A desktop application for astrophotographers to analyze and reject bad subframes based on quality metrics. Inspired by PixInsight's SubframeSelector.

No installation required! Download the latest release for your platform:

1. Download SubFrameSelector-macOS.zip from Releases
2. Extract the zip file
3. Right-click SubFrame Selector.app → Open (first time only, to bypass Gatekeeper)
4. Load your FITS folder and start analyzing!

1. Download SubFrameSelector-Windows.zip from Releases
2. Extract the zip file
3. Run SubFrame Selector.exe
4. Load your FITS folder and start analyzing!

• Load FITS files from one or multiple folders
• Parallel analysis using configurable CPU cores
• Quality metrics:
  • FWHM (star sharpness) - in pixels or arcseconds
  • Eccentricity (star roundness)
  • SNR (signal-to-noise ratio)
  • Star Count
  • Background level
• Interactive plot with sigma bands for outlier detection
• Click-to-select bad frames on the plot or file list
• Refresh to rescan folders after deletion (uses cached data)
• Safe deletion to Recycle Bin (recoverable)
• Tooltips on all controls for easy learning

1. Open Folder - Select a folder containing FITS files
2. + Add Folder - Optionally add more folders to analyze together
3. Analyze - Calculate quality metrics for all frames (adjust CPU cores as needed)
4. Select Metric - Choose which metric to display (FWHM, Eccentricity, etc.)
5. Identify Outliers - Points outside the gray sigma bands are potential bad frames
6. Click to Select - Click points on the plot or checkboxes in the file list
7. Delete Selected - Move bad frames to Recycle Bin
8. Refresh - Rescan folders and update plots with remaining files

The plot shows statistical boundaries using Median Absolute Deviation (MAD):

• Green line: Median value
• Dark gray band: ±1σ from median (~68% of frames)
• Light gray band: ±2σ from median (~95% of frames)
• Outside bands: Statistical outliers - candidates for rejection

The app reads camera/telescope metadata from FITS headers to calculate FWHM in arcseconds:

• Pixel Size: XPIXSZ, PIXSIZE
• Focal Length: FOCALLEN, FOCAL

Image scale formula: (pixel_size_μm / focal_length_mm) × 206.265 arcsec/pixel

# Clone the repository
git clone https://github.com/HiranD/Sub-Frame-Selector.git
cd Sub-Frame-Selector

# Install dependencies
pip install -r requirements.txt

# Run the application
python run.py

• Python 3.10+
• customtkinter
• matplotlib
• astropy
• photutils
• numpy
• scipy
• send2trash

# Install PyInstaller
pip install pyinstaller

# Build (creates dist/SubFrame Selector.app on macOS)
pyinstaller subframe-selector.spec --clean -y

sub-frame-selector/
├── src/
│   ├── analysis/       # Core analysis engine
│   │   ├── analyzer.py     # Parallel processing orchestrator
│   │   ├── fits_reader.py  # FITS file I/O
│   │   ├── star_detector.py # Star detection & PSF fitting
│   │   ├── metrics.py      # Quality metrics calculation
│   │   └── statistics.py   # MAD-based sigma bands
│   └── gui/            # User interface
│       ├── app.py          # Main application window
│       ├── toolbar.py      # Buttons and controls
│       ├── file_panel.py   # File list with checkboxes
│       └── plot_panel.py   # Interactive matplotlib plot
├── run.py              # Entry point
├── requirements.txt
└── subframe-selector.spec  # PyInstaller config

MIT