Universal DNG Converter

Professional multi-format image converter with DNG output support for astronomical and photographic applications.

Overview

Universal DNG Converter is a robust, production-ready tool that converts various image formats to Adobe's Digital Negative (DNG) format while preserving metadata and providing intelligent scaling options. Originally designed for astronomical imaging workflows, it excels at handling high-dynamic-range data from FITS files while supporting standard photographic formats.

Step-by-Step Guide (Local Clone → First DNG)

1. Prerequisites

• Python 3.8–3.12
• Recommended: virtual environment
• Optional: opencv-python for EXR (install later if needed)

2. Clone

git clone https://github.com/dot-gabriel-ferrer/universal-dng-converter.git
cd universal-dng-converter

3. Environment

python -m venv .venv
source .venv/bin/activate  # Windows: .venv\\Scripts\\Activate.ps1

4. Install

Development (with tests, lint, type checking):

pip install -U pip
pip install -e .[dev]

Runtime only:

pip install .

Add EXR later:

pip install opencv-python

5. Sanity Check

universal-dng-converter --help

You should see CLI options including scaling and batch flags.

6. Single Conversion

universal-dng-converter --input tests/test_images/test_grayscale.jpg --output output.dng

If --output is a directory, the name is derived automatically.

7. Batch Conversion

Si el input es un directorio se convierte en lote automáticamente:

universal-dng-converter --input tests/test_images --output converted_dngs --recursive

Para filtrar extensiones actualmente usa todo el directorio (filtro por extensiones: mejora futura).

8. Scaling Methods

• auto (default): heuristic min/max
• linear: raw min→0, max→65535 mapping
• percentile: robust (defaults p2/p98) — use for noisy FITS
• none: no scaling (may clip or appear dark) Example:

universal-dng-converter --input image.fits --output out.dng --scaling percentile --pmin 1 --pmax 99

9. Python API

from universal_dng_converter import ImageConverter

conv = ImageConverter(scaling="auto")
dst = conv.convert_to_dng("tests/test_images/test_color.png", "color.dng")
print("Created:", dst)

for src, out in conv.batch_convert("tests/test_images", "batch_out", recursive=False):
    print(src, '->', out)

10. Validate Output

exiftool color.dng | grep -E "(Software|Bits Per Sample|DNG Version)"

Open in Lightroom / darktable (pseudo-DNG: TIFF + tags, no sensor CFA raw pipeline).

11. Tests & Quality

pytest -q
pre-commit install
git add . && git commit -m "quality run"

Hooks: black, flake8, mypy, etc.

12. Optional Dependencies & Troubleshooting

*Actualmente se fuerza compression="lzw"; se añadirá opción para none.

13. Limitations

• Outputs TIFF with DNG-like tags (not full raw sensor pipeline)
• No compression toggle yet (planned)
• Color profile not embedded (assumed sRGB)

14. Roadmap (Planned)

• Configurable compression (none | lzw | deflate)
• Parallel batch mode
• Enhanced metadata (camera model passthrough)
• Extension filtering flag
• EXR automated tests

15. Contributing (Quick)

1. Fork & branch
2. pip install -e .[dev]
3. pre-commit install
4. Add tests, run pytest -q
5. Open PR

✨ Features (Summary)

• 🎯 Multi-format support: FITS, PNG, JPEG, TIFF, BMP, GIF (first frame), EXR
• 🔧 Intelligent bit-depth conversion: 8-bit and 16-bit output with smart scaling
• 📊 Advanced scaling methods: Auto, linear, percentile, or no scaling
• 🏷️ Metadata preservation: Extracts and embeds FITS headers and EXIF data
• ⚡ Batch processing: Process single files or entire directories
• 🔄 Recursive directory scanning: Process subdirectories automatically
• 📝 Comprehensive logging: Detailed progress and error reporting
• 💻 Professional CLI: Full command-line interface with extensive options
• 🐍 Python API: Programmatic access for integration into workflows

