# How to fine-tune prompts to adjust a dual-character scene to a single-character scene

<figure><img src="/files/QLP9ljflf0KNkAlTHr4M" alt=""><figcaption></figcaption></figure>

When we render double-character scenes on the ReadKidz storyboard page, we often encounter a series of problems such as <mark style="color:red;">the characters merging together</mark>, <mark style="color:red;">lack of interaction between characters</mark>, and <mark style="color:red;">chaotic character features</mark>!

<figure><img src="/files/Q6ncJwN9PTynEFnuJZZD" alt=""><figcaption><p>ReadKidz Advanced Tutorial</p></figcaption></figure>

How should we fine-tune the dual-character scene to achieve satisfactory picture book production?

<mark style="color:purple;">In fact, we just need to modify the two-person scene into a single-person scene, which will not affect the structure and integrity of the entire story!</mark>

**Step 1 :**

After entering the <mark style="color:purple;">PPT page,</mark> review the picture book subtitles, as well as the especially important placeholder images!

<mark style="color:purple;">What is a placeholder image? It is the character portrait created when creating a character, and by selecting the corresponding character portrait based on the prompt words and subtitles, you can consistently generate illustrations with that character! Therefore, the selection and deletion of placeholder images play a crucial role in the image!</mark>

<figure><img src="/files/dzurNzjFNvla0HozV31e" alt=""><figcaption><p>ReadKidz Advanced Tutorial</p></figcaption></figure>

**Step 2 :**

Find the subtitles for the corresponding <mark style="color:purple;">two-person scenes</mark>, and click “<mark style="color:purple;">AI Images</mark>” to enter the image generation page!

<figure><img src="/files/ugMnufXzBLHcSi0Etjh4" alt=""><figcaption><p>ReadKidz Advanced Tutorial</p></figcaption></figure>

<figure><img src="/files/PEvIXrOQPKfATdhXpe3U" alt=""><figcaption><p>ReadKidz Advanced Tutorial</p></figcaption></figure>

**Step 3 :**

We will extract the keywords and divide them according to a <mark style="color:purple;">certain format</mark>!

<figure><img src="/files/scFqCX9r3XMdotFSFxef" alt=""><figcaption><p>ReadKidz Advanced Tutorial</p></figcaption></figure>

**Step 4 :**

Then, based on the subtitle's corresponding story scene, <mark style="color:purple;">keep the important character</mark>!

For example："But remember," said the Fairy Godmother, "you must be home before the clock strikes midnight!"

We only need to <mark style="color:purple;">keep the character "Fairy Godmother"</mark>! Then we need to adjust the original prompt format!

<mark style="color:purple;">1.Delete the number of people in the scene, the description of the actions of the two characters, and the character information of the character "Cinderella"!</mark>

2.The atmosphere in the space-time description <mark style="color:purple;">must not contain interactive prompt words!</mark>

3.Character information prompt left <mark style="color:purple;">cannot have any prompts that interact with other characters!</mark> <mark style="color:red;">Note: Be sure to add "solo" at the beginning of the character information to ensure the stability of single-player scenes!</mark>

<figure><img src="/files/RPCI7LlJOSZpwa2bwI4s" alt=""><figcaption><p>ReadKidz Advanced Tutorial</p></figcaption></figure>

**Step 5 :**

Based on the keywords of the roles, remove the placeholder images of unnecessary roles in <mark style="color:purple;">“Portrait reference”</mark> to generate a stable <mark style="color:purple;">single-player scene image</mark>!

<figure><img src="/files/xo0cXxDx4R3PdG8OsagU" alt=""><figcaption><p>ReadKidz Advanced Tutorial</p></figcaption></figure>

<figure><img src="/files/if3xqfZNMIpVASOL8nm9" alt=""><figcaption><p>ReadKidz Advanced Tutorial</p></figcaption></figure>


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://readkidz.gitbook.io/readkidz/readkidz-advanced-tutorial/how-to-fine-tune-prompts-to-adjust-a-dual-character-scene-to-a-single-character-scene.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
