CLI

Requirements

• Node.js v20 or higher

Installation

npm install -g dill-pixel

Available Commands

Create a New Project

Create a new Dill Pixel project:

dill-pixel create my-project

Options:

• --use-pnpm - Use PNPM as package manager (recommended)
• --use-yarn - Use Yarn as package manager

Example:

dill-pixel create my-game --use-pnpm

Update Dill Pixel

Update to the latest version:

dill-pixel update

Install Peer Dependencies

Install required peer dependencies:

dill-pixel install

Audio Tools

Convert Audio

Convert audio files to the correct format:

dill-pixel audio compress [input-directory]

Default input directory: ./src/assets/audio

Output will be placed in [input-directory]/output

Generate Captions

Generate caption files from CSV:

dill-pixel audio captions [input-directory] [output-directory]

Default paths:

• Input: ./src/assets/audio/captions
• Output: ./src/assets/json

Voiceover CSV Generation

Generate voiceover CSV files from your i18n locale files:

dill-pixel vo generate [input-directory] [output-directory]

Default paths:

• Input: ./src/locales
• Output: ./src/assets/audio/vo/csv

The tool supports parsing .js, .ts, and .json locale files. It processes the following formats:

1. Simple string values:

{
  "greeting": "Hello world"
}

Generated CSV output:

FILENAME    LINE         SCRATCH_VOICE     SCRATCH_GENDER    SCRATCH_LANGUAGE    CAPTION
greeting    Hello world  en-US-Wavenet-D   female           en-US              Y

2. Array of options (using pipe separator):

{
  "greeting": "[Hello|Hi|Hey]"
}

Generated CSV output:

FILENAME      LINE    SCRATCH_VOICE     SCRATCH_GENDER    SCRATCH_LANGUAGE    CAPTION
greeting_0    Hello   en-US-Wavenet-D   female           en-US              Y
greeting_1    Hi      en-US-Wavenet-D   female           en-US              Y
greeting_2    Hey     en-US-Wavenet-D   female           en-US              Y

3. Dynamic replacements with options:

{
  "greeting": {
    "replace": "Hello {name}",
    "replaceOpts": "[John|Jane|Alex]"
  }
}

Generated CSV output:

FILENAME      LINE          SCRATCH_VOICE     SCRATCH_GENDER    SCRATCH_LANGUAGE    CAPTION
greeting_0    Hello John    en-US-Wavenet-D   female           en-US              Y
greeting_1    Hello Jane    en-US-Wavenet-D   female           en-US              Y
greeting_2    Hello Alex    en-US-Wavenet-D   female           en-US              Y

The generated CSV files will include the following columns:

• FILENAME - Unique identifier for the voice line
• LINE - The text to be voiced
• SCRATCH_VOICE - Voice ID (default: en-US-Wavenet-D)
• SCRATCH_GENDER - Voice gender (default: female)
• SCRATCH_LANGUAGE - Voice language (default: en-US)
• CAPTION - Caption flag (default: Y)

Text formatting:

• <strong>text</strong> is converted to *text*
• <em>text</em> is converted to _text_
• Other HTML tags are removed

Using Generated CSV Files with Google Text-to-Speech

The generated CSV files are formatted to work with Google Cloud Text-to-Speech service. Here’s how to use them:

1. Prerequisites:

   • A Google Cloud account with Text-to-Speech API enabled
   • Google Cloud credentials configured

2. CSV Format: The generated CSV files contain the necessary metadata for voice generation:

   • FILENAME: Used as the output audio file name
   • LINE: The text to be converted to speech
   • SCRATCH_VOICE: Google Cloud voice ID (e.g., ‘en-US-Wavenet-D’)
   • SCRATCH_GENDER: Voice gender preference
   • SCRATCH_LANGUAGE: Target language code
   • CAPTION: Whether to generate captions (Y/N)

3. Example Workflow:

   # 1. Generate CSV files from your locale files
   dill-pixel vo generate ./src/locales
   
   # 2. Use Google Cloud Text-to-Speech to generate audio files
   # Example using gcloud command (you'll need to adapt this for your specific setup)
   gcloud text-to-speech synthesize-speech \
     --voice-name=en-US-Wavenet-D \
     --text="$(cat path/to/generated.csv)" \
     --output-dir=./src/assets/audio/vo

4. After Generation:

   • Audio files will be created using the FILENAME from your CSV
   • Generate captions using dill-pixel audio captions

Version Information

Display the current version:

dill-pixel version

Project Structure

When creating a new project using dill-pixel create, it will generate a standard project structure with all necessary configuration files. For more details on the project structure, see the Quick Start Guide

The easiest way to get started with Dill Pixel is to use the CLI to create a new project. Run the following command in your terminal:

npx dill-pixel create my-application

Tip

Replace my-application with whatever you’d like to name your project

This will create a new directory called my-application with all the necessary files and dependencies pre-configured.

Using PNPM (Recommended)

Tip

If you’re using PNPM, you can use the following command instead:

npx dill-pixel create my-application --use-pnpm