name: mccurry-photo-processing description: Batch process Sony RAW (.ARW) files with Steve McCurry-inspired narrative color grading using darktable. Use when user needs professional warm, cinematic color grading for Sony camera RAW files.
Steve McCurry Photo Processing
Transform your Sony RAW photos with professional color grading inspired by Steve McCurry's iconic National Geographic photography style.
Features
- Professional Color Grading: Warm, rich tones inspired by McCurry's legendary style
- Batch Processing: Process 50-200+ images efficiently
- High Quality: Output JPEGs at 85-95% quality with no visible loss
- Open Source: 100% free using darktable
- Command Line: Automated workflow for efficient processing
- Multiple Presets: Standard and vibrant variants
Quick Start
1. Install darktable
# Check if darktable is installed
bash scripts/install_darktable.sh
# Or install manually
brew install darktable
2. Process Your Photos
# Basic usage: process all ARW files
python scripts/batch_process.py raw/ output/
# Custom quality
python scripts/batch_process.py raw/ output/ --quality 95
# Use vibrant preset
python scripts/batch_process.py raw/ output/ --preset assets/mccurry-vibrant.xmp
3. Validate Output
# Check output quality
python scripts/validate_output.py output/
# Export detailed report
python scripts/validate_output.py output/ --json
Steve McCurry Color Style
This skill replicates the distinctive visual characteristics of Steve McCurry's photography:
- Warm Golden Tones: Rich oranges, reds, and earth colors
- High Contrast: Punchy but retains shadow and highlight detail
- Saturated Midtones: Colors pop without looking unnatural
- Golden Hour Feel: Amber/yellow color cast like Kodachrome film
- Deep Shadows: Rich blacks with preserved texture
- Vibrant Blues/Teals: Cool tones provide visual balance
- Cinematic Quality: Film-like smoothness with storytelling mood
Directory Structure
your-project/
├── raw/ # Input: Sony ARW files
└── output/ # Output: Processed JPEGs
Presets
mccurry-narrative.xmp (Standard)
Balanced McCurry-inspired grading suitable for most scenarios. This is the default preset.
Best for:
- General photography
- Portraits
- Street photography
- Documentary work
mccurry-vibrant.xmp (Vibrant)
More saturated variant for high-impact images.
Best for:
- Landscapes
- Travel photography
- Images with strong colors
- Festival/cultural events
Detailed Usage
Batch Processing Script
python scripts/batch_process.py INPUT_DIR OUTPUT_DIR [OPTIONS]
Arguments:
INPUT_DIR- Directory containing ARW filesOUTPUT_DIR- Directory for processed JPEGs (created if needed)
Options:
--quality QUALITY- JPEG quality 85-100 (default: 92)--preset FILE- Custom XMP preset file--verbose- Show detailed processing information
Examples:
# Process with default settings (92% quality, standard preset)
python scripts/batch_process.py ~/Photos/Raw ~/Photos/Processed
# Maximum quality for print
python scripts/batch_process.py raw/ output/ --quality 98
# Vibrant preset for landscapes
python scripts/batch_process.py landscape_raw/ landscape_out/ --preset assets/mccurry-vibrant.xmp
# Verbose output for debugging
python scripts/batch_process.py raw/ output/ --verbose
Validation Script
python scripts/validate_output.py OUTPUT_DIR [OPTIONS]
Arguments:
OUTPUT_DIR- Directory containing processed JPEGs
Options:
--json- Export detailed JSON validation report--verbose- Show all warnings (not just errors)
Examples:
# Quick validation
python scripts/validate_output.py output/
# Full report with JSON export
python scripts/validate_output.py output/ --json --verbose
Installation Script
bash scripts/install_darktable.sh
Checks for darktable installation and installs via Homebrew if needed.
Performance
Processing Speed:
- ~5-15 seconds per image (depends on image size and Mac specs)
- 100 images: 10-25 minutes
- 200 images: 20-45 minutes
System Requirements:
- macOS 11.0 (Big Sur) or later
- 8GB RAM minimum (16GB recommended for large batches)
- 50GB free disk space (for RAW + JPEG storage)
- Apple Silicon (M1/M2) or Intel Core i5+
File Sizes:
- Input ARW: ~24-42MB per file
- Output JPEG (92%): ~3-8MB per file
- Output JPEG (98%): ~6-12MB per file
Supported Cameras
All Sony cameras that produce ARW files, including:
- Sony Alpha A7 series (A7, A7R, A7S, A7III, A7IV, A7RV, etc.)
