengths:

Natural character movement
Good facial expressions
Realistic physics
High-resolution output

Prompt Tips for Kling:

Simple, clear language - Translation works better with clarity
Focus on character action - Kling excels at people
Specify resolution - 1080p for best quality
Natural movements - Walking, talking, gesturing work great

Example Kling Prompts:

Young woman walking through park, smiling, looking at camera, natural movement, sunny day, 1080p

Man sitting at desk, typing on laptop, looking focused, office environment, daytime, realistic

Couple holding hands walking on beach at sunset, romantic, slow motion, cinematic

Luma Dream Machine

Overview:

Fast generation (~120 seconds)
5-second clips
Good at natural motion
Simple interface

Luma Characteristics:

Speed - Fastest generation among major platforms
Natural physics - Realistic motion and gravity
Simplicity - No complex parameters
Dreamlike quality - Slightly surreal, artistic

Prompt Approach:

[ACTION] in [ENVIRONMENT], [STYLE], [MOOD]

Examples:

Butterfly landing on flower, garden setting, macro photography, peaceful

Skateboader doing kickflip, urban skatepark, slow motion, energetic

Waterfall in jungle, lush vegetation, misty, serene and beautiful

Best Use Cases:

Quick iterations/testing
Natural outdoor scenes
Animal/nature subjects
Abstract/dreamlike content

Stable Video Diffusion (Open Source)

Overview:

Image-to-Video only
14 or 25 frames
576×1024 resolution
Local/cloud deployment

Technical Parameters:

motion_bucket_id (1-255) - Motion intensity
cond_aug (0.0-1.0) - How much to vary from input image
fps - Output frame rate
seed - Reproducibility

Advantages:

Open source - Free, customizable
Local control - Run on your hardware
No censorship - No content restrictions
Fine-tunable - Train on custom data

Example Code:

from diffusers import StableVideoDiffusionPipeline

pipe = StableVideoDiffusionPipeline.from_pretrained(
    "stabilityai/stable-video-diffusion-img2vid-xt"
)

frames = pipe(
    image=input_image,
    motion_bucket_id=127,  # Medium motion
    fps=7,
    decode_chunk_size=8
).frames

Best For:

Developers wanting control
Custom fine-tuning
Batch processing
Research/experimentation

🎬 Production Workflows

Text-to-Video Workflow

Step 1: Concept & Planning

Define video goal (product showcase, storytelling, abstract art)
Write shot list with descriptions
Identify style references

Step 2: Craft Initial Prompt

[Shot Type] of [Subject] [Action] in [Environment],
[Camera Movement], [Lighting], [Style], [Mood]

Example:

Medium shot of barista making latte art, pouring milk into cup, cozy coffee shop with warm lighting,
slow dolly in on cup, soft window light from left, cinematic style, peaceful morning atmosphere

Step 3: Generate & Iterate

Generate first version
Analyze what worked / what didn't
Adjust prompt:
- Subject not clear? → More detailed subject description
- Wrong camera move? → Specify camera movement more clearly
- Bad lighting? → Add specific lighting terms
- Morphing issues? → Simplify action, add negative prompts

Step 4: Refine

Generate 3-5 variations with tweaked prompts
Select best result
Note successful prompt patterns for future

Step 5: Extend/Enhance (Platform-Dependent)

Runway: Interpolate between clips
Pika: Extend duration
Manual: Use video editing to combine clips

Image-to-Video Workflow

Step 1: Create/Select Source Image

Option A: Generate Image (Midjourney, DALL-E, Stable Diffusion)

Craft detailed image generation prompt
Important: Character/subject should be centered and clear
Avoid extreme compositions (works better with balanced framing)

Option B: Use Photo/Artwork

High resolution preferred (1024px+ on longest side)
Clear subject
Good lighting in image
Avoid motion blur in source (unless desired)

Step 2: Craft Motion Prompt

[Subject Movement], [Camera Movement], [Motion Speed], [Style Consistency]

Examples:

Subject turns head left and smiles, camera slowly zooms in, smooth motion, maintain photographic style

