Best thermal cameras for home inspection

What the camera needs to reveal in a building

A thermal camera does not see “through” walls: it measures the infrared radiation emitted by the surface it is aimed at. In a building, it therefore reveals the consequences of hidden defects, as differences in surface temperature. The most common targets are thermal bridges (floor-to-wall junctions, window reveals, lintels, roller-shutter boxes), missing or settled insulation in lofts and partitions, air leaks around windows, sockets and service penetrations, and the layout of underfloor heating or heating pipes.

These differences are often small: a few tenths of a degree to two or three degrees between a sound area and a faulty one. That is why, in this guide, thermal sensitivity (NETD) weighs more than pixel count. A very detailed but low-contrast image will show an insulation defect less clearly than a more modest image that can separate small differences.

Getting the measurement conditions right

The quality of a building survey depends as much on conditions as on the device. An insulation defect only shows up if heat is actually flowing through the wall. A difference of at least 10 °C between inside and outside, maintained for several hours, is a common recommendation, which is why the heating season is the best time. The qualitative infrared method is described in EN 13187, which details these precautions.

• Avoid sunshine: a sunlit façade stores heat for hours and hides defects. Outdoors, prefer the end of the night or early morning under an overcast sky.
• Avoid rain and strong wind: evaporating water cools surfaces and wind evens out temperatures.
• Work from the inside for thermal bridges: ventilated cladding or external wall insulation hides much of what happens inside the wall.
• Depressurize the home when looking for air leaks: a blower-door test makes infiltration appear as cold streaks.

To go further, read when to scan a house with a thermal camera so the images are useful.

Distance, target size and resolution

The typical situation used by the selector is an area of about 10 cm whose temperature you want, seen from 3 m: roughly the size of a thermal bridge in a room corner or an insulation gap in a ceiling. What matters is not the advertised resolution but the surface area one pixel covers at that distance, calculated from the IFOV.

Example: with a 256 × 192 sensor and a 50° field of view, the IFOV is about 3.4 mrad. At 3 m, one pixel covers about 1 cm. For a reliable temperature we apply a five-pixel margin, so an area of about 5 cm: the 10 cm target is comfortably covered. With an 80 × 60 sensor and a field of view of about 36°, the IFOV is about 7.8 mrad: one pixel covers 2.3 cm and the measurable area exceeds 11 cm, so the device is at its limit. Outdoors, 10 m from a façade, those figures more than triple. For façade surveys, a 320 × 240 sensor or better quickly becomes necessary. The reasoning is explained in NETD vs IFOV: which one really matters.

What matters, and what matters less

• Sensitivity (NETD): 50 mK is fine; 40 mK or lower clearly helps with small differences.
• Emissivity and reflected temperature settings: essential for figures you can rely on. Without them, the camera shows contrast, but the displayed values remain approximate.
• Radiometric images and reporting software: they let you revisit measurements at the office and hand a document to a client or a contractor.
• A reasonably wide field of view: indoors, you cannot step back very far. A narrow field means more images to cover a wall.
• What matters less: frame rate (a building does not move, 9 Hz is enough), high temperature range (you stay between −20 and +100 °C) and “displayed” resolution produced by processing, which adds no measured information.

Common pitfalls

Glass is opaque to thermal infrared: a camera aimed at a window sees a reflection of whatever is opposite, often the cold sky or the operator, not the room behind. Shiny metal surfaces (aluminium frames, ducts) also reflect their surroundings and show false temperatures. A cold patch is not always a defect: furniture against a wall, a passing draught or a damp area produce similar images. And a single image is hard to read: always compare a suspect area with a sound area of the same kind. To understand how the surface affects the reading, see why emissivity skews temperatures.

Phone attachment or standalone camera?

To check your own home, a good smartphone module is often enough. The best ones offer sensitivity and sensors equivalent to mid-range standalone cameras. They do, however, depend on the phone, its connector and the maker’s app, which does not always offer proper reporting. For professional use (energy assessors, contractors, engineering firms), a standalone camera is more comfortable: instant handling, full measurement settings, reporting software and ruggedness. That is why both formats are ranked separately above. The detailed comparison is in phone attachment vs standalone thermal camera.

What budget should you plan?

For an occasional check at home, a device in the €€ band (€200 to €500) with a native sensor of at least 160 × 120 already locates the main defects. For regular use, with usable measurements and reports, aim for the €€€ to €€€€ bands (€500 to €2,000): a 256 × 192 to 320 × 240 sensor, 40 to 50 mK NETD and full settings. €€€€€ devices make sense for intensive professional work, long-distance façade surveys or interchangeable lenses. These price markers are explained on the price ranges page.

Frequently asked questions

Do I need a thermal camera for an energy performance certificate?

No. Thermography is not part of the regulatory method for energy performance certificates such as the French DPE. It remains a useful complement for locating defects the certificate does not show, for example before insulation work.

Can I survey a building in summer?

It is difficult for insulation, because the difference between inside and outside is too small. An air-conditioned home can provide one, in the opposite direction. Air-leak detection with depressurization is possible all year round, but the image is clearer with a marked temperature difference.

Should I measure from inside or outside?

Inside is usually more reliable for thermal bridges and insulation, because you aim directly at the wall in question and avoid wind and sky. Outside gives a quick overview but is more sensitive to conditions and façade finishes.

Is a 160 × 120 sensor enough?

For a home and measurements at a few metres, often yes, provided sensitivity is good. It becomes limiting for large rooms, façades or small details, as the resolution calculation above shows. The selector runs that calculation for your distance and target.