As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
For underfloor heating the useful pick is the FOTRIC TK5 (FOTRIC TK5). It is the only one of the three whose listing states a true 320 x 240 detector with no interpolation, a 40 mK thermal sensitivity and a 49 x 37 degree field of view, which is the combination that lets you see a whole manifold’s worth of pipe runs in one frame and still resolve the small temperature step between a flowing loop and a sluggish one. The FLIR E8 Pro (FLIR E8 Pro) is the better choice if you also do electrical and building work, because its 320 x 240 detector with MSX and cloud image handling suit report writing as much as floor tracing. The DCTOPT 320 x 240 (DCTOPT) is the budget route, but read its resolution claim carefully: the listing says 320 x 240 in the title and 240 x 240 super resolution in the bullets, which is not the same thing.
Why underfloor heating is a low-contrast problem
An underfloor heating loop is not a hot object. It is a pipe at perhaps 35 to 45 degrees Celsius buried 50 to 70 millimetres below a screed, and by the time that heat reaches the surface it has spread sideways and dropped. What the camera actually sees is a soft stripe a degree or two warmer than the floor beside it. That is the whole difficulty of this job, and it is why the two specifications that matter most here are not the ones buyers usually compare.
Thermal sensitivity, quoted in millikelvin, sets the smallest temperature difference the detector can distinguish. A 40 mK camera resolves four hundredths of a degree in theory; a 50 mK camera resolves five hundredths. On a floor where the pipe signature is one to three degrees above background, both will work, but the more sensitive instrument gives you a cleaner edge on the stripe, which matters when two loops run close together and you are trying to work out which one is which.
Field of view sets how much floor you can see in one frame. A narrow lens forces you to tile several images together to cover a room, and tiling is where mistakes happen, because the camera’s own temperature scale shifts between frames. A wide lens such as the 49 x 37 degrees published for the FOTRIC TK5 covers a large part of a room in one shot, so the whole manifold’s pipe pattern sits on a single consistent scale.
What each listing publishes about the detector
The single most important number in this category is whether the pixel count is a real detector or a software enlargement. The FOTRIC TK5’s listing is explicit about this: it says the camera uses a genuine 320 x 240 infrared detector with 76,800 raw pixels and specifically contrasts itself with imagers that rely on software interpolation. The FLIR E8 Pro’s listing also states a 320 x 240 thermal resolution with 76,800 pixels, so the two are comparable at the detector level.
The DCTOPT 320 x 240’s listing is the one to read slowly. The title says 320 x 240, while the bullet text describes a 2.8 inch LCD paired with 240 x 240 super resolution technology. Those are two different figures, and super resolution generally means the camera is capturing fewer native pixels and enlarging them. The listing does not publish the native detector size, so treat the 320 x 240 in the title as unconfirmed until the manufacturer states it plainly.
| Pick | Stated thermal resolution | Stated sensitivity | Stated field of view | Stated temperature range |
|---|---|---|---|---|
| FLIR E8 Pro | 320 x 240, 76,800 pixels, MSX | Not stated in the listing bullets | Not stated in the listing bullets | -20 to 550 C |
| DCTOPT | 320 x 240 in title; 240 x 240 super resolution in bullets | Under 50 mK | Not stated in the listing bullets | -20 to 550 C (-4 to 1022 F) |
| FOTRIC TK5 | 320 x 240 true, 76,800 raw pixels | 40 mK (0.04 C) | 49 x 37 degrees | -20 to 450 C (-4 to 842 F) |
Note that FOTRIC TK5’s stated upper limit is lower than the other two, which is irrelevant for floors but relevant if you also inspect plant. Note also that the listing does not publish an accuracy figure for the FLIR E8 Pro or the FOTRIC TK5, while the DCTOPT 320 x 240 states 2 degrees Celsius or 2 percent.
Reading a floor: what the camera will and will not show you
A thermal camera shows surface temperature, not pipe temperature and not flow. From a warm stripe you can infer that a loop is carrying hot water, and from the absence of a stripe you can infer that it is not. You cannot read flow rate from the image, and you cannot tell a balanced loop from an over-supplied one purely by how bright it looks, because floor coverings, air movement and furniture all change the surface reading.
The practical method is comparative. Run the heating until the floor is stable, then look at all the loops in the same frame if the lens allows. The loop that is cooler than its neighbours is the suspect. A camera with a wide field of view, like the FOTRIC TK5’s 49 x 37 degrees, lets you do that comparison in one image rather than three, which removes the scale-shift problem entirely.
Where the listing’s stated accuracy matters is in the absolute number. The DCTOPT 320 x 240 publishes 2 degrees Celsius or 2 percent, which is enough to say a loop is warm or cold but not enough to argue about a two-degree imbalance. Neither the FLIR E8 Pro nor the FOTRIC TK5 publishes an accuracy figure in the supplied bullets, so if you need to record absolute temperatures for a warranty claim, ask the manufacturer for the accuracy specification before buying.
The three cameras worth considering
FLIR E8 Pro
The all-rounder: 320 x 240 detector, MSX edge enhancement and cloud image storage make it as useful for the report as for the floor.
DCTOPT 320 x 240
The budget option with a stated sub-50 mK sensitivity and a 25 Hz frame rate, but its resolution wording needs reading twice.
FOTRIC TK5
The floor specialist: true 320 x 240 detector, 40 mK sensitivity and a wide 49 x 37 degree field of view for whole-room sweeps.
The mistake that makes a floor survey useless
The commonest error is scanning too early. Underfloor heating responds slowly because the screed has to charge before the surface shows the pipe pattern. A camera used ten minutes after the manifold opens will show a diffuse warm patch with no visible loop structure, and the operator concludes the camera is not sensitive enough. It usually is. The floor is simply not ready.