Camera orbits 90° clockwise around subject, subject remains still, slow smooth motion, keep composition

Subject's hair and clothes blowing in wind, camera static, gentle motion, maintain lighting

Step 3: Platform Selection

Best Platforms for Image-to-Video:

Runway Gen-3 - Most control, highest quality
Pika Labs - Modify Region for selective animation
Stable Video Diffusion - Local control, free

Step 4: Generate with Motion Controls

Runway Gen-3:

Upload image
Write motion prompt
Use Motion Brush to paint areas for movement
Adjust Camera Control sliders

Pika:

Upload image
Add prompt describing motion
Use -motion parameter (1-4)
Modify Region for selective animation

Step 5: Iterate

Too much motion? Lower motion parameter or simplify prompt
Too static? Increase motion or be more specific about movement
Morphing? Reduce complexity, add negative prompts
Wrong direction? Specify direction clearly (left, right, toward camera, etc.)

Multi-Shot Sequence Workflow

Creating Longer Narratives with Multiple Clips

Step 1: Storyboard Write shot list:

Shot 1: Wide establishing shot - City skyline at sunset
Shot 2: Medium shot - Character walking down street
Shot 3: Close-up - Character's face looking up
Shot 4: POV shot - What character sees (sign/building)
Shot 5: Medium shot - Character smiles and enters building

Step 2: Generate Each Shot Individually

Craft specific prompt for each shot
Generate all shots
Select best take for each

Step 3: Consistency Techniques

Visual Consistency:

Use same style keywords across all prompts
Match lighting descriptions
Maintain color palette (e.g., "warm orange and teal grading" in all)

Character Consistency (Challenging with Current AI):

Use Image-to-Video with same reference image
Describe character identically in each prompt
Use Runway's Gen-3 with character reference image
Consider generating character reference first, then I2V for each shot

Step 4: Post-Production

Import clips to video editor (Premiere, DaVinci Resolve, CapCut)
Arrange in sequence
Add transitions (cuts, fades, etc.)
Color grade for consistency
Add sound design
Export final video

Product Showcase Workflow

Goal: Highlight product from multiple angles with professional look

Shot 1: Hero Shot (Establishing)

Wide shot of [PRODUCT] on [SURFACE], [ENVIRONMENT],
slow 360° orbit camera, studio lighting with soft shadows,
product photography style, clean and professional

Shot 2: Detail Shot

Extreme close-up of [PRODUCT DETAIL - logo, texture, feature],
slow push in, macro lens, shallow depth of field,
dramatic lighting, premium feel

Shot 3: Context/Lifestyle Shot

Medium shot of person using [PRODUCT], [ENVIRONMENT],
natural movement, soft window lighting,
lifestyle photography, relatable and warm

Shot 4: Dynamic Shot

FPV camera racing around [PRODUCT], fast movement,
dramatic lighting, high-energy music video style,
exciting and modern

Combine into 30-second product video

🚀 Advanced Techniques

Consistency Across Generations

Challenge: AI video models generate different results each time.

Techniques:

1. Seed Control (Platform-Dependent)

Runway: Save seed from good generation, reuse
Stable Video Diffusion: Specify seed in code
Locks random variation, same prompt = same result

2. Reference Image Workflow

Generate perfect frame as image (Midjourney/DALL-E)
Use Image-to-Video mode with that reference
Character/style stays consistent

3. Detailed Consistent Descriptions

// Define character once, reuse exactly:
CHARACTER: "25-year-old Asian woman, long straight black hair,
red leather jacket, white t-shirt, blue jeans, brown eyes"

// Use in every prompt:
Close-up of [CHARACTER] smiling, ...
Medium shot of [CHARACTER] walking, ...
Wide shot of [CHARACTER] standing, ...