- Sony Alpha A9 series (A9, A9II, A9III)
- Sony Alpha A6000 series (A6000, A6100, A6400, A6600, etc.)
- Sony Alpha A1
- Sony ZV series
- Older Sony NEX cameras
Workflow Tips
For Best Results
- Shoot in RAW: Ensure your Sony camera is set to save ARW files
- Expose Properly: The preset works best with well-exposed images
- Organize First: Sort and cull images before batch processing
- Test First: Process 2-3 sample images to verify the look
- Adjust if Needed: Try the vibrant preset or adjust quality settings
When to Use Each Preset
Standard (mccurry-narrative.xmp):
- Portraits and people
- Indoor photography
- Balanced lighting conditions
- When you want subtlety
Vibrant (mccurry-vibrant.xmp):
- Landscapes and nature
- Outdoor travel photography
- Sunset/golden hour shots
- When you want maximum impact
Quality Settings Guide
| Quality | File Size | Use Case |
|---|---|---|
| 85-90% | 2-5MB | Web, social media, proofs |
| 90-94% | 3-8MB | General use, client delivery (recommended) |
| 95-100% | 6-15MB | Print, archival, portfolio |
Troubleshooting
darktable-cli not found
# Run installation script
bash scripts/install_darktable.sh
# Or create symlink manually
sudo ln -sf /Applications/darktable.app/Contents/MacOS/darktable-cli /usr/local/bin/darktable-cli
Processing very slow
- Close other applications to free up RAM
- Process in smaller batches (50 images at a time)
- Check Activity Monitor for CPU/memory usage
- Ensure sufficient disk space
Colors don't look right
- Try the alternate preset (vibrant vs. standard)
- Ensure input images are properly exposed
- Verify you're processing ARW files, not JPEGs
- Check your monitor color calibration
Some images fail to process
- Check if ARW files are corrupted (try opening in Sony's software)
- Ensure sufficient disk space
- Verify darktable version is up to date
- Run with
--verboseflag to see detailed errors
Advanced Usage
Creating Custom Presets
- Open darktable GUI application
- Import a sample Sony ARW file
- Adjust color grading modules to your taste
- Export the style as XMP
- Use your custom XMP with
--presetoption
Processing Specific Files
# Process only files matching a pattern
cd raw/
python ../scripts/batch_process.py \
--preset ../assets/mccurry-narrative.xmp \
DSC001*.ARW \
../output/
Integration with Other Tools
# Process then optimize for web
python scripts/batch_process.py raw/ output/ --quality 90
cd output/
mogrify -resize 2048x2048\> -quality 85 *.jpg
Reference Documentation
For more detailed information:
- references/mccurry-style-analysis.md - Deep dive into McCurry's visual style
- references/darktable-modules.md - Technical details of preset parameters
- references/troubleshooting.md - Extended troubleshooting guide
FAQ
Q: Can I use this with other camera brands? A: The presets work with any RAW format darktable supports (Canon CR2/CR3, Nikon NEF, etc.). Just change the file pattern in the script.
Q: Will this work with JPEG files? A: It will process JPEGs, but results are much better with RAW files since they contain more color data.
Q: Can I adjust the strength of the effect? A: Yes, edit the XMP preset in darktable GUI or create custom presets with different intensity.
Q: Does this delete my original RAW files? A: No, the original ARW files are never modified. Only JPEGs are created in the output directory.
Q: Can I undo the processing? A: The processing creates new JPEGs without modifying originals, so you can always re-process with different settings.
Credits
Color Grading Style: Inspired by Steve McCurry's iconic photography Processing Engine: darktable open-source RAW processor Skill Created For: Claude Code by Anthropic
License
This skill and its presets are provided as-is for personal and commercial use. darktable is licensed under GPL-3.0.
Version
Version: 1.0.0 Last Updated: January 2026 Compatible With: darktable 4.4+
Made with Claude Code - https://claude.com/claude-code