As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
Raythink · Raythink EX10 Pro 2.8" Dual-Lens Thermal Camera, 320×240 Super Resolution | 10h Continuous Runtime, Laser Pointer, Wi-Fi, 25Hz Handheld Infrared Camera for Home & Electrical Detection, -4°F~1022°F, 8G
View on AmazonPrice and availability are shown on Amazon and can change at any time.
A standalone thermal camera with a 2.8 inch display, a 2 MP visible camera for fusion and picture in picture, nine palettes, a laser pointer, 8 GB of internal storage, a stated 10 hours of runtime and an IP54 housing rated for 6.6 ft drops. It measures -4 to 1022 F with plus or minus 2 C or 2 percent, and resolves differences as small as 40 mK. One word in the first bullet decides whether the headline figure means anything: the thermal resolution is boosted to 320 by 240 in real time, and the listing never states the size of the detector doing the looking.
Boosted is the verb that decides this purchase
Every thermal camera has two resolutions. One is the microbolometer array, a physical grid of elements each of which measures the infrared arriving at it. The other is the size of the image the camera saves. They are frequently not the same number.
A camera with a native 320 by 240 detector says so in three words: 320 by 240 detector. This listing instead says it boosts thermal resolution to 320 by 240 in real time, and the product name carries the phrase Super Resolution. Both are the vocabulary of a processing step applied to something smaller.
The reason this matters is that the element count sets the smallest real thermal feature the camera can separate. Upscaling adds pixels. It does not add information, because the heat pattern was sampled at the detector pitch and nothing downstream can recover what was never sampled.
In fairness there are two different techniques hiding behind the same marketing word, and they are not equally empty:
- Interpolation. The processor invents intermediate pixels from their neighbours. The image looks smoother at a glance and contains no additional detail. A small hot spot that fell between elements is still invisible.
- Multi frame micro scanning. The optics or the detector are dithered by a fraction of an element between frames and several slightly offset readings are combined. This genuinely recovers detail, at the cost of needing a steady hand and a static subject, because anything moving smears.
The listing does not say which of those is happening, and it does not publish the native format, so there is no way to work it out from the page. That is the single question to put to the seller before buying, and the answer changes what the camera is worth. Our guide on comparing thermal camera resolution works through why the detector figure is the one to compare.
Note what is not being said here. A boosted figure is not a lie, and plenty of respected cameras upscale their output. The complaint is narrower: the figure a buyer compares is absent, and the figure that is present is the one that cannot be compared.
Forty millikelvin is the figure that is not in doubt
The same bullet carries a specification that is completely unambiguous, and it is the best thing on the page.
Thermal sensitivity, properly NETD, is the smallest temperature difference the camera can separate from its own electronic noise. At 40 mK the camera can distinguish surfaces 0.04 C apart. The listing converts the figure correctly, which is worth noting because a kelvin and a degree Celsius are the same size interval and plenty of listings get this wrong in both directions.
Many cameras at this level publish 50 mK, and a good number publish nothing at all. Forty is a strong figure, and unlike the resolution claim there is no processing trick available: sensitivity is a property of the detector and its electronics.
Whether sensitivity or element count matters more depends entirely on the job:
- Insulation, damp, underfloor heating, draughts. These appear as fractions of a degree spread across large areas. Sensitivity decides whether the pattern emerges from the noise at all. Element count barely matters, because the feature is bigger than the frame.
- Electrical terminations, small components, bearings. These are large temperature differences concentrated in small areas. Here the element count decides whether the hot spot occupies enough of the frame to be seen and measured, and sensitivity is never the limit.
So the awkward conclusion is that the listing leads with electrical panels, which is the application where the unstated detector format matters most, while the figure it publishes most convincingly serves building work instead. Our explainer on what NETD means covers why the two figures answer different questions.
The range is in Fahrenheit and the sensitivity in Celsius
The fourth bullet states -4 to 1022 F with plus or minus 2 C or plus or minus 2 percent accuracy, in the same sentence. Converting the range gives -20 to 550 C, which is a conventional span for a camera of this type and a useful one: it reaches cold bridging at the low end and flue and exhaust temperatures at the high end.
