> For the complete documentation index, see [llms.txt](https://docs.seaart.ai/guide-1/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.seaart.ai/guide-1/seaart-ai-basic-function/ai-audio.md).

# AI Audio

Turn an idea into a song, make a character in an image or video speak with Lip Sync, or add matching sound effects - all in one Audio workspace.

## Overview

To open SeaArt AI Audio, click **Create** in the top navigation bar, then select **Audio** in the left navigation.

<figure><img src="https://2219884424-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FigAtVLBrlaI8jVruJfC8%2Fuploads%2FjqZHa9wdg37JKk81pglp%2Faudio.webp?alt=media&amp;token=16fcece5-6798-4d0f-9543-71a090ac0b6b" alt="SeaArt AI Homepage"><figcaption></figcaption></figure>

<figure><img src="https://2219884424-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FigAtVLBrlaI8jVruJfC8%2Fuploads%2FrV7TMhNfXgGbyYVnreui%2FAudio2.webp?alt=media&amp;token=b70ac1c5-d091-4b9b-8ca8-de6c62518c47" alt="SeaArt AI Audio"><figcaption></figcaption></figure>

| Tool         | Output                                                     | Use it for                                              |
| ------------ | ---------------------------------------------------------- | ------------------------------------------------------- |
| Song         | A complete song with vocals, or instrumental music         | Creating music, background tracks, or songs with lyrics |
| Lip Sync     | A lip-synced video, using text-to-speech or uploaded audio | Making a character in an image or video speak           |
| Sound Effect | A sound effect, or a video with matched sound effects      | Creating sound effects or adding them to a video        |

Choose **Song** for music, **Lip Sync** when a character needs to speak, and **Sound Effect** when you need a sound effect or a video with matched sound.

The credit cost for each generation is shown on the **Generate** button.

## Song

Song is a text-to-audio tool for generating a complete song from lyrics or a description. Use **Vocal** when the song needs a singer, and **Instrumental** for music without vocals.

<figure><img src="https://2219884424-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FigAtVLBrlaI8jVruJfC8%2Fuploads%2FM1Yl0MKS28ZQb6BFHNA6%2Faudio3.webp?alt=media&amp;token=09e3ec8b-65b4-40b9-946a-f997551e4613" alt="SeaArt AI Text to Song"><figcaption></figcaption></figure>

### Create a song

1. Open the Song tab.
2. Select a model from the model panel.
3. Choose a Generation Mode: **Vocal** or **Instrumental**.
4. For Vocal, enter lyrics or a song description. For Instrumental, skip this step.
5. Set the style by entering a style or selecting tags.
6. Click **Generate**.

### Lyrics

* Supports English, Chinese, Japanese, Korean, Spanish, Portuguese, German, French, Italian, and Russian.
* Use structural terms such as verse and chorus for better results.
* Click the magic wand in the Lyrics field to use ✨**Quick Generate**, which creates a complete lyrics template that you can edit.
* Instrumental mode does not require lyrics.

### Style

Style tags tell the model what the song should sound like. Enter a style in the Style field or combine tags:

* **#Genre** - Pop, R\&B, Rock, Disco, Electronic, Folk, Hip-hop, Blues, Classical, Game Music, and more.
* **#Emotion** - the mood of the song.
* **#Voice** - vocal type. Available in Vocal mode.
* **#Instrument** - instrumentation and arrangement.

### Tips

* If you do not have finished lyrics, enter a song description instead and use style tags to set the direction.
* Match the style tags to the mood and use of the song.
* Use **Quick Generate** when you want a structured starting point, then replace the template content with your own lyrics.

{% file src="/files/P7YlgfL5IlM2Qw8X4ATC" %}

## Lip Sync

Create a lip-synced video in which a character in an uploaded image or video speaks the provided audio. Use **Create Audio** for text-to-speech voice generation, or **Upload Audio** to use an existing recording.

<figure><img src="https://2219884424-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FigAtVLBrlaI8jVruJfC8%2Fuploads%2FtG9LjZ7wrD6Py5sXLfBR%2Faudio4.webp?alt=media&amp;token=ff4dc522-091a-4458-b267-3f90b1f9c729" alt="SeaArt AI Lip Sync - Text to Speech"><figcaption></figcaption></figure>

### Create a lip-sync video

1. Open the **Lip Sync** tab.
2. Upload an image or video, or select one from History.
3. Choose **Create Audio** or **Upload Audio**.
4. For Create Audio, enter the dialogue, select a voice, adjust pacing and emotion, and use **Listen** to check the voice duration.
5. For Upload Audio, upload an existing audio file.
6. Optional: enable **Private Creation**.
7. Click **Generate**.

**Input requirements:**

* **Video** - MP4 or MOV; up to 100 MB; 2–30 seconds; 512–2160 px.
* **Image** - JPG or PNG; up to 5 MB; up to 4096 px.

### Voice settings

* **Timbre** - choose from categories such as Professional, Male, Female, Boy, and Girl, with multiple voices in each category.
* **Pacing** - adjust the speed of the voice.
* **Emotion** - choose Happy, Sad, Angry, Fearful, Disgusted, Surprised, or Neutral.
* **Listen** - preview the voice and confirm its duration before generating.

### Tips

* Enter at least 30 characters for better results. The maximum is 200 characters.
* Keep the dialogue natural and suited to the character.
* Use a clear, front-facing face with minimal occlusion.
* Match the voice and emotion to the character and scene.
* Use Listen to check the pacing and duration before you generate the video.

## Sound Effect

Sound Effect is a text-to-audio tool that supports two workflows: generate a sound effect from a description, or upload a video and let SeaArt create a matching sound description and add the sound effects to the video.

<figure><img src="https://2219884424-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FigAtVLBrlaI8jVruJfC8%2Fuploads%2FncRKdTfQ6yQt3Y6qPPSQ%2Faudio5.webp?alt=media&amp;token=1de28aa8-1876-4071-b2ec-ec2ea7225e8c" alt="SeaArt AI Sound Effect"><figcaption></figcaption></figure>

### Generate from a description

1. Open the **Sound Effect** tab.
2. Enter a description of the sound effect in **Enter sound effect inspiration description**.
3. Optional: select a **Recommended Sound**, or click **Change** for more options.
4. Open **Advanced Config** to adjust the duration or add negative prompts.
5. Optional: enable **Private Creation**.
6. Click **Generate**.

<figure><img src="https://2219884424-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FigAtVLBrlaI8jVruJfC8%2Fuploads%2FrxT6OSOy8QjpHWSLx862%2Faudio6.webp?alt=media&amp;token=71c514e3-48fe-4ffa-936d-43c1418b71eb" alt="Match a Sound Effect to a video"><figcaption></figcaption></figure>

### Match sound effects to a video

1. Open the **Sound Effect** tab.
2. Under **Or select video for voiceover**, upload a video.
3. SeaArt generates a sound description that matches the video. Review it and click **Confirm** to use it, or **Cancel** to discard it.
4. The description is placed in **Enter sound effect inspiration description**. Edit it if needed.
5. Adjust **Advanced Config** if needed.
6. Optional: enable **Private Creation**.
7. Click **Generate**. The result is the video with the matched sound effects added.

**Advanced Config:**&#x20;

* **Duration** - set the length of the sound effect.
* **Negative Prompts** - describe sounds you do not want to include.

{% file src="/files/KRpRQVJGgqZC621tXYSs" %}

### Tips

* Describe the sound source, action, material, and environment. For example: "Coins dropping on a wooden table."
* When matching sound to a video, review the generated description before confirming, and edit it to refine the result.
* Use Negative Prompts to exclude unwanted sounds.
