AI Model Wiki

    Skip to main content
    25 min
    Lesson 2 of 4

    Video Generation and Editing Techniques

    From the course Meta Movie Gen Studio

    Master Movie Gen's generation and editing workflows for creating professional multimedia content.

    Text-to-Video Generation

    Effective Prompting

    Scene Description Framework:
    [SUBJECT] + [ACTION] + [ENVIRONMENT] + [CAMERA] + [MOOD]
    
    Example: "A golden retriever puppy running through autumn leaves 
    in a sunlit forest clearing, tracking shot following the dog, 
    warm and joyful atmosphere"
    

    Cinematic Techniques

    # Action sequences
    "Slow motion close-up of coffee being poured into a ceramic mug,
    steam rising, warm kitchen lighting, shallow depth of field"
    
    # Establishing shots
    "Aerial drone shot slowly descending over a coastal city at golden hour,
    waves crashing against cliffs, cinematic wide angle"
    
    # Character-driven scenes
    "A chef carefully plating a dessert in a professional kitchen,
    focused expression, overhead camera angle, warm ambient lighting"
    

    Image-to-Video Generation

    Animating Still Images

    • Upload a reference image as the starting frame
    • Describe the desired motion and action
    • Movie Gen maintains visual consistency with the source
    • Ideal for product demonstrations and hero animations

    Personalized Video Generation

    • Upload 1-5 reference images of a subject
    • Generate videos featuring that specific subject
    • Maintains identity across different scenes and actions
    • Useful for personalized marketing and content creation

    Movie Gen Edit

    Text-Based Video Editing

    Editing Instructions:
    - "Change the background to a tropical beach"
    - "Make it nighttime with city lights"
    - "Add falling snow to the scene"
    - "Change the person's outfit to a red dress"
    - "Remove the car from the background"
    

    Localized Editing

    • Specify which elements to modify
    • Surrounding context preserved automatically
    • Temporal consistency maintained through edits
    • Multiple sequential edits supported

    Audio Generation

    Synchronized Sound Design

    Audio Generation Prompts:
    - "Forest ambience with birds chirping and leaves rustling"
    - "Upbeat electronic music with driving beat, 120 BPM"
    - "Dramatic orchestral score building to climax"
    - "City traffic sounds with distant sirens"
    

    Audio-Video Synchronization

    • Sound effects timed to visual events
    • Music mood matches scene emotional tone
    • Environmental audio matches setting
    • Smooth transitions between audio elements

    Key Takeaways

    • •Structured prompting with Subject + Action + Environment + Camera + Mood produces best results
    • •Personalized video generation maintains subject identity from reference images
    • •Text-based editing modifies existing videos while preserving temporal consistency
    • •Audio generation produces synchronized sound effects, music, and ambience
    0 of 4 lessons marked complete

    Progress is kept in this browser only.