4. Style Lock Techniques Include exact same style keywords in every prompt:

..., cinematic style, 35mm film aesthetic, warm color grade,
film grain texture, shot on Arri Alexa

Controlling Difficult Concepts

Challenge Areas:

1. Text/Logos in Video

Current AI limitation: Text often warps or is illegible
Workaround: Generate clean shot, add text in post-production
Alternative: Use Image-to-Video with text baked into source image, minimal motion

2. Specific Brand/Product Details

Issue: AI may not render exact product accurately
Solution: Use Image-to-Video with actual product photo
Technique: Photograph product, animate with subtle motion (rotation, camera move)

3. Complex Multi-Character Interactions

Issue: AI struggles with precise choreography between characters
Solution: Keep interactions simple (hugging, high-five) vs complex (fight choreography)
Alternative: Generate characters separately, composite in post

4. Lip Sync to Specific Audio

Current limitation: Most text-to-video doesn't sync lips to speech
Workarounds:
- HeyGen/D-ID for talking head videos
- Generate video without speaking, add voiceover
- Use shots where mouth isn't visible (over-shoulder, distance, etc.)

Hybrid Workflows

Combining AI Video with Traditional Tools

Workflow 1: AI + VFX

Generate base video with AI
Export to After Effects / Nuke
Add VFX elements (particles, effects, corrections)
Color grade and finalize

Workflow 2: AI + 3D

Create 3D environment (Blender, Unreal Engine)
Generate AI character/subject
Composite AI subject into 3D environment
Render final

Workflow 3: AI + Live Action

Shoot live action plate
Generate AI elements (backgrounds, effects, creatures)
Composite AI into live footage
Blend with color matching

Workflow 4: Multi-Platform AI

Generate character in Midjourney (consistency)
Animate with Runway Image-to-Video
Enhance with Topaz Video AI (upscale, denoise)
Final edit in Premiere/DaVinci

Upscaling & Enhancement

Challenge: Most AI video is limited resolution (720p-1080p)

Upscaling Tools:

1. Topaz Video AI

AI upscaling to 4K/8K
Denoising and stabilization
Frame interpolation (increase fps)

2. DaVinci Resolve Super Scale

Built into Resolve Studio
AI upscaling
Free with Studio license

3. AVCLabs Video Enhancer

AI enhancement
Colorization
Face refinement

Workflow:

AI Video (720p) → Topaz Video AI (upscale to 4K) →
Color Grade → Denoise → Export 4K

Speed Ramping & Time Manipulation

Creating Variable Speed in Single Clip

In Post-Production:

Generate clip at normal speed
Import to editing software with optical flow
Apply speed ramps:
- Start slow motion (50%)
- Ramp to normal (100%)
- Ramp to fast motion (200%)

Software:

DaVinci Resolve: Retime Curve
Premiere Pro: Time Remapping
Final Cut Pro: Retiming
After Effects: Time Remapping with frame blending

Creative Uses:

Action emphasis (slow-mo at impact)
Emotional beats (slow-mo at realization)
Energy bursts (normal to fast)

❌ Troubleshooting Common Issues

Issue: Morphing / Warping

Symptoms:

Subject's face/body distorts unnaturally
Objects melt or transform mid-clip
Structural inconsistencies

Causes:

Action too complex
Too much motion
Ambiguous subject description

Fixes:

Simplify action - "walking" instead of "dancing backflip"
Reduce motion complexity - Single action per clip
Be more specific - Detailed subject description locks appearance
Use Image-to-Video - Provide clear starting frame
Negative prompts (Pika) - -neg "morphing, distortion, warping"
Shorter duration - 3s instead of 10s

Issue: Jittery / Unstable Motion

Symptoms:

Camera shakes unexpectedly
Subject movement stutters
Frame-to-frame inconsistency

Causes:

Conflicting motion instructions
AI struggling with motion coherence

Fixes:

Specify "smooth" - "smooth dolly in" vs just "dolly in"
Reduce motion complexity - Single camera move vs combined
Lower motion parameter (Pika: -motion 1 or 2)
Post-stabilization - Use warp stabilizer in editing software
Specify stabilization - "Steadicam smooth" or "locked on tripod"

Issue: Wrong Lighting

Symptoms:

Lighting doesn't match description
Inconsistent light direction
Muddy or flat lighting

Fixes:

