Camera Instructions

Overview

Camera Instructions in ACT 3 AI allow you to define how a scene or shot should be captured visually. They are written as part of your shot prompts or entered directly in the Editor workspace to give the AI rendering engines detailed cinematic direction.

These instructions mimic real-world cinematography, covering framing, movement, lensing, and perspective.

Key Capabilities

• Framing: Define wide shots, close-ups, over-the-shoulder views, or extreme close-ups.
• Camera Movement: Add pans, tilts, dollies, cranes, or handheld effects.
• Lenses & Angles: Simulate wide-angle, telephoto, fisheye, POV, or Dutch tilt shots.
• Focus & Depth of Field: Control aperture (f-stops), shallow vs. deep focus, rack focus transitions.
• Composition: Rule of thirds, leading lines, character placement, and horizon level.
• Cinematic Styles: Apply descriptors like “cinematic,” “documentary-style,” “surreal,” or “noir lighting.”

How to Use

1. Open a scene or shot in the Editor workspace.
2. In the prompt field, add descriptive camera instructions (e.g., “wide establishing shot of a futuristic city, slow dolly in”).
3. Select rendering engine (e.g., Google Veo 3, Runway Integration, WAN AI Integration).
4. Preview in Timeline or Preview Pane.
5. Refine by editing shot prompts or re-rendering.

Example Instructions

• “Close-up of the hero’s face, handheld shaky camera, low lighting”
• “Wide establishing shot of desert, drone aerial view, sweeping pan left to right”
• “Over-the-shoulder shot of the detective looking at evidence, shallow depth of field, background blurred”
• “Tracking shot through a neon city street, low angle, Steadicam smoothness”

Integration with Workflow

• Script Editor → Add camera notes in script margin.
• Storyboards & Panels → Align drawn storyboards with camera prompts.
• Top-Down View → Place cameras and paths on a virtual map.
• Build Video → Ensure final rendered video follows intended framing.

Best Practices

• Be concise but specific (10–20 words works best).
• Combine cinematic terms with visual descriptions for stronger results.
• Break complex camera moves into multiple shots instead of one long prompt.
• Match camera style to genre (e.g., handheld for horror, Steadicam for action, static shots for drama).

Troubleshooting

• AI ignores instructions → Simplify or split into smaller prompts.
• Unnatural angles → Use real cinematography terms (e.g., “low angle,” “crane shot”) instead of vague phrases.
• Blurry results → Adjust resolution settings in the render engine or use draft previews for testing.

See Also

• Shot Prompting
• Scenes · Shots · Story Beats
• Storyboard & Panels
• Top-Down View
• Google Veo 3 Integration
• Runway Integration
• WAN AI Integration

Contact Us if you have any problems using our product, or if you have questions.