liturgy.display

A reading management system controlled by a webpage "remote" and processed by voice recognition.

Overview

liturgy.display shows scripture/reading text on screen and advances slides by voice recognition word count using the Vosk speech-recognition engine. Designed for simple setup and offline/edge operation.

Features

• Offline speech recognition (Vosk)
• Configurable words-per-slide depending on screen size and preference
• Simple config in .env
• Basic support for USCCB readings (no API key required)

Prerequisites

• Python 3.8+ (or adjust for your runtime)
• pip
• Microphone access on the host machine
• A Vosk model (downloaded and extracted locally)

Download models: https://alphacephei.com/vosk/models

Installation

1. Clone repository

   git clone /path/to/liturgy.display
   cd liturgy.display
   

2. (Optional) Create and activate a virtualenv

   python -m venv venv
   source venv/bin/activate
   

3. Install dependencies

   pip install -r requirements.txt
   

Configuration

Create a .env file in the project root with at least:

WORDS_PER_SLIDE=40
MODEL_PATH=/your/path/to/vosk/model

• WORDS_PER_SLIDE: number of words shown per slide before auto-advance
• MODEL_PATH: path to the extracted Vosk model directory

Running

Start the app (example):

python main.py

Troubleshooting

• Model not found: verify MODEL_PATH points to the extracted model folder
• Microphone issues: ensure OS microphone permissions are granted and the correct device is selected.