Be extremely specific - "soft window light from left" not just "good lighting"
Reference time of day - "golden hour sunset" locks color temp
Specify light quality - "soft diffused" vs "hard directional"
Include shadows - "long shadows" or "no shadows"
Color temperature - "warm 3000K" or "cool 6500K"

Issue: Inconsistent Style Across Shots

Symptoms:

Shot 1 looks cinematic, Shot 2 looks like home video
Color grading changes between clips
Aesthetic mismatch

Fixes:

Copy-paste style block - Use identical style keywords in every prompt

Style: cinematic, 35mm film, warm color grade, film grain, Arri Alexa

Lock aspect ratio - Same dimensions for all shots
Post color grade - Match all shots in DaVinci Resolve
Reference same films/artists - "Blade Runner 2049 cinematography" in all prompts

Issue: Subject Not Matching Description

Symptoms:

Character has wrong hair color, clothing, features
Object doesn't match description

Fixes:

More detailed description - Age, ethnicity, specific features, exact clothing
Use reference image - Image-to-Video with exact subject
Regenerate multiple times - Select best match
Character consistency workflow - Generate character image first, then animate

Issue: Unwanted Elements Appearing

Symptoms:

Extra people/objects in frame
Background elements not requested
Artifacts or glitches

Fixes:

Negative prompts (Pika) - -neg "people, crowds, extra objects"
Simplify environment - Minimalist description reduces clutter
Be specific about "only" - "single person, no one else in frame"
Clean up in post - Rotoscope/mask out unwanted elements

Issue: Action Not Happening

Symptoms:

Static shot when movement requested
Subject doesn't perform described action

Fixes:

Increase motion parameter (Pika: -motion 3 or 4)
More explicit action - "actively running forward" vs just "running"
Simpler action - AI may not understand complex verbs
Motion Brush (Runway) - Paint specific areas to move
Image-to-Video - Show action mid-way in source image (running pose, not standing)

📚 Prompt Templates Library

Category: Portraits

Cinematic Portrait:

Close-up of [CHARACTER DESCRIPTION], looking directly at camera,
slow dolly in, golden hour sunlight from left, shallow depth of field,
cinematic portrait style, Rembrandt lighting, contemplative mood, 35mm film aesthetic

Fashion Portrait:

Medium close-up of [MODEL DESCRIPTION], turning head slowly to camera, confident expression,
studio lighting with rim light, high contrast, fashion photography style, editorial, bold and striking

Documentary Portrait:

Medium shot of [PERSON] in [ENVIRONMENT], natural expression, slight handheld camera movement,
natural window light, documentary photography style, authentic and raw, intimate

Category: Product

Luxury Product Showcase:

360° orbit around [PRODUCT], slow and smooth rotation,
studio lighting, black background, shallow depth of field,
product photography style, premium feel, reflections on surface

Lifestyle Product:

Close-up of hands [USING PRODUCT], natural movement,
soft window light, cozy home environment,
lifestyle photography, relatable and warm, authentic

Tech Product:

Slow dolly in on [TECH PRODUCT], sleek and modern,
cool blue lighting, minimalist white background,
futuristic style, clean and professional, sci-fi aesthetic

Category: Nature

Epic Landscape:

Aerial drone shot rising above [LANDSCAPE], revealing vast scale,
golden hour lighting, dramatic clouds, sweeping camera movement,
National Geographic style, epic and grand, cinematic

Intimate Nature:

Macro close-up of [NATURE ELEMENT - flower, insect, dewdrop],
slow push in, soft natural light, shallow depth of field,
nature documentary style, peaceful and serene, beautiful

Weather Drama:

Wide shot of [LANDSCAPE] during [WEATHER - storm, fog, rain],
camera slowly panning across scene, dramatic lighting, moody atmosphere,
atmospheric and intense, cinematic

Category: Urban

Cyberpunk City:

FPV drone racing through neon-lit cyberpunk city at night,
fast movement, rain-slicked streets, neon signs reflecting,
Blade Runner aesthetic, dynamic and energetic, futuristic

Urban Lifestyle:

Steadicam following person walking through busy city street,
natural movement, golden hour, urban documentary style,
authentic city life, energetic and real

