Story App - AI-Powered Voice Script Generator

A modern web application that allows storytellers, voice-over artists, and content creators to write scripts with detailed voice instructions, generate AI voice audio, and export high-quality audio files.

Features

• Character Management: Create and manage characters with different voice profiles
• Script Editor: Write and organize dialogue lines with character assignments
• Voice Instructions: Add detailed instructions for tone, pacing, emphasis, and emotion
• AI Voice Generation: Generate high-quality voice audio using OpenAI's TTS models
• Audio Preview: Listen to individual lines or play the entire script
• Export Options: Export as individual audio files (ZIP) or a single combined file
• Script Formatting: AI-powered script formatter to enhance dialogue with professional voice instructions
• Story Management: Save, load, and manage multiple stories

Installation

Prerequisites

• Node.js (v14 or higher)
• NPM or Yarn
• OpenAI API key

Setup

1. Clone the repository:

   git clone https://github.com/sumitchatterjee13/VocalCanvas.git
   cd story-app

2. Install dependencies:

   Using installation scripts:

   • Windows: Run install.bat by double-clicking or from command prompt
   • macOS/Linux: Run ./install.sh in terminal (you may need to set permissions with chmod +x install.sh)

   Manually:

   npm install
   # or
   yarn install

3. Environment setup:

   • Rename .env.local.example to .env.local
   • Add your OpenAI API key:

     OPENAI_API_KEY=your_api_key_here
     

   • Optionally configure AI models:

     FORMATTER_MODEL=gpt-4o-mini
     

4. Start the application:

   Using run scripts:

   • Windows: Run run.bat by double-clicking or from command prompt
   • macOS/Linux: Run ./run.sh in terminal (you may need to set permissions with chmod +x run.sh)

   The application will automatically open in your default web browser.

   Manually:

   npm run dev
   # or
   yarn dev

5. If starting manually, open http://localhost:3000 in your browser

How to Use

Creating Characters

1. Navigate to the "Characters" panel on the left side
2. Enter a character name
3. Select a voice from the dropdown menu (options include alloy, echo, fable, onyx, nova, and more)
4. Click "Add Character"

Writing Dialogue

1. In the "Add Dialogue Line" section:

   • Select a character from the dropdown
   • Write the dialogue text
   • Optionally add voice instructions
   • Click "Add Line"

2. Manage existing lines:

   • Play: Listen to individual lines
   • Edit: Modify the text or instructions
   • Delete: Remove a line
   • Drag and drop: Reorder lines

Using the AI Script Formatter

1. Click the "Format Script" button in the top navigation bar

2. In the popup, paste your raw dialogue text in this format:

   CHARACTER: Dialogue text. (Acting directions)
   

3. Click "Format with AI"

4. The AI will transform your dialogue into professionally formatted lines with detailed voice instructions:

   CHARACTER: Dialogue text without acting directions.
   
   (Voice Affect: Low, hushed, and suspenseful; Tone: Deeply serious and mysterious; 
   Pacing: Slow, deliberate; Emotion: Restrained yet intense; 
   Emphasis: On key words; Pronunciation: Slightly elongated vowels; 
   Pauses: Strategic pauses after important phrases)
   

5. Click "Import Formatted Script" to use the enhanced script in your story

Audio Controls

• Play Line: Listen to a specific dialogue line
• Play All: Listen to the entire script sequentially
• Regenerate: Create fresh audio for a particular line that doesn't sound right
• Export Line: Download a single line as a WAV file
• Export ZIP: Download all lines as separate audio files in a ZIP archive
• Export Single: Combine all lines into a single audio file

Story Management

• Save: Save your current story to local storage
• New Story: Create a new, blank story
• Load Story: Load a previously saved story

Technical Details

• Built with Next.js, React, and TypeScript
• Uses OpenAI's GPT-4o-mini model for script formatting and enhancement
• Uses OpenAI's Text-to-Speech models for audio generation
• Responsive design with Tailwind CSS

Limitations

• Requires an OpenAI API key
• Audio generation consumes OpenAI API credits
• Maximum script size is limited by OpenAI's API token limits

Example Script Formatting

Before formatting:

DETECTIVE: The footprints lead to the garden. (slowly, examining the ground) They stop right here.
GARDENER: (nervously) I was just pruning the roses this morning. Nothing unusual.
DETECTIVE: (suspiciously) The roses, you say? (pauses) The murder weapon was a pruning shear.

After AI formatting:

DETECTIVE: The footprints lead to the garden. They stop right here.

(Voice Affect: Analytical and focused; Tone: Authoritative with underlying tension; Pacing: Methodical, slowing down at "right here"; Emotion: Intense concentration mixed with suspicion; Emphasis: On "footprints" and "stop"; Pronunciation: Crisp consonants; Pauses: Brief pause after "garden", longer pause before "They stop")

GARDENER: I was just pruning the roses this morning. Nothing unusual.

(Voice Affect: Anxious and defensive; Tone: Attempting casual but strained; Pacing: Slightly rushed, especially on "Nothing unusual"; Emotion: Nervousness masked with forced calm; Emphasis: Unnatural emphasis on "just" and "Nothing"; Pronunciation: Slight tremor in voice; Pauses: Quick intake of breath before speaking, minimal pauses)

DETECTIVE: The roses, you say? The murder weapon was a pruning shear.

(Voice Affect: Sharp and cutting; Tone: Deliberately provocative and accusatory; Pacing: Slow and measured for impact; Emotion: Controlled intensity building to revelation; Emphasis: Heavy emphasis on "roses" and "pruning shear"; Pronunciation: Elongated "s" in "roses"; Pauses: Significant pause after "you say?", dramatic pause before revealing the murder weapon)

License

APACHE-2.0 License

Acknowledgments

• Voice models and text formatting by OpenAI
• Icons from Heroicons

---

Note: This project requires an OpenAI API key and will consume API credits when generating audio or formatting scripts.