Answer

Short answer:

Yes, you can control camera movement in Rendair AI videos using Motion Presets (such as zoom, pan, or drone shot) and text prompts. For more precise control over the animation's trajectory, you can also define specific Start and End frames to guide the video from one state to another.

Overview

Architectural visualization often requires more than a static image to convey space and depth. Rendair AI’s Image-to-Video tool allows you to animate still renderings into short, 5-second clips.

Rather than requiring complex animation software or manual keyframing, the platform uses a semantic approach. You define the intent of the camera movement, whether it is a slow pan across a facade or a cinematic drone shot, and the AI interprets the depth and perspective of your image to execute that movement naturally. This bridge between static rendering and animation is designed for quick client presentations where motion helps clarify spatial relationships.

How it works

The process abstracts away the technical complexities of camera rigging. You provide the visual context, and the system handles the motion path.

1. Select the Tool: Navigate to the Image-to-Video feature in the Create dashboard.

2. Upload Input: Upload your base render (or a generated image from a previous step).

3. Choose Motion Control:

• Presets: Select a pre-defined camera movement from the interface (e.g., Zoom In, Pan Left, Cinematic Drone).

• Prompting: Alternatively, describe the specific movement in the text prompt area (e.g., "Slow camera drift to the right").

• Start/End Frames (Advanced): For specific transitions, such as moving from a construction site to a finished building, you can upload a "Start Frame" and an "End Frame." The AI will generate the motion required to morph between them.

1. Generate: The system processes the depth map of the image and applies the requested motion, delivering a video file in seconds.

Capabilities

Rendair AI provides several layers of control depending on how specific your needs are.

• Directional Presets: You can apply standard architectural photography movements with a single click.

• Pan: Horizontal movement (Left/Right) to show context or width.

• Zoom: Push in (Dolly In) to focus on details or Pull out (Dolly Out) to reveal surroundings.

• Tilt: Vertical movement (Up/Down) for high-rise or atrium shots.

• Cinematic Styles:

• Drone Shot: Simulates an aerial flyover or orbiting motion, ideal for exterior perspectives.

• Wide Shot: Creates a subtle, establishing motion suitable for large interior spaces.

• Transition Control: By defining both a start and end image, you can force the camera to travel between two specific visual states. This is particularly effective for time-lapse effects or "before and after" reveals.

• Text-Based Nuance: You can combine presets with text instructions to refine the speed or mood, such as specifying "slow, smooth motion" to avoid rapid or jarring cuts.

Inputs and outputs

Inputs

• Base Image: JPG, PNG, WEBP, or HEIC (up to 16MB).

• Optional End Frame: A second image to guide the animation's final state.

Outputs

• Format: MP4 video file.

• Duration: Approximately 5 seconds.

• Resolution: Dependent on the input image aspect ratio and quality.

When to use this

• Social Media Teasers: Static images often get lower engagement; a simple pan or zoom adds necessary visual interest for Instagram or LinkedIn.

• Client Presentations: Use a "Zoom In" motion to draw the client's eye to a specific material or design feature without saying a word.

• Spatial Clarification: Use a "Drone" or "Orbit" motion to help clients understand the depth of a building and its relation to the site.

• Progress Visualization: Use Start/End frames to animate a renovation proposal, showing the existing condition morphing into the proposed design.

Limitations or notes

• Duration is Fixed: Videos are generated as short clips (approx. 5 seconds). This is intended for loops and clips, not full-length walkthroughs.

• Generative Edges: When panning or zooming out, the AI must "invent" data that exists outside the frame of your original image. While usually accurate, check the edges of the video for hallucinations or artifacts.

• No Manual Keyframes: You cannot manually plot a spline or camera path (e.g., "go forward 2 meters, then turn left"). The control is directional and semantic, not coordinate-based.

• Credit Usage: Video generation consumes credits. These credits do not expire as long as you have an active subscription.