Architectural:

Slow tilt up revealing modern skyscraper, camera at ground level,
blue hour twilight, symmetrical composition, architectural photography,
clean and geometric, impressive scale

Category: Action

Sports Action:

Slow motion 120fps, close-up of [ATHLETE] [ACTION],
intense focus, sweat visible, dramatic lighting,
sports photography style, powerful and dynamic

Chase Scene:

Tracking shot following [SUBJECT] running through [ENVIRONMENT],
fast movement, handheld camera shake, urgent pace,
action movie style, intense and chaotic

Slow Motion Drama:

Extreme slow motion 240fps, [SUBJECT] [DRAMATIC ACTION],
hair and clothing moving in slow-mo, particles visible in light,
cinematic style, epic and beautiful, dramatic

Category: Artistic

Surreal:

Wide shot of [SURREAL SCENE], dreamlike movement,
volumetric lighting, ethereal atmosphere, floating elements,
surrealist art style, mysterious and otherworldly

Abstract:

Close-up of [ABSTRACT SUBJECT], flowing movement,
colorful lighting, out-of-focus bokeh, experimental style,
abstract art, vibrant and artistic

Anime Style:

Medium shot of [CHARACTER] in anime style, [ACTION],
vibrant colors, Studio Ghibli aesthetic, hand-drawn look,
whimsical and nostalgic, detailed and beautiful

🎓 Pro Tips & Best Practices

Prompt Writing Strategy

1. Start Simple, Add Complexity

// Version 1 (Basic):
Woman walking in park

// Version 2 (Better):
Medium shot of young woman walking through park, sunny day

// Version 3 (Professional):
Medium shot of young Asian woman in summer dress walking through green park,
slow dolly forward, golden hour sunlight, shallow depth of field,
cinematic style, peaceful and serene

// Version 4 (Expert):
Medium shot of 25-year-old Asian woman with long black hair, wearing white summer dress,
walking slowly through lush green park, smiling softly, looking at camera,
slow dolly forward following her movement, golden hour sunlight from right creating rim light,
shallow depth of field f/2.8, bokeh background, 35mm film aesthetic, warm color grade,
peaceful and romantic atmosphere, natural motion blur

2. Layer Your Prompt Build prompts in blocks:

SUBJECT: [Detailed character/object]
ACTION: [What they're doing]
CAMERA: [Movement and framing]
ENVIRONMENT: [Where and when]
LIGHTING: [Quality and direction]
STYLE: [Aesthetic reference]
MOOD: [Emotional tone]
TECHNICAL: [Optional specs]

3. Use Reference Language

// Instead of generic:
"looks good"

// Use specific references:
"Wes Anderson symmetrical composition"
"Roger Deakins cinematography"
"Blade Runner 2049 lighting"
"Emmanuel Lubezki natural light"

Testing & Iteration

Scientific Approach:

1. Change One Variable at a Time

Base prompt:
"Woman walking in city, camera follows"

Test 1 - Camera angle:
"Woman walking in city, LOW ANGLE camera follows"

Test 2 - Lighting:
"Woman walking in city, camera follows, GOLDEN HOUR"

Test 3 - Style:
"Woman walking in city, camera follows, CINEMATIC STYLE"

2. Keep Prompt Journal

Save successful prompts
Note what worked / what didn't
Build personal prompt library
Track patterns across platforms

3. A/B Testing Generate 2 versions simultaneously:

Version A: Slow dolly in
Version B: Fast dolly in

Compare results, learn which works better

Platform Selection Guide

Choose Platform Based on Need:

Need	Best Platform	Why
Highest quality	Runway Gen-3	Best overall quality, most control
Character consistency	Runway Gen-3	Best at maintaining subject across frames
Fast iteration	Luma Dream Machine	Fastest generation (~2 min)
Selective animation	Pika Labs	Modify Region feature
Natural human motion	Kling AI	Excels at realistic people movement
Free/Open source	Stable Video Diffusion	Local deployment, no cost
Talking head	HeyGen / D-ID	Specialized for avatar speech
Complex camera moves	Runway Gen-3	Best camera control
Product showcase	Runway Gen-3	Clean, professional results
Experimental/artistic	Luma Dream Machine	Dreamlike, artistic quality