The accuracy wording needs unpacking, because the two terms are not alternatives you get to choose between. The standard reading of plus or minus 2 C or 2 percent is whichever is larger at the measured temperature:
- At 30 C, the 2 percent term is 0.6 C, so the fixed 2 C term governs.
- At 100 C, both terms are 2 C and they coincide.
- At 500 C, the 2 percent term is 10 C and the fixed term is irrelevant.
Anyone reading plus or minus 2 C and expecting it to hold across the whole span is going to be five times out at the top of the range.
It is also worth being clear that a 2 C accuracy alongside a 0.04 C sensitivity is not a contradiction. Sensitivity governs comparison: this terminal is warmer than the two beside it. Accuracy governs the absolute number: this terminal is at 68 C. Most thermal inspection is comparative, which is why a modest accuracy figure does far less damage than it first appears.
The gap that does damage is emissivity. The bullets mention nine palettes, four modes, six user spots and a laser, and never mention an emissivity setting. Without one, every reading on bare or polished metal is low, often dramatically, and the first bullet points this camera at electrical panels full of exactly that. Our guide on setting emissivity explains the size of the error.
IP54 and 6.6 ft are both specific, and one is weaker than it looks
Credit where it belongs: this listing publishes two durability figures in a category where most publish adjectives.
IP54 decodes cleanly. The 5 means dust protected, so ingress is not entirely prevented but not enough enters to interfere with operation. The 4 means protection against splashing water from any direction. What it stops short of is a jet, a hose and immersion. Rain on a roof survey is within the rating. A washdown bay, a food plant hose down or a dropped instrument in a puddle is not.
The 6.6 ft drop figure is the more unusual one and the more useful. Instruments rarely die from water. They die from being dropped off a ladder, a bonnet or a bench, and 6.6 ft covers a fall from a standing hand onto a hard floor.
The battery and storage figures belong in the same section because they are what a standalone body buys you over a camera that clips to a phone. Ten hours of continuous operation and 8 GB of internal storage mean no app, no pairing, no phone battery drain and no handling a glass slab with gloves on in a loft. The trade is that the capacity, the charging time and whether the cell can be replaced are all unstated, so a camera that is tired after a few years of cycles may not be serviceable.
Reading the five bullets as a specification sheet
| Figure | What the listing says |
|---|---|
| Output thermal resolution | 320 by 240, described as boosted in real time |
| Native detector format | Not stated in the listing bullets |
| Thermal sensitivity | 40 mK, stated as 0.04 C |
| Temperature range | -4 to 1022 F, equal to -20 to 550 C |
| Measurement accuracy | Plus or minus 2 C or plus or minus 2 percent |
| Display | 2.8 inch |
| Visible camera | 2 MP |
| Imaging modes | Thermal, visible, fusion, picture in picture |
| Colour palettes | Nine |
| Measurement points | Automatic max, min and centre, plus six user spots |
| Alarms | User set high and low |
| Laser | Integrated pointer |
| Storage | 8 GB internal |
| Runtime | Up to 10 hours continuous |
| Ingress protection | IP54 |
| Drop rating | 6.6 ft |
| Connectivity | Mobile app and Windows PC software |
| Field of view | Not stated in the listing bullets |
| Frame rate | Not stated in the listing bullets |
| Focus, fixed or adjustable | Not stated in the listing bullets |
| Minimum focus distance | Not stated in the listing bullets |
| Emissivity adjustment | Not stated in the listing bullets |
| Display resolution | Not stated in the listing bullets |
| Battery capacity | Not stated in the listing bullets |
| Radiometric file export | Not stated in the listing bullets |
Two of the blanks deserve a note. Frame rate is missing, and it is not a trivial specification: some low cost thermal cameras run at 9 Hz, and the image lags and smears when you pan along a wall or a run of conduit. Field of view is missing too, which means you cannot work out how much of a wall you capture from a given standoff, and with focus also unstated you do not know the closest distance at which a terminal block or a circuit board will be sharp.
