Infrared optics follow the same geometric principles as visible optics, but the material constraints are radically different. Germanium, silicon, chalcogenide glass, and zinc selenide are the workhorses of infrared optics — none of which you’d find in a camera store. Understanding the optics side of your thermal imaging module is essential for making informed system-level decisions.
Focal Length and the IFOV
The instantaneous field of view (IFOV) is the solid angle subtended by a single detector pixel. It’s the fundamental angular resolution unit of your imaging system:
IFOV (radians) = pixel pitch (μm) / focal length (mm) × 10⁻³
For a 640×512 module with 12 μm pixel pitch and a 25 mm lens:
IFOV = 12 × 10⁻⁶ / 0.025 = 0.48 mrad
This means each pixel sees a 0.48 mrad slice of the world. At 1000 m range, one pixel covers 0.48 m. You need at least 2 pixels on a target for reliable detection, so detection range for a 1 m target:
R_detect = target_size / (IFOV × min_pixels) = 1.0 / (0.00048 × 2) = 1041 m
Total Field of View
The horizontal and vertical FOV depend on both the focal length and the array format:
HFOV = 2 × arctan(detector_width / (2 × focal_length))
For a 640×512, 12 μm pitch detector (7.68 mm × 6.14 mm) with a 25 mm lens:
HFOV = 2 × arctan(7.68 / 50) = 17.6°
VFOV = 2 × arctan(6.14 / 50) = 14.1°
This 17.6° × 14.1° FOV covers a ground footprint of 307 m × 246 m at 1000 m altitude. For wide-area surveillance, you might want 35 mm or even 50 mm on a 1280×1024 detector to maintain resolution while covering more ground.
The F-number Trade-off
The F-number (or f-stop) is the ratio of focal length to entrance pupil diameter:
F/# = focal length / aperture diameter
A faster lens (lower F/#) collects more light — critical for infrared imaging where photon flux is the fundamental sensitivity limit. However:
| F/# | Light Collection | Aberration Challenge | Cost |
|---|---|---|---|
| F/1.0 | Baseline (maximum) | High | High |
| F/1.2 | 44% less | Moderate | Moderate |
| F/1.6 | 64% less | Lower | Lower |
| F/2.0 | 75% less | Low | Low |
For uncooled LWIR where NETD is already limited, F/1.0 lenses are common despite their cost. For cooled MWIR where detector sensitivity is excellent, F/2.0 or F/4.0 is often acceptable, enabling smaller aperture telescopes at long focal lengths.
Athermalization: The Infrared-Specific Challenge
Thermal expansion causes all optical materials to change refractive index with temperature — dn/dT. In visible optics, this effect is small enough to ignore for most applications. In infrared optics, it’s dominant.
An un-athermalized germanium LWIR lens will shift focal point by tens of micrometers over a 60°C temperature range. For a fixed-focus system, this shows up as a blurry image at temperature extremes.
Three approaches to athermalization:
- Active focus control: Motor-driven focus that compensates for temperature — works well but adds complexity and power consumption
- Passive athermalization: Design the lens barrel from materials with complementary thermal expansion coefficients (e.g., aluminum barrel + germanium lens) — the barrel expands as the lens focal length changes, maintaining focus automatically
- Diffractive elements: Add a diffractive optical element (DOE) that provides negative dn/dT to offset the positive dn/dT of the refractive elements
IRmodules integrates passively athermalized lenses with all SPECTRA series modules as standard. The lens and barrel assembly is tuned for the specified operating temperature range (typically -40°C to +70°C) without requiring active focus control.
Choosing Focal Length: A Quick Reference
| Application | Typical Focal Length | Format | Approximate HFOV |
|---|---|---|---|
| Wide-area UAV survey | 9–13 mm | 640×512 | 35–50° |
| Standard UAV ISR | 19–25 mm | 640×512 | 17–23° |
| Medium-range ground station | 50–75 mm | 640×512 | 6–9° |
| Long-range surveillance | 100–200 mm | 1280×1024 | 4–8° |
| Vehicle/weapon sight | 35–50 mm | 640×512 | 9–13° |
The focal length determines everything downstream: FOV, range performance, lens size, and cost. Get this decision right early, and the rest of the optical integration becomes manageable. Get it wrong, and you’re redesigning the housing halfway through production.
