AI Videos with Sound

​

Overview

• Videos with integrated dialogue
• Synchronized character speech
• Scene cuts and transitions
• Professional camera movements

​

Before you start

• Access to Kive workspace
• Basic understanding of video prompting

​

Steps

​

Step 1: Access Video Generation

1. Open the Generation section
2. Switch to Video mode using the mode selector
3. Click the Settings icon on the right side

​

Step 2: Configure Video Settings

• Mode: Premium (for best quality)
• Aspect Ratio: 16:9 (or your preferred format)

​

Step 3: Master Video Prompting

​

Prompting structure

• Early prompt = early in video
• Later prompt = later in video
• Describe actions in chronological order

1. Set the Scene (light initial description)
2. Focus on Speaking Character
3. Add Dialogue (in quotes)
4. Sprinkle Scene Descriptions throughout

​

Advanced techniques

First scene description, "character says this"
Cut to: New scene description, "character says that"

• Include camera directions in your description
• Examples: “camera pans left”, “zoom in on face”, “aerial shot”

​

Step 4: Use the Preview

• Rough idea of colors and lighting
• General aesthetic understanding
• Draft visualization (not final quality)

​

Step 5: Generate Your Video

1. Review your prompt and settings
2. Click Generate
3. Render time: approximately 1 minute
4. Set up multiple variations while waiting

​

Step 6: Review with Sound

1. Play the video with sound enabled
2. Check dialogue synchronization
3. Save successful generations to boards

​

Prompting

​

Basic prompt structure

[Scene setting], [character description] says "[dialogue]", 
[action description], [camera movement]

​

Example prompts

Office setting with natural light, professional woman 
in blue suit says "Welcome to our presentation", 
she gestures to the screen, camera slowly zooms in

Busy coffee shop, barista smiles and says "What can I get you?",
Cut to: Customer thinking, then says "Large cappuccino please",
Cut to: Close-up of coffee being made, steam rising

​

Character consistency tips

• Describe characters clearly at first mention
• Use consistent descriptors throughout
• Reference previous character descriptions for cuts

​

Settings

​

Techniques

​

Create conversations

Person A in red jacket says "How are you?",
Person B smiles and responds "Great, thanks!",
They shake hands, camera circles around them

​

Scene transitions

Sunrise over mountains, birds chirping,
Cut to: Hiker on trail says "What a beautiful morning",
Cut to: Summit view, hiker raises arms in triumph

​

Emotional moments

Close-up on tearful face, soft lighting,
"I never thought this day would come" she whispers,
Camera pulls back to reveal celebration scene

​

Troubleshooting

​

Audio Sync Issues

• Keep dialogue concise and clear
• Place dialogue quotes immediately after character description
• Avoid overlapping speech in single shots

​

Scene Consistency

• Maintain consistent character descriptions
• Use “same [character/location]” references
• Keep lighting and time of day consistent

​

Render Quality

• Always use Premium mode for final outputs
• Check preview for general composition
• Generate multiple variations for options

​

Tips

1. Start Simple: Begin with single character, single scene
2. Build Complexity: Add cuts and camera moves gradually
3. Use Starting Points: Check homepage for prompt templates
4. Batch Generation: Queue multiple variations while waiting
5. Sound Design: Describe ambient sounds for richer audio

​

Quick reference

​

See also

• Explore AI character models for consistent characters
• Learn about AI studios for visual presets
• Master AI product shots for commercial content