Historical Echoes - Victorian & Georgian Chat

A web application featuring historical characters from 1750-1899, spanning both Georgian and Victorian eras. Chat with characters from 18th-century London to 19th-century Victorian society, using Ollama for language generation and Amazon Polly for text-to-speech with synchronized mouth animation.

Features

• Historical Characters: Speak in authentic Georgian and Victorian English (1750-1899)
• Multiple Eras: Characters from both Georgian (1750-1830) and Victorian (1830-1899) periods
• Text-to-Speech: Amazon Polly neural voice synthesis
• Mouth Animation: Viseme-based lip sync with SVG avatar
• Local AI: Ollama for language generation (no external API calls)
• Responsive UI: Beautiful historical-themed interface

Available Characters

Georgian Era (1750-1830)

• The Correspondent: A witty and charming Londoner from 1780s
• Sir Reginald: A witty and charming gentleman from 1780s London
• Lady Catherine: An educated woman from 1780s London
• Mr. Thompson: A successful merchant from 1780s London
• The Casual Correspondent: A more relaxed Londoner from 1780s

Victorian Era (1830-1899)

• Charles Dickens: The celebrated Victorian novelist and social critic
• Mr. Boz: A Victorian novelist and social observer (Dickensian character)
• The Victorian Gentleman: A refined gentleman from 1850s-1890s London

Prerequisites

Before running this application, ensure you have:

1. Node.js 22+ installed

   node -v  # Should show v22 or higher

2. Ollama installed and running

   # Install Ollama (if not already installed)
   # Visit: https://ollama.ai/download
   
   # Start Ollama service
   ollama serve
   
   # Pull the required model (in a new terminal)
   ollama pull llama3.2:3b-instruct

3. AWS CLI configured with Polly access

   # Install AWS CLI (if not already installed)
   # Visit: https://aws.amazon.com/cli/
   
   # Configure AWS credentials
   aws configure
   # Enter your AWS Access Key ID, Secret Access Key, and region (e.g., us-east-1)

4. AWS Polly Permissions: Ensure your AWS user has the following permissions:

   • polly:SynthesizeSpeech
   • polly:StartSpeechSynthesisTask

Installation

1. Clone or download this project

   cd "/Users/dylan/Documents/Cursor Projects/1800s CHAT"

2. Install dependencies

   npm install

Running the Application

1. Start Ollama (if not already running)

   ollama serve

2. Start the Georgian Chat server

   npm start
   # or
   node server.js

3. Open your browser Navigate to: http://localhost:8787

Usage

1. Start a conversation: Type a message in the input field and press Enter or click Send
2. Listen to responses: The avatar will speak with synchronized mouth movement
3. Try these example prompts:
   • "Good day! What news from London?"
   • "Tell me about the latest fashions"
   • "What do you think of the King's policies?"
   • "How is the weather treating you?"

Configuration

Changing the Voice

To use a different Amazon Polly voice, edit server.js and modify the VoiceId parameter:

const params = {
  OutputFormat: "mp3",
  VoiceId: "Amy", // Change this to another voice
  // ... other parameters
};

Popular British voices:

• Amy (British English, Female)
• Emma (British English, Female)
• Brian (British English, Male)
• Arthur (British English, Male)

Adjusting the Character

Modify the SYSTEM_PROMPT in server.js to change the character's personality:

const SYSTEM_PROMPT = `You are "The Correspondent," a Londoner living in 1760–1800. 
// ... customize the character description
`;

Changing the AI Model

To use a different Ollama model, update the model name in the /chat endpoint:

body: JSON.stringify({
  model: "llama3.2:3b-instruct", // Change this
  // ... other parameters
})

Troubleshooting

Common Issues

1. "Ollama API error"

   • Ensure Ollama is running: ollama serve
   • Check if the model is installed: ollama list
   • Pull the model: ollama pull llama3.2:3b-instruct

2. "TTS API error" or "Failed to generate speech"

   • Verify AWS credentials: aws sts get-caller-identity
   • Check AWS region in server.js (default: us-east-1)
   • Ensure Polly permissions are granted

3. Audio not playing

   • Check browser console for errors
   • Ensure browser supports Web Audio API
   • Try refreshing the page

4. Mouth not animating

   • Check browser console for viseme errors
   • Ensure audio is playing (check volume)
   • Try shorter messages first

Debug Mode

Enable debug logging by adding this to server.js:

// Add after imports
process.env.DEBUG = 'true';

Project Structure

georgian_chat/
├── server.js          # Express.js backend
├── package.json       # Dependencies and scripts
├── public/
│   └── index.html     # Frontend with chat UI and avatar
└── README.md          # This file

API Endpoints

• POST /chat - Send message to Ollama, get character response
• POST /tts - Convert text to speech with viseme data
• GET /health - Health check endpoint

Technical Details

Viseme Mapping

The application maps Amazon Polly visemes to mouth shapes:

• sil, PP, FF, TH, DD, kk, CH, SS, nn, RR → Closed mouth
• aa → Open mouth
• E, I → "E" mouth shape
• O, U → "O" mouth shape

Audio Processing

• Uses Web Audio API for precise timing
• Base64-encoded MP3 audio from Amazon Polly
• Viseme timestamps synchronized with audio playback

Future Enhancements

• Streaming responses: Real-time text generation
• Emotion system: Facial expressions based on sentiment
• Voice selection: Multiple character voices
• Offline mode: Replace Polly with local TTS
• Docker deployment: Containerized setup

License

ISC License - feel free to modify and distribute.

Support

For issues or questions:

1. Check the troubleshooting section above
2. Verify all prerequisites are installed
3. Check browser console for error messages
4. Ensure all services (Ollama, AWS) are running properly