Equipment Needed:
- DSLR or Mirrorless Camera: Any modern camera with 20+ megapixels (Canon, Nikon, Sony, etc.) Higher resolution is better, but even a good smartphone camera can work in a pinch. Fixed focal length (prime) lenses are preferred for consistency.
- Tripod: For stable, consistent images.
- Reference Markers: Printed QR codes or ArUco markers. Reflective targets (can be made with reflective tape on cardboard). Tennis balls or ping pong balls (bright colors work well).
- Measuring Tools: Laser distance meter for scale reference. Long tape measure. Spirit level
- Software: Agisoft Metashape ($179 for Standard edition) OR free alternatives like Meshroom, COLMAP, or even WebODM
Process:
- Prepare the Dish: Place reference markers in a grid pattern across the dish surface. For an 18m dish, aim for markers every 1-2 meters. Include some markers at known distances for scale reference.
- Capture Photos: Take photos from multiple positions around the dish, high and low. Ensure 60-80% overlap between adjacent images. Capture both wide shots and detail shots. For best results, take like 100+ photos of the entire structure. Take photos under consistent lighting (overcast days are best).
- Process the Images: Import photos into photogrammetry software. Align photos to create a sparse point cloud. Generate a dense point cloud. Create a mesh and texture if desired. Will be very nice in presentations! Scale the model using your known reference distances.
- Analysis: Export the point cloud as a CSV or PLY file. Use MATLAB, Python (with NumPy/SciPy), or specialized surface analysis software and use appropriate fitting functions to “fit” the point cloud to an ideal parabola. Calculate root mean square (RMS) error from the ideal surface. Generate contour maps of deviation. More great images for presentations!
Special Considerations for Radio Dishes:
- Dish Position: Ideally position the dish at zenith or a consistent angle for the full survey.
- Reference Frame: Establish the feed support point and dish center as key reference points!
Example Workday Plan:
- Morning Session: Set up markers across the dish (2-3 hours)
- Midday Session: Photograph dish from multiple angles (2-3 hours)
- Afternoon/Evening: Process initial results on-site to verify quality (1-2 hours) don’t leave until the software workflow is confirmed!
- Follow-up Analysis: Detailed surface error mapping and calculations (1-2 days or more)
This approach has been successfully used by amateur radio astronomy groups and small observatories to characterize dishes of similar size to DSES’s 18m antenna. The resulting data can directly inform your noise temperature calculations and help verify the antenna efficiency value (I believe 69%) we are currently using in our models.
For DSES specifically, this could be implemented as a weekend workshop activity that would not only produce valuable technical data but also serve as an educational opportunity for members.
Do you see a way to improve this proposed process? Comment and critique welcome and encouraged.