Multi-Shot Sequences

Create cinematic sequences with multiple shots while maintaining consistency

Planning Your Sequence

Before generating, plan your sequence like a script:

1.Define the location and environment
2.Introduce your characters and their appearance
3.Plan shot transitions and camera movements
4.Write detailed prompts that reference previous shots

Complete Example

import asyncio
from ministudio import VideoOrchestrator, VideoGenerationRequest
from ministudio.providers import VertexAIProvider

async def main():
    # Initialize
    provider = VertexAIProvider()
    orchestrator = VideoOrchestrator(provider)
    
    # Define sequence
    shots = [
        VideoGenerationRequest(
            prompt="Wide shot of a modern glass skyscraper at sunrise, golden light reflecting off the windows",
            duration_seconds=6,
            aspect_ratio="16:9",
            quality="high"
        ),
        VideoGenerationRequest(
            prompt="Inside a sleek office conference room. A woman in business attire stands by a window overlooking the city",
            duration_seconds=8,
            aspect_ratio="16:9",
            quality="high"
        ),
        VideoGenerationRequest(
            prompt="Close-up of the same woman looking confident, city skyline visible through window behind her",
            duration_seconds=5,
            aspect_ratio="16:9",
            quality="high"
        ),
        VideoGenerationRequest(
            prompt="The woman walks toward a table with laptop and documents, the cityscape still visible",
            duration_seconds=6,
            aspect_ratio="16:9",
            quality="high"
        ),
    ]
    
    # Generate sequence
    results = await orchestrator.generate_sequence(shots)
    
    # Process results
    for i, result in enumerate(results):
        if result.success:
            print(f"Shot {i+1}: {result.video_path}")
            print(f"Duration: {result.duration_seconds}s")
        else:
            print(f"Shot {i+1} failed: {result.error}")
    
    print(f"Sequence complete! Total duration: {orchestrator.get_state().total_duration}s")

if __name__ == "__main__":
    asyncio.run(main())

The orchestrator automatically extracts the last frame from each shot to maintain continuity in the next shot.

Tips for Better Sequences

Maintain Consistency

•Use the same character descriptions across shots
•Reference visual elements from previous shots
•Mention time of day or lighting consistency

Camera Work

•Start with wide shots to establish location
•Move to medium shots for interactions
•Use close-ups for emotional moments

Pacing

•Longer shots (8-10s) for establishing scenes
•Medium shots (6-8s) for action
•Short shots (3-5s) for emphasis or dialogue

Managing State Between Sequences

# Complete first sequence
results1 = await orchestrator.generate_sequence(indoor_shots)

# Reset state for different location
orchestrator.reset()

# Generate new sequence without interference
results2 = await orchestrator.generate_sequence(outdoor_shots)