🚀 Quick Start (Direct Install)

Installation (from PyPI when published)

pip install universal-dng-converter  # (pending publication)

EXR support:

pip install universal-dng-converter[exr]

If you need LZW compression support install:

pip install imagecodecs

Basic Usage

Convert a single image:

universal-dng-converter --input image.fits --output ./output/

Batch convert with advanced options:

universal-dng-converter \
  --input images/ \
  --output dng_output/ \
  --recursive \
  --bit-depth 16 \
  --scaling percentile \
  --verbose

Python API (Alt example)

from universal_dng_converter import DNGImageConverter
from pathlib import Path

converter = DNGImageConverter()
result = converter.convert_to_dng(
    input_path=Path("image.fits"),
    output_dir=Path("./output/"),
    bit_depth=16,
    scaling_method="auto"
)

📖 Documentation

• Installation Guide - Detailed setup instructions
• Usage Guide - Comprehensive usage examples
• API Reference - Complete API documentation
• Development Guide - Contributing and development setup

🎯 Use Cases

Astronomical Imaging

• Convert FITS files from telescopes and cameras to DNG for Adobe Lightroom/Photoshop
• Preserve astronomical metadata and handle high-dynamic-range data
• Batch process entire observation sessions

Photography Workflows

• Convert RAW-like formats to standardized DNG
• Maintain EXIF metadata across format conversions
• Integrate into automated processing pipelines
• Archive scientific datasets in standardized format
• Convert microscopy and medical imaging data

📋 Requirements

System Requirements

• Python 3.8 or higher
• 2GB+ RAM (for large astronomical images)
• Cross-platform: Windows, macOS, Linux

Dependencies

• astropy - FITS file handling and astronomical calculations
• Pillow - Core image processing capabilities
• tifffile - Advanced TIFF/DNG output generation
• numpy - High-performance numerical operations

🔧 Supported Formats

🎛️ Scaling Methods

The converter offers four intelligent scaling methods optimized for different data types:

• 🤖 Auto: Automatically selects optimal method based on image characteristics
• 📈 Linear: Full dynamic range mapping (min→0, max→max_value)
• 📊 Percentile: Robust scaling using 1st-99th percentiles (recommended for noisy data)
• 🔒 None: Preserves original values without modification

🏗️ Project Structure

universal-dng-converter/
├── 📁 src/
│   └── universal_dng_converter/     # Main package
│       ├── __init__.py              # Package initialization
│       ├── converter.py             # Core conversion logic
│       └── cli.py                   # Command-line interface
├── 📁 tests/                        # Test suite
│   ├── __init__.py
│   └── test_converter.py           # Unit tests
├── 📁 scripts/                      # Standalone scripts
│   └── convert-to-dng              # Direct execution script
├── 📁 docs/                         # Documentation
│   ├── installation.md
│   ├── usage.md
│   └── development.md
├── 📁 examples/                     # Usage examples
│   └── basic_usage.py
├── pyproject.toml                   # Modern Python packaging
├── requirements.txt                 # Runtime dependencies
├── requirements-dev.txt            # Development dependencies
└── README.md                        # This file

🤝 Contributing

We welcome contributions! Please see our Development Guide for details on:

• Setting up the development environment
• Code style guidelines (Black + isort)
• Testing procedures (pytest + coverage)
• Submitting pull requests

Quick Development Setup

git clone https://github.com/dot-gabriel-ferrer/universal-dng-converter.git
cd universal-dng-converter
pip install -e ".[dev]"
pre-commit install
pytest

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Adobe Systems for the DNG specification
• AstroPy Project for excellent FITS file handling
• Pillow Contributors for comprehensive image format support
• Python Scientific Community for the foundational tools

⭐ Star this repository if you find it useful!

Elías Gabriel Ferrer Jorge

License

MIT License

Version

1.0.0 - Professional multi-format DNG converter