The PC software bullet promises temperature analysis, which implies but does not state radiometric files, where every pixel keeps its own temperature so you can place a new measurement spot after the fact. That is the difference between a thermal photograph and a thermal measurement you can revisit, and it is worth confirming.
Buy it for the sensitivity, not for the resolution headline
This suits someone doing comparative inspection on a budget who wants an instrument rather than a phone accessory. Insulation surveys, underfloor heating runs, HVAC airflow, tracing a warm pipe through a wall, finding what is hotter than its neighbours. For that work the 40 mK figure is the one that decides whether you see the pattern, and the 10 hour battery, the 8 GB of storage and the reporting software make up a coherent day of work.
Skip it, or at least pause, in three cases. If you are buying on the 320 by 240 figure, get the native detector format in writing first, because the listing has told you that number is produced rather than sensed. If you need absolute temperatures on metal, the absence of any emissivity mention is disqualifying until confirmed otherwise. And if your environment involves hoses or washdown, IP54 is not the rating you need.
Recommended Tools
Products shown because their documented specifications fit this guide. Price and availability are on Amazon and can change.
Frequently Asked Questions
What is the difference between a detector resolution and a boosted resolution?
A thermal camera has a physical sensor, a microbolometer array, with a fixed number of elements. That count sets the smallest real thermal feature the camera can separate. The image the camera writes to a file can be any size you like, because processing can interpolate between elements or combine several slightly offset frames. A listing that states a detector format is telling you about the sensor. A listing that says resolution is boosted to a figure in real time is telling you about the output. This listing does the second and never does the first.
Does 40 mK really mean 0.04 degrees Celsius?
Yes, and the conversion in the listing is correct. A millikelvin is a thousandth of a kelvin, and a kelvin and a degree Celsius are the same size interval, so a difference of 40 mK is a difference of 0.04 C. The figure is a noise specification, properly called NETD: it describes the smallest temperature difference the camera can distinguish from its own electronic noise. It says nothing about how accurate the absolute reading is, which is a separate figure.
Why is the range quoted in Fahrenheit and the accuracy in Celsius?
There is no technical reason, and it is worth converting so you can compare the camera against others. The stated range of -4 to 1022 F is -20 to 550 C. The accuracy is given as plus or minus 2 C or plus or minus 2 percent, which almost always means whichever is larger. Below 100 C the fixed 2 C term dominates. Above it the percentage does, so near the top of the range the tolerance is around 10 C rather than 2 C.
Does the lack of an emissivity adjustment matter?
It matters a great deal if you point the camera at metal. Emissivity describes how efficiently a surface radiates infrared, and polished or bare metal radiates very poorly, so a camera set for a painted surface will read a shiny busbar or a copper pipe far below its real temperature. The bullets list nine palettes, four imaging modes, a laser pointer and six user spots, and never mention emissivity. On a camera whose first bullet talks about electrical panels, that is the omission to ask about before buying.
What does IP54 allow and not allow?
The 5 covers dust: ingress is not completely prevented but not enough gets in to interfere with operation. The 4 covers water: protection against splashing from any direction. It does not cover jets, a hose or immersion, which are IP65, IP66 and IP67 respectively. In practice that makes the camera fine in rain, in a loft or on a dusty site, and not suitable for a washdown area or a deliberate rinse. The 6.6 ft drop figure is the more unusual specification here and it is the one that describes how instruments actually die.
Is the laser pointer a measurement aid?
Only indirectly. The laser marks a point in the visible world so you can tell a colleague which breaker you mean, or confirm you are aimed at the right pipe among several. It does not mark the measured area on the thermal image, and because the laser, the thermal lens and the visible lens sit in different places on the body, the laser dot and the measured spot diverge as you move closer to the target. Use the on screen markers for measurement and the laser for pointing.
Affiliate disclosure: Akermin earns a commission from qualifying Amazon purchases made through links on this page. Our editorial picks are based on documented specifications and owner feedback, not commissions.