📷 Camera IR Lab

Explore what wavelengths your phone's camera can detect. Test if it can see near-infrared. Honest answer about thermal imaging at the bottom.

Pixel at reticle

—

— · tap the video to move the reticle

Full-frame histogram

Red/green/blue channel distribution across the whole image. A pure NIR source (TV remote LED in dark room) creates a small spike in the red channel only.

Experiments to try

The TV remote test. Point any TV/AC/streaming-box remote at the camera lens from 5–10 cm. Press a button. You should see a bright glow flash on the LED. This is 940 nm near-infrared light — completely invisible to your eyes, but your camera sees it.

Front vs rear camera. Repeat experiment 01 on the front (selfie) camera with the Swap button. It's usually brighter on the front camera because the IR-cut filter is weaker (face-unlock systems need to see IR).

Switch to "IR-emphasis" mode. Now repeat. The remote LED looks white/pink against a darkened background. Visible-spectrum reds in the room get partially suppressed because their G and B channels are also lit.

The sun and incandescent bulbs. Look at sunlight or an old-style filament bulb in IR-emphasis mode. They have huge IR content; LEDs and fluorescents do not. This is how you can spot incandescent vs LED lighting through a camera.

The thermal test (negative result). Point the camera at your hand, then a cold glass of water, then a hot mug of coffee. They all look the same. The camera cannot see thermal IR (8–14 µm). This proves cameras don't measure temperature.

Dark fabrics and inks. Some black inks and dark fabrics are transparent to NIR. View dark plastic, black headphones, or printed text in IR-emphasis mode — the NIR sometimes passes through and reveals what's underneath.

The honest physics: what your camera can and can't detect

Silicon's sensitivity range. The CMOS image sensor in your phone is made of silicon, which detects photons from roughly 400 nm (violet) to 1100 nm (near-infrared). The peak is around 600 nm (yellow-green). It's most sensitive to wavelengths your eyes don't even see strongly.

The IR-cut filter. Phone cameras include a tiny glass filter glued over the sensor that blocks most light above ~650 nm. Without it, photos would look pink and washed out — IR leakage corrupts color accuracy. But these filters are not perfect: enough NIR leaks through that bright IR sources (remotes, sunlight, incandescent bulbs) are still visible. The front camera's filter is typically weaker because phones use NIR for proximity sensing, focus assist, and face unlock.

Why this isn't thermal imaging. Objects at room temperature emit electromagnetic radiation, but at wavelengths around 10 µm — that's ten thousand nanometers, ten times longer than silicon can detect. Thermal imaging requires a different sensor entirely (a microbolometer, which detects heat by measuring tiny temperature changes in a vanadium oxide or amorphous silicon array). FLIR-style cameras cost $200+ even for the cheapest models because that sensor doesn't fit on a silicon wafer.

What you can deduce. Comparing different light sources in IR-emphasis mode tells you about their spectral content (incandescent vs LED vs sun). Comparing front and rear cameras tells you about your phone's filter strength. Examining how much NIR leaks through different materials tells you about their transmittance — useful, occasionally, for material science demonstrations. But no measurement on this page corresponds to temperature.

The IR-emphasis math

The "IR-emphasis" view applies output = max(0, R − (G+B)/2) × 4 to every pixel. A pure NIR source (which leaks roughly equally into R, G, B but slightly stronger in R) produces a higher R−(G+B)/2 than a visible red object (whose green and blue components are smaller relative to red). The ×4 amplification makes faint IR sources visible. It's a crude filter — visible reds still show somewhat — but it's clear enough to spot a TV remote LED in a brightly lit room.