PNG to WebP Converter

A high-performance Python script to batch convert PNG images to WebP format with multithreading, GPU acceleration, and customizable quality settings.

✨ Features

• 🚀 Multithreaded processing - Dramatically faster batch conversions
• 🎯 GPU acceleration - NVIDIA CUDA support via FFmpeg
• 🔄 Dual processing engines - Choose between Pillow (CPU) or FFmpeg
• 📁 Smart folder organization - Automatically creates organized output folders
• 🎛️ Flexible quality control - Lossy, lossless, and custom quality options
• 🛡️ Robust error handling - Thread-safe processing with detailed progress reporting
• 🔍 Case-insensitive detection - Finds all PNG files regardless of case
• 🌟 Transparency preservation - Maintains PNG alpha channels

📋 Requirements

Core Dependencies

• Python 3.6+
• Pillow library: pip install Pillow

Optional (for GPU acceleration)

• FFmpeg with CUDA support
• NVIDIA GPU with compatible drivers
• CUDA toolkit

🚀 Quick Start

Install Dependencies

pip install Pillow

Basic Usage

# Convert all PNGs in a folder (single-threaded)
python png_to_webp.py /path/to/images

# Fast multithreaded conversion
python png_to_webp.py /path/to/images -t 4

# GPU-accelerated conversion
python png_to_webp.py /path/to/images --ffmpeg --gpu

📖 Detailed Usage

Command Line Options

Option	Short	Description	Default
input_folder	-	Path to folder containing PNG files	Required
--output	-o	Custom output folder path	input_folder/webp_output
--quality	-q	WebP quality (0-100)	95
--lossless	-l	Use lossless compression	False
--threads	-t	Number of processing threads	1
--ffmpeg	-	Use FFmpeg instead of Pillow	False
--gpu	-	Enable GPU acceleration	False

Processing Modes

🔧 Pillow Mode (Default)

• Best for: Small to medium batches, simple setup
• Pros: Easy installation, reliable, good quality
• Cons: Slower for large batches

# Single-threaded
python png_to_webp.py ./images

# Multi-threaded (recommended)
python png_to_webp.py ./images -t 6

⚡ FFmpeg Mode

• Best for: Large batches, advanced users
• Pros: Faster processing, GPU support
• Cons: Requires FFmpeg installation

# FFmpeg with 8 threads
python png_to_webp.py ./images --ffmpeg -t 8

# FFmpeg with GPU acceleration
python png_to_webp.py ./images --ffmpeg --gpu -t 2

🎯 Performance Guide

Thread Recommendations

System Type	Recommended Threads	Example Usage
4-core CPU	4-6 threads	-t 4
8-core CPU	6-12 threads	-t 8
16-core CPU	12-20 threads	-t 16
GPU processing	1-2 threads	--gpu -t 2

Performance Comparison

Method	Speed	Setup Difficulty	Best For
Pillow (1 thread)	1x	Easy	Testing, small batches
Pillow (8 threads)	6x	Easy	Medium batches
FFmpeg (8 threads)	8x	Medium	Large batches
FFmpeg + GPU	15x	Hard	Massive batches

💡 Usage Examples

Basic Conversions

# Convert with default settings
python png_to_webp.py ./photos

# High quality conversion
python png_to_webp.py ./photos -q 98

# Lossless conversion
python png_to_webp.py ./artwork --lossless

Performance Optimized

# Fast multi-threaded conversion
python png_to_webp.py ./batch_images -t 8 -q 85

# Maximum speed with GPU
python png_to_webp.py ./huge_dataset --ffmpeg --gpu -t 2

# Custom output location
python png_to_webp.py ./input -o ./compressed -t 6

Specific Use Cases

# Web optimization (smaller files)
python png_to_webp.py ./website_images -q 80 -t 4

# Archive quality (perfect preservation)
python png_to_webp.py ./archives --lossless -t 2

# Production pipeline (balanced)
python png_to_webp.py ./production --ffmpeg -q 90 -t 8

📁 File Organization

The script automatically creates organized output:

your_images/
├── photo1.png
├── photo2.png
├── screenshot.PNG
└── webp_output/           # ← Auto-created
    ├── photo1.webp
    ├── photo2.webp
    └── screenshot.webp

⚙️ Quality Guidelines

Quality Range	Use Case	File Size	Visual Quality
60-75	Web thumbnails	Very small	Good
75-85	General web use	Small	Very good
85-95	High-quality images	Medium	Excellent
95-100	Professional work	Large	Near perfect
Lossless	Archival/editing	Largest	Perfect

🔧 GPU Setup (Optional)

NVIDIA GPU Requirements

1. Compatible GPU: GTX 900 series or newer
2. NVIDIA drivers: Latest version
3. CUDA toolkit: Version 11.0+
4. FFmpeg with CUDA: Compiled with --enable-cuda

Verify GPU Support

# Check if FFmpeg supports CUDA
ffmpeg -hwaccels

# Test GPU conversion
python png_to_webp.py ./test_folder --ffmpeg --gpu -t 1

📊 Sample Output

Found 156 PNG files to convert using FFmpeg...
Using 8 thread(s) for processing
GPU acceleration enabled (requires compatible hardware)

✓ Converted (GPU): image001.png -> image001.webp
✓ Converted (GPU): photo_large.PNG -> photo_large.webp
✓ Converted (GPU): screenshot.png -> screenshot.webp
✗ Error converting corrupted.png: Invalid image format
✓ Converted (GPU): artwork.png -> artwork.webp

FFmpeg conversion complete! 155/156 files converted successfully.

🐛 Troubleshooting

Common Issues

"No PNG files found"

# Check file extensions and path
ls *.png *.PNG

GPU acceleration not working

# Verify CUDA support
nvidia-smi
ffmpeg -hwaccels

Out of memory with many threads

# Reduce thread count for large images
python png_to_webp.py ./4k_images -t 2

Permission errors

# Check folder permissions
chmod 755 /path/to/folder

Performance Tips

• Large images: Use fewer threads to avoid memory issues
• Many small images: Use more threads for better throughput
• GPU mode: Stick to 1-2 threads (GPU handles parallelism)
• Mixed sizes: Start with 4-6 threads and adjust

🔄 Migration from Other Tools

From ImageMagick

# Old: magick *.png -quality 90 output_%03d.webp
# New: 
python png_to_webp.py ./ -q 90 -t 4

From cwebp

# Old: for f in *.png; do cwebp "$f" -o "${f%.png}.webp"; done
# New:
python png_to_webp.py ./ -t 8

📈 Batch Processing Tips

For Massive Datasets (1000+ images)

# Process in chunks to monitor progress
python png_to_webp.py ./chunk1 --ffmpeg --gpu -t 2
python png_to_webp.py ./chunk2 --ffmpeg --gpu -t 2

For Production Workflows

# Create a processing script
#!/bin/bash
python png_to_webp.py ./raw_images --ffmpeg -q 90 -t 8 -o ./web_ready
python png_to_webp.py ./high_res --lossless -t 4 -o ./archive

🤝 Getting Help

# View all options
python png_to_webp.py --help

# Test with single file
python png_to_webp.py ./single_image_folder -t 1

📝 License

This script is provided as-is for educational and practical use. Feel free to modify and distribute.

---

💡 Pro Tip: Start with default settings and 4 threads, then optimize based on your hardware and needs!