> ## Knowledge Base Index
> Fetch the complete knowledge base index at: https://1-more-shot.crisp.help/sitemap.xml
> Use this file to discover available pages before exploring further.
> Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash).

# How to use the Automatic Video feature?

The **Auto Video** tool is optimized to be a **visual generator** focused primarily on the singer character and is not designed for complex scene adherence. 

**Why Prompts Aren't Always Followed?** 
The system is designed to intentionally **simplify complex scenes** or ignore elements that could lead to poor or inconsistent results. It prioritizes a **high-quality, stable video** over strictly adhering to every detail of an overly complex prompt. Additionally, the tool is currently optimized **only to generate the singer character**, which is why it struggles with multiple characters or other complex elements. 

**Recommended Workflow** 
For the best possible results with Quick Generate, we recommend focusing your prompt on the **main character** and a **simple environment**:
* **Prompt**: Keep the prompt simple, specifying only 2 or 3 **locations** and an **outfit** if desired.
* **Image: **The ideal image to upload should be a full-face visible shot of the singer, without accessories like hats, large sunglasses, or masks. Our system needs to clearly see the facial details to maximize consistency.

**Accessory Consistency (Hats, Glasses, etc.)** 
Maintaining accessories like hats or sunglasses is challenging because the AI is focused on the face. To increase your chances of success:
* **Make it Part of the Identity**: When writing your prompt, explicitly state that the accessory is an essential part of the singer's appearance or costume. For example, instead of just "wearing a hat," write: "The singer is famous for their signature black fedora which must be worn in every shot."
* **Be Specific**: Describe the accessory clearly (e.g., "a sleek, silver aviator-style sunglasses," not just "sunglasses").
* **Visual Reinforcement**: If the accessory appears clearly in the uploaded reference image, the AI has a better visual cue to follow, though consistency is still not guaranteed.

Here a quick tutorial about the automatic video tool: https://youtu.be/Ic8D0wC5Bic