Thermal cameras
A thermal camera listing usually carries two resolutions and only mentions one of them loudly. FLIR states 1440 × 1080 image and 80 × 60 thermal in the same paragraph; TOPDON states 320 × 240 derived from 160 × 120. The detector figure is the instrument; the other is a picture.
Three cameras, and the ranking follows the detector rather than the headline.
The Bosch GTC 12V-450-13 quotes 256 × 192 px and nothing else — no separate upscaled image number, which reads as a native detector specification. At 49,152 measured points it is two and a half times the TOPDON detector and ten times the FLIR ONE. It is also the only camera here with an ingress rating and a drop rating, and the only one that publishes no accuracy figure.
The TOPDON TC004 Mini states 320 × 240 from 160 × 120 — upscaling, published as upscaling. It states ± 2 °C (36 °F), the tightest accuracy in the hub, a −20 to 550 °C (-4 to 1022 °F) range, Wi-Fi, PC analysis and 14 hours of runtime. For work that has to end in a report rather than an impression, it is the practical choice.
The FLIR ONE Gen 3 states 1440 × 1080 image and 80 × 60 thermal in the same sentence, and gives its accuracy with its conditions: ± 3 °C (37 °F) or ± 5% with the device between 15 and 35 °C (95 °F). It plugs into a phone. Four thousand eight hundred measured points is enough for the building-fabric problems most people buy a thermal camera for.
Read the smaller number. It is the instrument.
The commonest domestic use, and one that works with a modest sensor because the differences are large.
Where a thermal image shows layout clearly, which is genuinely useful before drilling or lifting.
Where a hot connection stands out, and where accuracy and interpretation both matter.
Evaporation cools a surface, so damp can show as a cool patch. It is an indication, not a measurement of water.
| Your use | What you need | What is oversold |
|---|---|---|
| Finding drafts | A modest sensor | High resolution |
| Pipe runs and underfloor | Any thermal camera | Temperature accuracy |
| Electrical inspection | Resolution and accuracy | Enhanced pixel counts |
| Any use | Understanding emissivity | The headline number |
A thermal camera is a delicate sensor behind a lens made of an unusual material. The lens should never be cleaned like glass, and the camera needs time to stabilize when moved between temperatures before its readings settle.
The camera itself is harmless; what it looks at often is not.
256 × 192 px · −20 to +450 °C (842 °F) · IP54 · 2 m (7 ft) drop · dual power
A camera that quotes 256 × 192 with no upscaled figure beside it, in an IP54 body rated for a 2 m (7 ft) drop, running on a 12 V Li-ion pack or AA cells.
Best for: Site use, where the camera shares a bag with other tools and has to work on the first press.
160 × 120 detector, 320 × 240 displayed · −20 to 550 °C (-4 to 1022 °F) · ± 2 °C (36 °F) stated · Wi-Fi
A standalone thermal camera stating a 320 × 240 output produced by upscaling a 160 × 120 detector, with ± 2 °C (36 °F) accuracy, a −20 to 550 °C (-4 to 1022 °F) range and Wi-Fi to a PC.
Best for: Regular inspection work where images have to be reviewed and reported afterwards rather than glanced at.
80 × 60 detector · 1440 × 1080 image · ± 3 °C (37 °F) or ± 5% stated · USB-C Android
A phone-attached thermal camera whose listing states both numbers plainly: an 80 × 60 thermal detector behind a 1440 × 1080 displayed image, with accuracy given as ± 3 °C (37 °F) or ± 5%.
Best for: A first thermal camera for home inspection, where finding the cold spot matters more than resolving it finely.
≈ $214
The detector figure is the smaller one, and it is usually the one qualified by the words "thermal", "IR" or "from". If a listing gives only one number and it is large, assume it is the image and look for the detector figure in the technical data.
Large, high-contrast features: a missing insulation batt, a cold bridge, an overheating connection in an open box, a blocked radiator. What it will not do is resolve one small component among several in a crowded panel.
It finds temperature differences, and evaporating moisture is often cooler — which is a hint, not a diagnosis. It does not see through walls, and it does not replace a moisture meter or a qualified assessment.
For drafts and insulation, surprisingly little, because the differences are large. For electrical inspection and small targets, true detector resolution starts to matter.
Real resolution is the detector's pixel count; enhanced is an upscaled image. Only the first determines what the camera can actually resolve.
It reflects rather than radiates, which is emissivity. Without accounting for it, the temperature reading on any polished surface is wrong.
It can show cool patches where evaporation is happening, which suggests where to look. It does not detect water and it does not confirm a leak.
A great deal. A small target seen from far away occupies too few pixels to be measured correctly, and the reading averages in its surroundings.
For finding subtle differences, often yes. Resolution decides what you can see; sensitivity decides how small a difference is visible at all.
Suited to: Site use, where the camera shares a bag with other tools and has to work on the first press.
Best figure: Quotes one resolution — 256 × 192 px — with no upscaled headline figure alongside it
Worst figure: No accuracy figure published, where both cheaper cameras state one
Suited to: Regular inspection work where images have to be reviewed and reported afterwards rather than glanced at.
Best figure: States both numbers: 320 × 240 output derived from a 160 × 120 detector
Worst figure: The 320 × 240 headline is upscaled — the detector is 160 × 120
Suited to: A first thermal camera for home inspection, where finding the cold spot matters more than resolving it finely.
Best figure: States the detector resolution — 80 × 60 — alongside the image resolution, which is honest and rare
Worst figure: An 80 × 60 detector is the lowest here — 4,800 real measured points
Targets, tripods and spares from the same aisle, at the price we noted. Check the thread matches the instrument you pick before ordering: it depends on the mount and the part number, which the listing states.
Instrument against instrument, where we have already published the comparison.
The instruments above come from the catalog of this site. Specifications are the ones the manufacturer states and the ones on the Amazon listing; prices are the Keepa reading for the date below, and both change. Ranking uses declared specifications, stated strengths weighed against stated limits, and price. Nothing here has had its accuracy checked by us, and the page says as much.
Last checked on
Prices quoted in these summaries are dated readings. The live price is shown next to each product.
Thermal cameras, line lasers and clamp meters swing by thirty dollars or more across the year, and the biggest "discounts" often land at prices the same model sold for six weeks earlier. UniverTrack tracks the real price history across retailers, spots fake reductions, and alerts you when it is genuinely the moment to buy.
