A thermal imaging camera measures the infrared energy leaving a surface and turns it into a picture of surface temperature. In building work it is a scanning tool: it shows where a wall, ceiling or floor is colder or warmer than it should be, which tells an inspector where to look closer.
Here is the fact everything else depends on. A thermal camera does not see moisture. It sees temperature difference. Wet material usually reads cooler than the dry material around it because evaporation pulls heat out of the surface, so water often shows up, but what the camera recorded was a cold spot. Thermal imaging locates suspect areas. A moisture meter confirms them. This page covers how the instrument works, which specifications matter for building work, how to read an image without fooling yourself, and where it fits a mold inspection.
Anything sold as a leak detector or a moisture camera is overstating what the instrument does, and an inspector who hands you a thermal image as proof of a moisture problem is doing the same.
Wet insulation and framing store and release heat differently than dry material, which shows as a lag during a heating or cooling cycle. That effect is real and useful. It is also produced by a dozen things that have nothing to do with water.
So the sequence on a competent inspection never changes. The camera finds an anomaly. A pin or pinless moisture meter goes on that exact spot and either confirms elevated moisture or does not. When the meter disagrees with the picture, the meter wins. A blue blotch on a ceiling looks like evidence, and it is only a reason to take a reading.
Every surface above absolute zero emits infrared radiation, and emits more as it gets warmer. The detector, a grid of elements called a microbolometer array, measures that radiation across the scene and assigns a color to each element.
What trips people up is emissivity, meaning how efficiently a material radiates infrared compared to a perfect emitter. Painted drywall, wood, brick and plaster run around 0.90 to 0.95 and image accurately. Shiny metal, foil-faced insulation, glass and glossy tile image badly. They under-report their own temperature and reflect infrared from everything else in the room, so point a camera at a foil-faced duct and you may be photographing the technician holding it.
The habit that catches this costs nothing. Take a second image from a different angle. Real thermal patterns stay put. Reflections move with you.
Resolution is the count of actual thermal pixels, and it decides whether a small anomaly is visible from across a room.
Think of it as pixels on target. A damp area the size of a dinner plate, seen from 15 feet on a 160 x 120 detector, lands on a handful of pixels and smears into the background.
Be careful with megapixel claims. Some cameras advertise a large image produced by upscaling, or by blending thermal data with a photo from a second visible-light lens. FLIR calls its version MSX. The blend is useful, because it puts recognizable edges like outlet covers and trim into a vague thermal blob, but it adds no thermal information. The number that counts is native detector resolution, and a manufacturer who buries it is telling you something.
Thermal sensitivity, written as NETD and quoted in millikelvin, is the smallest temperature difference the detector can distinguish. Lower is better. Entry cameras commonly quote under 100 or under 60 mK, professional building cameras under 40 or under 30 mK. For moisture work this matters as much as resolution and gets far less attention. A damp cavity might sit half a degree cooler than the wall around it, and a camera that cannot separate that from its own noise shows you a clean wall.
A wide lens covers a whole wall from a hallway. A narrow lens puts more pixels on a distant target, which a high ceiling or a roof needs. Fixed-focus cameras are sharp in a set range and vague outside it, and a soft image is not merely ugly, it is wrong, because it averages the target with everything around it.
If the wet area and the wall around it are the same temperature, the camera shows nothing and the wall is still wet. It is invisible.
Delta-T is the temperature difference across the assembly you are imaging, and creating one is half the skill. Run the HVAC hard for an hour or two beforehand, which drives a difference across the wall and makes wet material lag behind dry. Image after a temperature swing, such as late afternoon on a day that warmed up. Work at dawn or dusk on exterior surfaces.
The hardest condition to image in is a mild overcast day with nothing running: no sun load, no heating, no cooling, indoor and outdoor temperatures close together. The building sits at equilibrium and the camera shows a beautifully uniform wall. An inspector who says conditions are poor and offers to come back is being honest.
Most of what shows up on a scan is not a finding. Ordinary building features produce patterns that look exactly like problems.
Stated plainly: a thermal anomaly is a question, not an answer. It gets answered with a meter, a borescope through a small hole, or a look at the plumbing above.
| Class | What it is | Good for | Where it falls short |
|---|---|---|---|
| Phone attachment | A thermal module that plugs into a phone | An obvious problem: a drafty window, a cold corner, missing attic insulation | Low native resolution and sensitivity, no radiometric file for later analysis |
| Standalone handheld | Own screen, battery and storage | Trade work, HVAC, electrical, light building diagnostics | Wide variation in sensitivity and focus, and the cheap end oversells resolution |
| Professional building camera | Higher resolution, low NETD, manual focus, radiometric storage, reporting software | Documented inspections, moisture surveys, energy audits | Cost, and a learning curve the instrument does not remove |
Phone attachments deserve a fair hearing. For a homeowner working out why one room is always cold, they genuinely work at a fraction of the price. What they are not is documentation. An image headed for an insurance file or a remediation scope needs a radiometric file from a camera with enough detector behind it to be believed.