Avoiding Common Pitfalls

DON'T:

❌ Request copyrighted characters (Mickey Mouse, specific actors)
❌ Expect perfect text rendering
❌ Use vague descriptions ("looks cool", "nice lighting")
❌ Combine too many actions in one prompt
❌ Expect 100% consistency across separate generations
❌ Assume first generation will be perfect

DO:

✅ Be extremely specific and detailed
✅ Use cinematography vocabulary
✅ Reference time of day for lighting
✅ Specify camera movement clearly
✅ Keep actions simple and clear
✅ Generate multiple variations
✅ Use Image-to-Video for consistency
✅ Embrace happy accidents (AI may create unexpected beauty)

Rendering & Export Tips

1. Generation Settings:

Resolution: Max available for platform
Duration: Shorter clips (3-5s) = better quality than 10s
Seed: Save good seeds for later reuse

2. Post-Processing Workflow:

AI Video Export →
Upscale (Topaz Video AI to 4K) →
Color Grade (DaVinci Resolve) →
Sound Design (add ambient, music, SFX) →
Final Export (H.264 or ProRes)

3. Batch Processing:

Generate multiple variations overnight
Review all results next day
Select best, delete rest
Efficient use of generation credits

🎬 Real-World Example Prompts

Example 1: Coffee Shop Commercial

Shot 1 - Establishing:

Wide shot of cozy coffee shop exterior, morning light, people walking by,
slow dolly forward toward entrance, golden hour sunlight, shallow depth of field,
lifestyle commercial style, warm and inviting atmosphere

Shot 2 - Barista Making Coffee:

Close-up of barista's hands tamping espresso grounds, steam in background,
slow dolly in on portafilter, soft window light from left, shallow depth of field,
commercial product photography, warm tones, professional and artisan

Shot 3 - Coffee Pouring:

Extreme slow motion 120fps, espresso pouring into white cup, rich crema forming,
macro lens close-up, dark background with rim light, liquid motion sharp,
luxury commercial style, dramatic and beautiful

Shot 4 - Customer Enjoying:

Medium shot of young woman sitting by window, lifting cup and smiling,
slow zoom in, soft natural window light, shallow depth of field,
lifestyle photography, peaceful morning atmosphere, warm color grade

Example 2: Tech Product Launch

Shot 1 - Dark Reveal:

Extreme close-up of smartphone in darkness, screen suddenly lights up illuminating device,
slow dolly out revealing full phone, dramatic side lighting, black background,
tech commercial style, sleek and modern, futuristic

Shot 2 - 360° Showcase:

360° slow orbit around smartphone on pedestal, studio lighting revealing all angles,
black mirror surface reflection, dramatic three-point lighting, shallow depth of field,
product photography style, premium and luxurious

Shot 3 - Hands Interaction:

Close-up of hands picking up phone, finger swiping screen, interface glowing,
slight dolly in, soft gradient background, clean minimalist aesthetic,
Apple-style commercial, smooth and elegant

Example 3: Fashion/Beauty

Shot 1 - Face Focus:

Extreme close-up of model's eyes, slowly opening and looking at camera,
slow push in, golden hour lighting from 45° angle, catchlight in eyes,
fashion photography style, Rembrandt lighting, editorial and striking

Shot 2 - Hair Movement:

Medium close-up of model turning head, hair flowing in slow motion,
side lighting creating rim light on hair, dark background, shallow depth of field,
beauty commercial style, dramatic and elegant, 120fps slow-mo

Shot 3 - Full Portrait:

Full body shot of model in designer dress, walking toward camera confidently,
slow dolly back, studio with gradient background, butterfly lighting,
high fashion editorial, powerful and graceful

🎯 Conclusion

Mastering AI video generation requires:

