What Is NETD?

NETD (Noise Equivalent Temperature Difference, measured in mK / millikelvin) is defined as the target-background temperature difference that produces an output signal equal to the RMS temporal noise of the camera pixel — the core metric for thermal sensitivity. Standard measurement follows the ASTM E1543 and VDI/VDE 5585 industrial standards, relying on a precision temperature-controlled blackbody radiation source as a uniform IR reference target.

1. Required Test Hardware

• High-stability surface blackbody (emissivity ≥0.98, temperature stability ≤10 mK RMS, adjustable precise temperature)

• Dark shielded test chamber (blocks ambient stray infrared radiation)

• Tested thermal camera, configured with fixed integration time, frame rate, and F-number per specification; complete NUC (non-uniformity correction) before testing

• Frame-grabber / SDK to export raw pixel AD count data (unprocessed original digital signal)

2. Standard Test Method: Dual-Temperature Responsivity & Temporal Noise Method

Step 1: Configure Camera Parameters

• Uncooled LWIR microbolometer: maximum integration time, with standard fixed lens

• Cooled MWIR detector: integration time set to half-well dynamic range, lens removed for bare FPA (focal plane array) testing

• Complete a 2-point non-uniformity correction first to eliminate fixed pattern noise (FPN)

Step 2: Capture Three Groups of Raw Frame Data

• Low temperature T₁ (e.g., 20 °C): collect ≥64 continuous frames, average all frames pixel-by-pixel → signal array S₁ (AD counts at T₁)

• Mid fixed temperature T₀ (e.g., 25 °C): collect ≥128 sequential frames, take the standard deviation of each pixel's values across all frames, and place these standard deviations into an array representing temporal noise (AD counts at σₙ)

• High temperature T₂ (e.g., 30 °C): collect ≥64 continuous frames, average frames → signal array S₂ (AD counts at T₂)

Step 3: Calculate Responsivity and Final NETD

• Pixel responsivity: R = (S₂ − S₁) / (T₂ − T₁), in counts/°C

• Single-pixel NETD: NETD_pixel = σₙ / R, in Kelvin

• Final camera NETD: the average of all valid pixel NETD values (excluding dead/bad pixels), multiplied by 1000 to give the figure of merit in milliKelvin (mK)

3. Critical Test Conditions Affecting the Final NETD Result

• Specified test temperature point (commonly NETD @30°C ambient)

• Camera integration time, F-number (aperture), and frame rate

• Ambient temperature and humidity inside the test enclosure

• Spectral band under test (LWIR 8–14 μm / MWIR 3–5 μm)

4. Typical NETD Benchmark Values

FAQ

Q: What standards govern NETD measurement?

NETD measurement typically follows the ASTM E1543 and VDI/VDE 5585 industrial standards, using a precision temperature-controlled blackbody source as the reference target.

Q: Why does NETD testing use three temperature points?

Two temperature points (low and high) establish the pixel responsivity — the signal change per degree — while a third mid-point temperature is used to measure temporal noise. Combining responsivity and noise yields the NETD value.

Q: What NETD value should a procurement team expect from an industrial-grade uncooled camera?

High-grade uncooled industrial bolometers typically achieve NETD below 35 mK, compared with 40–80 mK for consumer-grade uncooled cameras and below 20 mK for cryogenically cooled photon detectors.

Q: Does F-number affect NETD test results?

Yes. The camera's F-number (aperture), along with integration time and frame rate, is a critical test condition that must be fixed per specification to obtain a valid, comparable NETD measurement.

Future Vision Technology is a China-based manufacturer with 18 years of experience in thermal cameras, supporting global partners through OEM/ODM services. View Thermal Camera Range