Wan2.1 Movie Maker

A web interface for generating and combining 10-second videos using the Wan2.1 text-to-video model.

Features

• Generate videos from text prompts
• Enhance prompts with detailed descriptions for better video generation
• Combine multiple clips into longer movies
• Batch process multiple prompts
• GPU usage monitoring to prevent conflicts
• Dark mode UI for reduced eye strain
• Automatic poster image generation for video thumbnails
• Restart capability for interrupted generation processes
• Voice sample feedback system for collecting and processing user feedback
• Voice exploration and generation tools for creating and managing voice samples
• Voice status monitoring for tracking generation progress
• IP address monitoring for servers with dynamic IPs
• Enhanced voice generation with female-dominant voice sets

Requirements

• Python 3.8+
• FFmpeg
• Wan2.1 model
• Flask
• CUDA-compatible GPU

Installation

1. Clone this repository
2. Ensure FFmpeg is installed
3. Set up the Wan2.1 model
4. Install Python dependencies

Usage

Starting the Web Interface

./run.sh --port=5001

Running a Test Server

For testing new features without affecting the main service:

./test_server.sh [port]

Example:

./test_server.sh 5002

This will start a test instance on port 5002 (or any specified port).

Enhancing Prompts

./enhance_only.sh --input=prompts.txt

Generating Videos with Enhanced Prompts

./run_enhanced_generation.sh --input=enhanced_prompts/enhanced_TIMESTAMP_only.txt --seconds=1

Batch Processing

./batch_generate.py prompts.txt --seconds=1

Poster Image Generation

The system includes an automatic poster image generation service that extracts frames from videos to use as thumbnails.

# Run manually (one-time)
python3 extract_frames.py

# Run in monitor mode
python3 extract_frames.py --monitor --interval 30

For more details, see README_POSTER_SERVICE.md.

Restarting Generation Process

If the video generation process stops unexpectedly, you can restart it from where it left off:

./restart_hdmy5movie_generation.sh

This script will automatically determine the next prompt to process and create any necessary directories.

For more details, see README_RESTART_GENERATION.md.

Voice Sample Feedback System

The system includes a feedback collection and processing system for voice samples:

# Start the feedback server
python3 feedback_server.py

# Update voice sample descriptions based on feedback
python3 update_descriptions.py

For more details, see FEEDBACK_SYSTEM.md.

Voice Exploration and Generation

The system includes tools for exploring and generating voice samples:

# Start the feedback server which also serves the voice explorer interface
python3 feedback_server.py
# Then navigate to http://localhost:8000/voice_explorer.html

# Generate voice samples with different parameters
./explore_voices.sh --quick

# Monitor voice generation status
./start_status_server.sh

For more details, see VOICE_SYSTEM.md.

IP Address Monitoring

The system includes scripts to monitor the server's public IP address and send notifications when it changes:

# Check for IP changes manually
./check_ip_change.sh

# Run the complete monitoring process
./monitor_ip_changes.sh

A cron job is set up to run this check automatically every hour.

For more details, see IP_MONITOR_README.md.

Voice Generation

Generating New Voice Sets

To generate a new set of 100 voices with 80% female voices:

./generate_more_female_voices.sh

To generate additional voice sets:

./generate_100_more_voices.sh

These scripts generate:

• 80 female voices with higher expressivity
• 20 male voices with varied characteristics

Preserving Voice Samples

To ensure all generated voices are preserved:

python3 merge_voice_samples.py --replace

For automatic preservation during generation:

./auto_merge_voices.sh

For more details, see VOICE_GENERATION.md.

Exploring Generated Voices

Start the feedback server and navigate to the Voice Explorer:

python3 feedback_server.py

Then open http://localhost:8000/voice_explorer.html in your browser.

Project Structure

• app.py: Main Flask application
• run.sh: Script to start the web interface
• test_server.sh: Script to start a test server on a specified port
• enhance_prompts.py: Script to enhance prompts
• batch_generate.py: Script for batch processing
• extract_frames.py: Script to extract frames from videos for poster images
• hdmy5movie_poster_service.service: Systemd service file for poster generation
• restart_hdmy5movie_generation.sh: Script to restart the generation process
• feedback_server.py: Server for collecting voice sample feedback
• update_descriptions.py: Script for updating voice sample descriptions based on feedback
• voice_feedback_db.json: Database of user feedback on voice samples
• voice_samples.js: Voice sample data with descriptions
• voice_explorer.html: Web interface for exploring voice samples
• listen.html: Simplified interface for playing voice samples
• voice_status.html: Interface for monitoring voice generation status
• start_feedback_server.sh: Script to start the feedback server
• start_status_server.sh: Script to start the status server
• explore_voices.sh: Script to generate voice samples with different parameters
• check_files.py: Script to check and monitor voice generation files
• check_ip_change.sh: Script to check for IP address changes
• send_ip_notification.sh: Script to send IP change notifications
• monitor_ip_changes.sh: Combined script for IP monitoring
• templates/: HTML templates for the web interface
• static/: Static files (CSS, JS)
• clips/: Generated video clips
• movies/: Combined movies
• hdmy5movie_videos/: Video files for the HDMY 5 Movie project
• hdmy5movie_posters/: Poster images for the HDMY 5 Movie project
• hdmy5movie_voices/: Voice samples for the HDMY 5 Movie project
• voices/: Directory containing general voice samples
• voice_poc/: Directory containing voice generation proof of concept implementations
• logs/: Log files for various services

License

MIT