Understanding cinematography - Camera, lighting, composition language
Platform knowledge - Strengths/weaknesses of each tool
Prompt engineering - Detailed, specific, layered descriptions
Iteration & testing - Scientific experimentation
Hybrid workflows - Combining AI with traditional post-production

Remember:

AI video is a TOOL, not magic - quality prompts = quality output
Current limitations (text, consistency) require creative workarounds
Technology evolving rapidly - techniques that fail today may work tomorrow
Best results combine AI generation with traditional filmmaking knowledge

Start creating! 🎬

Version: 1.0 Last Updated: 2025-10-24 Next Update: When new platforms/features release

Quick Reference Card:

PROMPT FORMULA:
[SHOT TYPE] of [SUBJECT] [ACTION] in [ENVIRONMENT],
[CAMERA MOVEMENT], [LIGHTING], [STYLE], [MOOD], [TECHNICAL]

REMEMBER:
✅ Be specific and detailed
✅ Use cinematography vocabulary
✅ Reference time of day
✅ Specify camera movement
✅ Keep actions simple
✅ Generate variations
✅ Embrace happy accidents

PLATFORMS:
• Runway Gen-3: Best quality, most control
• Pika Labs: Modify Region, creative control
• Kling AI: Realistic human motion
• Luma: Fast generation, dreamlike
• Stable Video: Open source, free

Now go create Hollywood-grade AI videos! 🎬🚀

Install Skill

SKILL.md

🎬 AI Video Prompting Skill - Hollywood Production Grade

📋 Table of Contents

🎯 Core Prompt Structure

Universal Prompt Formula (Works All Platforms)

Component Breakdown

1. Shot Type (Required)

2. Subject (Required)

3. Action (Required)

4. Environment (Required)

5. Camera Movement (Critical for AI Video)

6. Lighting (Defines Mood & Quality)

7. Style Modifiers (Aesthetic Direction)

8. Mood & Atmosphere (Emotional Tone)

9. Technical Specs (Optional but Powerful)

🎥 Camera Language - Professional Cinematography

Shot Angles (Camera Height & Orientation)

Camera Movements - Deep Dive

1. Pan (Horizontal Rotation)

2. Tilt (Vertical Rotation)

3. Dolly / Track (Camera Position Moves)

4. Zoom (Lens Focal Length Change)

5. Orbit / Arc (Circular Around Subject)

6. Crane / Jib (Vertical Sweeping)

7. Pedestal (Camera Raises/Lowers Vertically)

8. Handheld (Camera Operator Holds Camera)

9. Steadicam (Stabilized Handheld)

10. Drone / Aerial (Flying Camera)

Combining Camera Movements

🏃 Motion Control - Directing Movement

Subject Motion Speed

Motion Blur & Shutter

Speed Ramping

Parallax & Depth

👤 Character Animation - Directing Actors

Facial Expressions

Body Language

Character Interactions

Gaze & Eye Direction

Character Movement in Space

Performance Intensity

💡 Lighting & Atmosphere - Painting with Light

Natural Lighting Scenarios

Artificial Lighting Setups

Lighting Moods & Scenarios

Color Temperature & Grading

🎨 Scene Design - Building Worlds

Environment Categories

Atmospheric Elements

🎬 Platform-Specific Guides

Runway Gen-3 Alpha

Pika Labs 1.0

Kling AI (Kuaishou)

Luma Dream Machine

Stable Video Diffusion (Open Source)

🎬 Production Workflows

Text-to-Video Workflow

Image-to-Video Workflow

Multi-Shot Sequence Workflow

Product Showcase Workflow

🚀 Advanced Techniques

Consistency Across Generations

Controlling Difficult Concepts

Hybrid Workflows

Upscaling & Enhancement

Speed Ramping & Time Manipulation

❌ Troubleshooting Common Issues

Issue: Morphing / Warping

Issue: Jittery / Unstable Motion

Issue: Wrong Lighting

Issue: Inconsistent Style Across Shots

Issue: Subject Not Matching Description

Issue: Unwanted Elements Appearing

Issue: Action Not Happening

📚 Prompt Templates Library

Category: Portraits

Category: Product

Category: Nature

Category: Urban