NETD or IFOV: which one actually matters for your job

These are the two numbers that determine what a thermal camera can actually do. Almost no one explains them, and they often pull in opposite directions.

NETD, or the ability to tell two close temperatures apart

NETD is expressed in millikelvin. A NETD of 40 mK means the sensor can distinguish two surfaces whose temperatures differ by 0.04°C. Below that gap, the difference is drowned out by the detector's noise.

The smaller the number, the better the sensitivity.

A 0.08°C gap on a wall, seen by three sensors

Same scene, same resolution: only the detector's noise changes.

One essential caveat: a NETD without its measurement conditions means nothing. The value depends on the scene's temperature and the lens's aperture. Serious manufacturers write "< 40 mK at 25°C, F/1.0". Those who just advertise "< 40 mK" on its own allow no comparison at all.

A detail we found on three devices in the database: the HIKMICRO Pocket2, the InfiRay P2 Pro, and the HIKMICRO E1L advertise a NETD specified at F/1.0 while their lens actually opens to F/1.1. Real-world sensitivity is therefore lower than the displayed value.

IFOV, or the ability to see small

IFOV is expressed in milliradians and refers to the angle covered by a single pixel. It's calculated by dividing the field of view by the sensor's number of columns, which makes it possible to check the value a manufacturer advertises.

Its practical translation is immediate: multiply the IFOV by the distance in meters, and you get the size of the smallest visible detail in millimeters.

At five meters, a HIKMICRO G40 at 0.68 mrad can distinguish a 3.4 mm object. A Bosch GTC 400 C at 5.78 mrad can't go below 29 mm. On an electrical panel viewed from a safe distance, the first reads a terminal block, the second sees a blur.

And be careful: seeing an object doesn't mean you can measure it. A reliable reading requires the target to cover several pixels — in practice, three times the IFOV.

Why the two criteria pull against each other

For a given sensor, a narrow field of view improves IFOV and shrinks the covered area. A wide field does the opposite. Increasing the pixel count improves IFOV but reduces each detector's surface area, and therefore the amount of radiation captured, and therefore sensitivity.

Two devices in our database illustrate this trade-off almost to caricature.

The Fluke, among the most expensive devices on the market, has the worst NETD of all professional models in our database. That's not a flaw: it's a choice, consistent with its market — industrial electrical maintenance, where you're looking for a clear hot spot on a small component.

Which one to prioritize based on your job

Leak detection, moisture, thermal bridges: NETD dominates. You're looking for gaps of a few tenths of a degree over large areas. A cold wall behind an insulation defect, rising damp, a heated floor. Aim for 40 mK or better. IFOV barely matters, since you're two meters from the wall.

Electrical work, panels, switchgear: IFOV dominates. You're observing small components from a safe distance. The gaps are stark, often dozens of degrees. Aim for 2 mrad or better, and check the spot size at your working distance.

Full building diagnostics: both, with priority to NETD. You alternate between facades, window frames, and panels.

Electronics repair: neither one, or almost. You're three centimeters from a board. What matters is having a macro lens and the smallest area you can actually measure.

What our comparator does

Every device gets a score per use case, with published weightings. NETD carries 50% of the sensor family for leak detection, versus 25% for solar work. IFOV carries 65% of the optics family for solar work, versus 5% for electronics repair.

And you can move the sliders yourself if your job doesn't fit neatly into any category. Open the comparator.

All values cited come from manufacturer datasheets and the sources listed on each product page. No data is estimated.