The second error is emissivity. Bare screed, tile, timber and carpet all radiate differently, and a camera left on a default emissivity of 0.95 will read a shiny surface low. The listings here address this unevenly: the FOTRIC TK5 states adjustable emissivity as a built-in tool, the FLIR E8 Pro and the DCTOPT 320 x 240 do not mention it in the supplied bullets. On a floor with mixed coverings, an adjustable emissivity setting is the difference between a comparable survey and a set of numbers you cannot line up.
The third error is using a camera to prove a pipe is blocked. A camera shows the symptom, a cold patch or a cold loop. It does not show the cause, and it will not tell you whether the problem is a stuck actuator, air in the loop or a crushed pipe. For that you need pressure and flow measurements at the manifold. Where a camera earns its place is in narrowing down which loop to open first. If your work is mostly tracing hidden pipework rather than floor heating specifically, the approach in cameras for finding water leaks is a closer fit.
Where this category stops being the right answer
A thermal camera is a screening tool. It tells you where to look. It does not measure flow, it does not measure pressure and it does not commission a system. If the question is whether a manifold is balanced to its design flow rates, you need a flow meter on each loop, not an image of the floor. If the question is whether a zone valve is opening, you need a multimeter at the actuator, and the reasoning in cameras for electrical maintenance covers that overlap.
There is also a floor-covering limit. Thick carpet, underlay and solid timber all insulate the pipe signature, and at some point the temperature difference at the surface falls below what any handheld camera can resolve reliably. A 40 mK camera will reach further into that problem than a 50 mK one, but neither will see through a deep-pile rug. In those rooms the honest answer is to lift a corner or use an acoustic or flow-based method instead.
Finally, the camera does not replace a record. Photograph the manifold, note which loop is which, and keep the images so that next season’s survey can be compared against this one. A camera that stores images to a card or the cloud, as the FLIR E8 Pro and the DCTOPT 320 x 240 both state, makes that routine rather than a chore.
Choosing between the three, and the case for buying nothing
If underfloor heating is the main job and you want the clearest loop pattern with the fewest frames, the FOTRIC TK5’s true detector, 40 mK sensitivity and wide lens are the most directly matched to the task. If you are a general inspector who does floors occasionally and also writes reports, the FLIR E8 Pro’s MSX edge enhancement and cloud image handling give you a more versatile instrument, though its sensitivity and field of view are not published in the supplied bullets. If budget is the deciding factor, the DCTOPT 320 x 240 is the cheapest of the three on the evidence available, but its resolution claim is ambiguous and should be confirmed before purchase.
One product in the source data, the CaseSack Pocket2, is a carrying case rather than a camera, so it is excluded from the comparison. It is relevant only as an accessory: the listing describes a semi-hard case compatible with HIKMICRO Pocket2 and Pocket1 with a mesh pocket for a charge cord, which is a sensible purchase for protecting whichever camera you choose, but it is not an instrument and it does not belong in a ranking of cameras.
If you only need to check one room once, consider hiring or borrowing a camera rather than buying. The purchase makes sense when you will survey floors repeatedly, across multiple properties, or when you need images for a report. For a single domestic fault, the money is usually better spent on a heating engineer with flow meters. Readers who want to compare across the wider field can look at budget thermal cameras and high resolution thermal cameras for the two ends of the market.
Recommended Tools
Products shown because their documented specifications fit this guide. Price and availability are on Amazon and can change.
Frequently Asked Questions
What resolution do I actually need to see underfloor heating pipes?
A native 320 x 240 detector is the practical floor for this work, which is what the FLIR E8 Pro and the FOTRIC TK5 both publish. Lower native resolution with software enlargement, which is what the DCTOPT 320 x 240's bullet text may describe, will show the pipe stripe but with softer edges, making it harder to separate two adjacent loops. The listing for the DCTOPT 320 x 240 does not publish a native detector size, so confirm it with the manufacturer if this matters to you.
Does the camera need a wide lens for floor surveys?
It helps a great deal. A wide field of view, such as the 49 x 37 degrees published for the FOTRIC TK5, lets you capture several loops in one frame so they share a single temperature scale. With a narrower lens you must stitch images, and the scale shifts between frames, which makes comparison unreliable. The FLIR E8 Pro and the DCTOPT 320 x 240 do not publish a field of view in the supplied bullets.
Can a thermal camera tell me if a loop is blocked?
No. It shows you a cold patch or a cold loop, which is the symptom. It cannot distinguish between a blocked pipe, a stuck actuator, air in the system or a balancing problem. Confirming the cause needs pressure and flow measurements at the manifold. The camera's job is to tell you which loop to investigate first.
How long should I run the heating before scanning?
Long enough for the screed to charge, which in practice means until the floor surface temperature is stable rather than still rising. The listings do not publish a warm-up time because it depends on the floor build-up, not the camera. Scanning too early is the most common reason people conclude a camera is not sensitive enough.
Is a 40 mK camera meaningfully better than a 50 mK one for floors?
Yes, at the margin. The pipe signature on a floor surface is often only one to three degrees above background, and a more sensitive detector gives a cleaner edge on that stripe. The FOTRIC TK5 publishes 40 mK and the DCTOPT 320 x 240 publishes under 50 mK. The FLIR E8 Pro does not publish a sensitivity figure in the supplied bullets, so it cannot be compared on this point from the listing alone.
Do I need the cloud features for underfloor heating work?
Not for finding the fault, but they help with record keeping. The FLIR E8 Pro's listing states automatic cloud storage, editing and sharing through FLIR Ignite, and the DCTOPT 320 x 240 states a 4GB detachable card holding over 10,000 images. If you survey the same property season after season, having the previous images to hand is genuinely useful. If you are a one-off user, a memory card is enough.
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.