> ## Documentation Index
> Fetch the complete documentation index at: https://docs.rewatch.studio/llms.txt
> Use this file to discover all available pages before exploring further.

# Generating images

> Describe an image in conversation, redo it by its number, mention a character with @Name.

The “Image generator” tool (in the “Generators” section) works like a conversation. You write in the “Request” field, an assistant rewrites your request into a precise prompt, then launches the generation. Each image gets a number: you use it to redo it (“redo number 2 with warmer light”).

## The form that works

The assistant looks for: subject, framing, light, materials, mood, style. Give it these six elements.

```text theme={null}
[Precise subject] in [place], [close up, wide shot, top view] framing,
[type, time, color] light, [wood, metal, wet fabric] materials,
[calm, tense, joyful] mood, [photo, soft 3D, illustration] style.
```

## Before / after

> [tab] Weak
>
> ```text theme={null}
> A kitchen
> ```

> [tab] Improved
>
> ```text theme={null}
> A farmhouse kitchen at dawn, wide shot from the doorway, golden light through the window, worn wooden table, enamel mugs, calm mood, photo with film grain.
> ```
>
> Each addition removes a decision the AI would have made at random: framing, light, materials, rendering.

## Mentioning a character

Type `@Name` to mention one of your characters (created in [AI characters](https://docs.rewatch.studio/tools/characters.md)). Their reference images and their description go along with the request automatically. So write what they do, where and how, never their appearance.

> [tab] Weak
>
> ```text theme={null}
> @Maya, a woman with red curly hair and a yellow slicker, in the rain
> ```
>
> You redescribe Maya: you risk contradicting her.

> [tab] Improved
>
> ```text theme={null}
> @Maya waits under a bus shelter at night, she looks at her watch, fine rain, blue neon lights, medium three quarter shot.
> ```

## Redoing or varying an image

Refer to the number: “redo number 3 as a tighter shot”, “same scene in daylight”. The assistant takes the image's prompt, changes only what you ask, and attaches it as a reference if you want to keep the same subject. On an image, “Use as reference” attaches it to the next request, and “Animate” sends it to the [Video clips](https://docs.rewatch.studio/tools/video-clips.md) tool.

## Common mistakes

* A request that is too short (“a kitchen”): add light, framing and style.
* Describing the appearance of a mentioned character: say only what they do.
* Putting readable text in the image (poster, sign): image models draw words badly. Add the text in the editor.
* Several requests at once without a number: “number 2”, “the last one” designate the session's images, name them.
* An ambiguous request (which image, how many): the assistant asks you a question instead of launching.

## Settings that go with it

Next to the “Request” field:

* **Engine**: some engines accept more reference images or hold a face better. See [Choosing an engine](https://docs.rewatch.studio/models/choose.md).
* **Format**, **Resolution** and **Number of images**.
* **Attach**: import an image (PNG, JPEG or WebP) or take an image from the session. If an engine doesn't accept references, the app tells you before sending.
* The estimated cost is shown before sending. Above a certain threshold, the assistant asks for your confirmation with the amount: see [Spending credits wisely](https://docs.rewatch.studio/models/credits.md).

## Examples to copy

```text theme={null}
Cinematic portrait of a detective in the rain, medium shot, blue and red neon on her face, wet trench coat, tense mood, realistic photo.
```

```text theme={null}
An isometric miniature city flooded with water, top view, soft 3D rendering, pastel colors, calm mood.
```

Each request states the subject, the framing, the light and the style: that is what the assistant uses to write the prompt.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.