I've lost count of how many times someone has asked: "So, can this thing see through walls?" They hold up a thermal camera — a handheld, a drone module — and expect me to say yes. Hollywood made thermal vision look like X-ray with an orange filter. Real life? Not even close.

Here are five of the most stubborn thermal imaging myths I run into — the ones that keep surfacing in emails, Reddit, and YouTube comments. Each one gets a straight answer plus the physics behind it. If you're evaluating a module for a project, clearing these up now saves you from buying the wrong hardware.

How Thermal Imaging Works

Every object above absolute zero emits infrared. Hotter = more infrared. A thermal camera's microbolometer detector captures LWIR (long-wave infrared, 8–14 μm) and maps it to an image: bright for hot, dark for cold.

The one thing to remember: a thermal camera reads surface temperature only. It doesn't penetrate. It doesn't see through. It detects infrared coming off the outermost layer. That's it. Every myth below traces back to forgetting this.

Myth #1 — "Thermal Cameras See Through Walls"

No. Drywall, brick, concrete, wood — all opaque to LWIR. Infrared hits a wall and gets absorbed or reflected. Nothing passes through.

If infrared could pass through walls, your thermal image would be a washed-out mess of noise from outside, the sun, and everything else. The clean wall-surface reading you get proves the wall is blocking what's behind it.

What you can detect

A hot pipe inside a wall heats the drywall from within. Heat conducts to the surface. The camera picks up a warm stripe — not an image of the pipe, just a temperature anomaly on the surface. Same for cold spots (moisture) or hot spots (electrical faults). This is surface temperature mapping, not seeing through.

Myth #2 — "Thermal Cameras See Through Glass"

No. Glass is a mirror to LWIR. Point a thermal camera at a window and you see your own reflection, not what's outside. Silicon-oxygen bonds absorb infrared instead of transmitting it. Zero transmission.

This is exactly why thermal lenses use germanium or zinc selenide, not regular glass. Those materials are transparent to LWIR; silica glass is not.

Myth #3 — "Thermal Can't See Through Fog or Smoke"

Wrong — thermal does better than visible light here.

Visible light (~400–700 nm) scatters heavily off fog and smoke particles. LWIR (8,000–14,000 nm) waves are about 20× longer — they slip past small particles with far less scattering. Firefighters have used thermal imaging for decades because of this. It's not magic, just longer wavelengths.

Myth #4 — "Thermal Measures Temperature Exactly"

Someone reads "127.3°C" off the screen and writes it down as truth. Then another camera says 124.1°C. Now they're confused.

A thermal camera estimates temperature. It's a radiometric instrument, not a thermometer. Three things throw it off:

① Emissivity (ε). Human skin: ~0.97. Polished aluminum: ~0.1–0.2 (reflects most IR). Wrong ε setting = garbage reading. Always check an emissivity table for your target material.

② NETD is sensitivity, not accuracy. 25mK NETD means it can detect a 0.025°C difference — it doesn't mean readings are accurate to 0.025°C. Mixing up resolution and precision leads to bad data.

③ Distance and environment. At 20 meters, atmospheric absorption eats your signal. Humidity and ambient temperature matter too. Budget modules rarely compensate. High-end radiometric cameras do.

Myth #5 — "Thermal Only Works at Night"

Thermal works 24/7. It reads emitted infrared, not reflected visible light. Sun up or down makes no difference to the sensor.

The real issue is thermal crossover. Sunlight heats surfaces unevenly: a rock at 50°C, grass at 25°C, a person at 35°C. When background temps creep into body-temperature range, contrast collapses. Worst around sunrise and sunset. A target visible at 4 AM can vanish by 6 AM — not because the camera failed, but because the temperature gap closed.

Bonus — "Can Thermal See Through Clothes?"

No. Fabric blocks infrared. The camera sees the cloth surface temperature, nothing more. A warm phone in a pocket makes a faint heat bloom on the fabric — you're not seeing through to skin. This myth is not just wrong; it's harmful. It scares people away from useful technology based on something that doesn't exist.

All 5 Myths at a Glance

Choosing a Thermal Camera: What Actually Matters

• Resolution: 640×512 = 4× the detail of 320×256. Go higher for anything beyond basic hotspot detection.
• NETD: 25mK vs 40mK matters for subtle differences — moisture, early electrical faults.
• Output: PAL = near-zero latency (FPV drones). UVC = plug-and-play (PC/phone). Some modules do both.
• Lens FOV: Wide for awareness, narrow for range. Pick based on use case.
• Radiometry: Temperature data or just imagery? Radiometric costs more; non-radiometric is fine for hotspot-only work.

The market shifted fast. Modules that cost thousands five years ago are now in the hundreds. The physics hasn't moved an inch — which is why these five myths still need killing.

Frequently Asked Questions