Moisture is a minority of the work.
It decides where to put the meter and where a hole is worth drilling.
A mold inspection has a scoping problem: growth is usually hidden, and you cannot open every wall. The camera scans a large area quickly, flags the few spots out of line with their surroundings, and a moisture meter confirms. Where the meter agrees, a borescope through a small drilled hole shows what is in the cavity. That sequence is how a scope gets defined without exploratory demolition.
It also points at where the water came from. Under ANSI/IICRC S520 the property owner is responsible for identifying and correcting the moisture source, so finding the leak decides whether the problem returns after remediation.
Named without prices, because street pricing moves constantly. What price mostly buys is detector resolution, then sensitivity.
| Brand | Tier | Notes | What it does badly |
|---|---|---|---|
| FLIR | Phone attachment through professional | The default name in the category. MSX visible-image blending makes pictures far easier to interpret, and the reporting software is strong. | A confusing lineup where similar-looking models differ sharply in detector resolution, at a premium over equivalent specifications. |
| Seek | Phone attachment and compact standalone | Good value at the homeowner and light-trade end, and capable for the money. | Resolution and sensitivity fall short of professional building work, and the software is thin. |
| Hti | Budget standalone | The cheapest self-contained camera with its own screen, for an occasional user who will not tie up a phone. | Quoted image sizes often reflect interpolation, not native pixels, and build quality, focus and support are inconsistent. |
| Fluke | Professional | Built like their other instruments, meaning they survive job sites. | Priced for professionals, and aimed more at electrical and mechanical work than building envelope moisture. |
| Testo | Professional building diagnostics | Aimed most clearly at envelope work, with surface humidity and condensation risk shown alongside the image. | Less common in North America, so support and used inventory are harder to find. |
Yes, and it is one of the more rewarding tools a homeowner can rent or buy.
Reasonable uses. Finding drafts and air leaks. Checking attic insulation coverage before paying for more. Locating radiant floor loops. Working out why one room never gets warm.
Where the boundary is. A camera cannot tell you whether a cold patch is wet, let alone whether anything is growing behind the drywall. If you find something, the next step is a moisture meter, not a demolition saw. Hidden moisture is exactly the situation where opening a wall without containment turns one room into a whole building.
The rule is the same one on the main remediation page. A small, visible, isolated area with an obvious moisture source is reasonable for a careful property owner with proper PPE. Beyond that, or wherever growth is hidden, the HVAC system is involved, or anyone has health concerns, the work needs a built containment and a contractor.
Read more: moisture meters and how readings are interpreted · Read more: borescope inspection of wall cavities
No. Any claim otherwise is false. A thermal camera measures surface temperature, and mold has no thermal signature separating it from the material it grows on. What a camera can find is the moisture condition that allows growth, and only indirectly, by showing a temperature difference a moisture meter then has to confirm.
It can find the thermal effect of one. Wet material usually reads cooler because evaporation removes heat, so an active leak often appears as a cold pattern spreading from a source. That is a strong lead, not a confirmation. Every anomaly needs a moisture meter reading in the same place before anyone calls it water.
For serious building diagnostics, 320 x 240 native detector resolution is the practical minimum, because you need enough pixels on target to resolve a damp area from across a room. A 160 x 120 camera works at close range on obvious problems. Check native detector resolution rather than advertised image size, which is often upscaled.
Prices span a wide range and move constantly, so check current pricing before buying. Phone attachments are the least expensive by a wide margin, standalone handhelds cost several times more, and professional building cameras several times more again. What the money buys is detector resolution and thermal sensitivity, in that order.
For a homeowner, yes. They find drafts, missing insulation and cold spots perfectly well at a fraction of the cost of a handheld. What they cannot do is produce documentation. Native resolution and thermal sensitivity are both limited, and the images will not carry weight in an insurance claim or a remediation scope.
Because it reads the surface only. Infrared does not pass through drywall, plaster or wood the way X-rays pass through tissue. What appears on a wall is the effect of something behind it on that wall's surface temperature: a cold stud, a warm pipe, a damp cavity cooling the face. You are reading a symptom, never the object.
MI&T is an inspection and testing company. We do not perform mold remediation, we do not sell cameras or meters, and we take no referral fees, so we have no stake in which instrument anyone buys.
Thermal imaging is one of the tools we use to scope an inspection, and it is never the whole answer. The camera narrows a building down to a few suspect areas. Meters, cavity inspection and sampling establish what is there, how far it extends, and where the water came from. The result is a scope a contractor can price, rather than a picture of a blue patch.
Find your nearest MI&T location to schedule an inspection or clearance test.
