Building an Astro Imaging Quality Meter— Project Status
Project Status · Building an Astro Imaging Quality Meter
Where the AIQ project stands right now — updated as findings come in.
Forecasts lie — not maliciously, but enough that every suburban astrophotographer has lived the same moment. The forecast says average seeing; you can't find focus. It says clear transparency; it's a milky haze anyway. No single instrument answers the question that actually matters: is tonight worth it? Even the gold-standard Unihedron SQM-L only answers one part of that. It tells you how dark the sky is. It doesn't tell you whether seeing or transparency are helping or hurting, or whether you should actually set up. The AIQ was built to close that gap.
The AIQ Score
Darkness, seeing, and transparency combine into a single 0–100 composite score, something the SQM-L can't do on its own.
| Factor | Source | Weight | What it captures |
|---|---|---|---|
| Darkness | TSL2591 sensor (measured) | 0–40 pts |
Real-time sky brightness reading |
| Seeing | Forecast (manual input) | 0–30 pts |
Atmospheric steadiness |
| Transparency | Forecast (manual input) | 0–30 pts |
Atmospheric clarity — haze, thin clouds |
Below 30, don't set up. 30–34 is marginal — short sessions only. 35 and above, set up and image.
A Finding I Didn't Expect
A seven-point altitude survey, comparing the AIQ against the SQM-L from 30° to zenith, turned up something genuinely useful: your telescope sees the sky at 60°, not 90°, and darkness changes more than you'd think over that range.
Zenith is not the whole sky
0.75 mag — AIQ gradient, 30° → zenith
1.42 mag — SQM-L gradient, 30° → zenith
The SQM-L's wider light cone picks up more light-dome contamination away from zenith — another way the AIQ's hard aperture stop pays off in real-world use.
The Silicon Lottery
The TSL2591 is widely recommended for DIY sky quality meters — but nobody had published a completed, calibrated instrument using one. Screening 15 units revealed why: the chip is designed for smartphone ambient light sensing, not astronomical low-light capture, and manufacturing variance becomes visible right at the operating edge the AIQ depends on.
What first-time builders need to know
About 1 in 5 units is sensitive enough for true Bortle 5 dark-sky readings. Over half work well for Bortle 8/9 suburban skies. That's not a build mistake — it's a real, repeatable finding about the chip itself. Budget at least five breakout boards from different sources, and screen before you commit.
Two Units, Same Sky
Two independently built units, measured side by side on an optical calibration rail across seven altitude angles, agreed to within 0.1 mag — inside the SQM-L's own stated tolerance.
Spec Achieved
The nine original targets, current state:
| Target | Status | Note |
|---|---|---|
| Bortle range | Confirmed | 5–9, urban-optimized |
| Off-axis rejection | Confirmed | 0.75 vs 1.42 mag gradient |
| Sky brightness | Confirmed | mag/arcsec² + Bortle, live |
| Composite score | Confirmed | Thresholds still tuning |
| Environmental | Confirmed | Temp, humidity, dew margin |
| Cloud detection | In progress | Implemented, not field-validated |
| Sensor | Works, screened | ~20% Bortle 5 yield |
| Cost | ~$155 achieved | Under the ~$175 target |
| Openness | Confirmed | Field Guide, firmware, STL — live |
What's Next
- Longitudinal study correlating AIQ score with actual image quality, session by session
- Phase 2 weatherproofing and BNO055 tilt correction
- A stationary, field-deployed unit to bypass current WiFi range constraints
- Possible remote solar-powered station
Is This Useful?
The AIQ is designed for Bortle 5–9, limited mainly by TSL2591 low-light sensitivity — confirmed through indoor testing and field-validated at Bortle 9. Outdoor validation at Bortle 5 dark-sky sites is still ahead. Would you build one? Would you use a composite score? What would make this more useful to you?
Full parts list, calibration methodology, and build instructions live in Field Guide #3.
Clear skies / Pete // bortle9astro.com