# Composite Poster

Whether you're just starting out with AI art or you're a seasoned designer, you're bound to discover new techniques and find fresh inspiration in this process.

<figure><img src="/files/zaTWdJzFTLBg16YoY3b2" alt=""><figcaption></figcaption></figure>

## STEP 1 Open SeaArt Canvas

1. Open SeaArt Canvas and select the <mark style="background-color:yellow;">'Realtime Canvas.'</mark> SeaArt Canvas offers 'Generate Canvas,' 'Realtime Canvas,' and 'Design Canvas.'"

* Realtime Canvas allows you to view and adjust images in real-time.
* Generate Canvas enables direct use of AI to generate images, perform partial redraws, and erasing, among other functions.

<figure><img src="/files/YX2hDacGyiKSKVssYnzd" alt=""><figcaption></figcaption></figure>

## STEP 2 Determine the Poster Theme

Before we begin, we need to clarify the theme of the poster we want to create. For instance, our theme could be <mark style="background-color:yellow;">"Being Chased by a Giant Cat."</mark> With a theme in mind, we can use the generation mode to produce several sketches as references, providing us with a general creative direction.

* **Prompt:** A huge ferocious long-haired cat running, chasing a man, disaster film.

<figure><img src="/files/bXUI7dtcmDPihNmCI1NJ" alt=""><figcaption></figcaption></figure>

As you can see, the directly generated image does not meet our needs, and there is a significant gap between the actual result and our envisioned effect.

## STEP 3 Generate Poster Materials

Next, generate the materials needed for the poster individually and save them as cutouts.&#x20;

Open the Realtime Canvas, and first adjust the image generation parameters, selecting the appropriate model and lora.

* Model: RealVisXL V4.0
* Lora: Movie Aesthetic

<figure><img src="/files/Ow25q3xf9DWJ96yIDxmi" alt=""><figcaption></figcaption></figure>

The two sampler options in the settings control the generation mode and Realtime Canvas respectively. We will change the sampling method in Realtime Canvas **from LCM to DPM++ 2M Karras.** Although this will increase the image generation time, it enhances the quality of the generated images.

<figure><img src="/files/K1bD3U7ALruiDawxsmQE" alt=""><figcaption></figcaption></figure>

Once the parameters are set, we directly enter text into the prompt box to generate the cat material images.

* **Prompt:** A ferocious long-haired cat running, cute, huge cat, low pov.

<figure><img src="/files/mAGkRKOBPaEDjwRppZRJ" alt="" width="428"><figcaption></figcaption></figure>

After selecting the appropriate material image, click <mark style="background-color:yellow;">**"Save as New Layer"**</mark> in the top right corner to copy the image to the generation area. Remember to click the **pause button** during this operation.

<figure><img src="/files/fbh4Qeqto78pB1ewA0b6" alt="" width="434"><figcaption></figcaption></figure>

Next, click on the image and select the <mark style="background-color:yellow;">**"Remove Background"**</mark> feature to cut out the cat.

<figure><img src="/files/zB9SWpWDtMMxnIC9DVjk" alt="" width="464"><figcaption></figcaption></figure>

Then, click the layers button in the bottom left corner to open the layers panel, and **click the little eye icon to hide the cat layer.**

<figure><img src="/files/00WAgO2GxwjlseTk1ShH" alt="" width="563"><figcaption></figcaption></figure>

Repeat the steps above to continue saving other materials to the generation area.

<figure><img src="/files/y3yvY7UAL8UzzPfuaC68" alt="" width="292"><figcaption></figcaption></figure>

> **Prompt:** A terrified man, running in fear, low POV, white background

<figure><img src="/files/tOAWJR2NaWQNzLB8q4ef" alt="" width="303"><figcaption></figcaption></figure>

> **Prompt:** Darkness, leaf-covered clearing, jungle, path, huge dust and smoke, low perspective

## STEP 4 Generate Images

Adjust the size and position of the materials to align with the perspective.&#x20;

<figure><img src="/files/HZdgkevztmzy06lfEDKy" alt="" width="563"><figcaption></figcaption></figure>

Enter the scene prompt, describing the content of the image as completely as possible.

<figure><img src="/files/cdDV0NvkdP5nJHdWJL6T" alt=""><figcaption></figcaption></figure>

> **Prompt:** Cute giant cat running fast, chasing a man who is running, scared man, speed blur, smoke filled, rock splash, wide angle lens, motion blur, film lighting, Sony camera, backlighting, edge lighting, ray tracing, film grain

After generating a satisfactory image, click the <mark style="background-color:yellow;">**download**</mark> button in the top right corner to save the image. Finally, it's recommended to use <mark style="background-color:yellow;">**AI Image Upscaler**</mark> to add detail to the scene.

<figure><img src="/files/4YQCpzyZUjKk2NDAq7rY" alt=""><figcaption></figcaption></figure>

The final product effect is as follows:

<figure><img src="/files/WRnGdjSiIs0b6rT1vEDy" alt="" width="375"><figcaption></figcaption></figure>

Using the SeaArt Canvas allows you to more clearly adjust the image to generate the perfect picture in your mind!


---

# 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://docs.seaart.ai/guide-1/3-advanced-guide/composite-poster.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.
