NETD, pixel pitch, 384 against 640, frame rate, detection range. What each number on a thermal spec sheet actually describes, and which of them should change your mind.
Specifications only mean something once you know the job. A unit that is excellent for scanning woodland may be the wrong buy for watching a field edge at 300 metres. Work out your usual distances, your ground and your budget first, then read the numbers against that.
One standout figure on a box does not make a good unit. Compare the whole thing.
NETD stands for Noise Equivalent Temperature Difference. It is a measure of how small a temperature difference the sensor can separate from its own electrical noise, written in millikelvin (mK). Lower is more sensitive. A figure of 30 mK corresponds to a difference of 0.03°C.
Read it as a laboratory measurement rather than a promise. "NETD <30 mK" describes behaviour under defined test conditions. It does not guarantee that every difference of that size will stand out on a damp evening. Where sensitivity earns its money is exactly where contrast is poor: an animal lying in wet grass, or a cold night after rain.
When you compare two figures, check three things. Whether the rating describes the bare sensor or the finished unit. Whether both were measured the same way. And how the lens and processing contribute, because they can undo a good sensor.
NETD is not the same as temperature measurement accuracy. A sensitive monocular is built to reveal contrast, not to tell you an animal is 37.2 degrees.
Pixel pitch is the distance between the centres of neighbouring sensor pixels, in micrometres. You will see 12 µm and 17 µm quoted. Smaller pitch allows a physically smaller sensor, and with the same lens each pixel covers a narrower slice of the scene.
It does not follow that 12 µm beats 17 µm. A 640x512 sensor has the same pixel count either way. Pitch is one input alongside resolution, lens design and sensitivity, not a verdict.
Resolution is the more straightforward number:
Nearly three times the information, for comparable optics and field of view. A 384 unit still does useful work finding heat at sensible ranges. A 640 gives you more to examine when you are trying to work out what something is.
Check sensor resolution separately from display resolution. A sharp screen cannot invent detail the sensor never recorded.
Digital zoom enlarges a crop of the image you already have. On a 640x512 sensor, a 2x digital zoom is showing roughly a 320x256 patch stretched across the display. It looks closer. It is not more detailed. Maximum advertised magnification is worth very little unless the image holds together at it.
The lens does the real work. For a given sensor, a shorter focal length widens the field of view, and a longer one narrows it while putting more pixels across a distant subject. Wide helps when you are scanning woodland or walking a hedge line. Narrow helps when you are studying something across a field, at the cost of seeing what is around it.
Compare the quoted field of view, in degrees or metres at 100 metres. That tells you more than the diameter of the objective.
Frame rate, in hertz, is how often the image refreshes. A 50 Hz unit updates fifty times a second, which keeps movement smooth while you pan or follow a running fox. Lower rates look fragmented when anything moves.
Frame rate is not the same as lag. Sensor response and processing both affect how quickly what you see catches up with what is happening. If you can, swing a unit across a scene before you buy and watch for smear and delay.
A quoted range tells you how far away the unit may register that something is there, under conditions the manufacturer chose. It does not tell you that you could say what it is.
These distances move with subject size, weather, thermal contrast and whatever criteria the maker used. Buy around the level of detail you need at the distances you actually shoot. A heat blob at 900 metres and a identifiable deer at 900 metres are very different claims.
This matters beyond the spec sheet. You cannot lawfully take a shot at something you have not positively identified, and a thermal image is not identification on its own.
Image quality is the headline. Everyday usability is what determines whether the unit lives in your bag or in a drawer.
Where you can, compare shortlisted units outdoors on the same scene, at similar magnification, including something with poor contrast. That ten minutes tells you more than a fortnight of reading specifications.
NETD is Noise Equivalent Temperature Difference, measured in millikelvin. It describes how small a temperature difference the sensor can pick out from its own image noise. A lower number means more sensitivity, so 30 mK can separate a difference of 0.03°C under test conditions. It is measured in a laboratory, not promised in a wet field at dusk, and the lens and image processing matter as much as the figure.
Not on its own. Pixel pitch is the distance between neighbouring sensor pixels. A smaller pitch allows a more compact sensor, and with the same lens each pixel covers less of the scene. But a 640x512 sensor has the same number of pixels at either pitch, so pitch has to be read alongside resolution, lens and sensitivity rather than used as a score.
A 384x288 sensor captures 110,592 thermal pixels. A 640x512 captures 327,680, nearly three times as many. With comparable optics that means finer detail. A 384 unit is still useful for finding heat at sensible distances; 640 gives you more to work with when you need to make out what you are looking at.
No. Digital zoom enlarges part of the image the sensor already captured. A 2x digital zoom on a 640x512 sensor is showing you roughly a 320x256 area blown up to fill the screen. Processing can tidy its appearance, but no new detail arrives. Judge a unit at base magnification and at the zoom you will actually use.
It works without any visible light, but weather still degrades it. Rain, fog and atmospheric moisture cut contrast and useful range. Better sensitivity helps at the margins but does not remove the limit. Thermal also does not see through solid cover: you may pick up an animal through gaps in a hedge while branches hide the rest.
If you are still deciding between the two technologies rather than comparing units, start with thermal or night vision: which should you buy. For the night vision side of the specifications, read choosing night vision. For what you are allowed to do with either, our foxing kit guide covers the law on identification and lamping.
Last reviewed 11 November 2026.