Measuring CryoPaint Performance

PDRC (passive daytime radiative cooling) surfaces behave differently than ordinary paint, so the tool you use to measure them matters as much as the paint itself.

Before you measure

Paint the Whole Object, Not Just a Patch

This is the single most common reason a test comes back disappointing. Cooling is a rate per unit of area, so a patch on a larger metal roof, wall, or container is not a fair test. An uncoated section of metal can absorb nine times more heat than the CryoPaint can remove.

Coat the entire connected surface, or test on a small, fully separate object like a piece of scrap metal on a piece of cardboard. See why surface area decides the result.

Start here

Designing a Valid Test

The instrument matters less than the rig. Eliminate variables methodically and strategically for a cleaner test with less data noise.

Your ideal setup is like a sandwich, from top to bottom:

CryoPaint
Metal panel
Thermistor probe
Tape
Piece of cardboard
Towel
Table

Verify your thickness before anything else. Thickness is the variable you cannot see after the paint dries, and it is the most common silent test-killer. Check it two ways. First, arithmetic: divide the ounces you used by the square feet you covered. If a 4 oz sample covered 4 sq ft, you are at the bare 1 oz/sq ft minimum; if it covered 8 sq ft, the film is too thin for a meaningful test. Second, a cheap wet-film thickness gauge: 1 oz/sq ft laid down in one pass reads about 12.5 wet mils, and a single roller coat reads only about 4. That is why rolling takes 3.5 coats to reach the rolled minimum and about 6 to 7 coats to reach the 2 oz/sq ft we recommend for a test. If you rolled it on, count your coats before you blame the paint: two roller coats is roughly 0.6 oz/sq ft, under the minimum, and a thin film gives a weak result on the best day of the year.
Paint a large surface area. The larger the surface area you paint, the more cooling power it has. Cooling is about -100 W/m², so the total grows with the square meters you paint. Of course, if you spread the paint too thin, it will counteract the surface area you've added. The sweet spot is about 2 oz. per square foot for a test.
Isolate the sample thermally. The optimal way to do this is to get a 1 sq. ft. piece of metal and put it on top of a piece of cardboard or a folded towel. That way, the ground, table, or other surface you're placing it on isn't conducting heat into the metal itself
Point it at open sky. Lay it flat and face it up, out from under eaves, trees, and walls. Judge the view from the panel's position: the more of its field of view that is sky rather than ground and structures, the more cooling there is to measure. Also, be aware of where you're at when you are measuring it. If you hover right over it, that's going to affect the result!
Place the probe on the underside, toward the center. if you're using a thermistor or thermometer probe, attach the sensor to the back of the panel where the sun cannot hit it, and keep it in the center. The metal conducts the surface temperature through to the sensor.
Shade the readout, not just the probe. Meters and dataloggers drift when they bake in the sun. Keep the thermistor housing shaded and cool so the electronics are not adding their own error to the number. (you could even put it in a CryoPainted box)
Log the wind speed and other live weather with every reading. Sustained wind pushes the surface back toward air temperature, so a reading taken in wind is partly a wind measurement. See the wind and convection guidance. Additionally, taking your live current weather allows you to more accurately consider which values may have been at play when analyzing in post.
Measure the air temperature at the test, in the shade. Every performance number is a difference against air temperature, so how you measure the air matters as much as how you measure the panel. Use a thermometer a few feet from the panel, shaded from the sun and up off the hot ground. Do not use the weather app on your phone: that reading comes from a station that can be miles away and several degrees different from the air right over your test.
Give it time, and average your readings. Let the panel sit undisturbed for about 15 minutes so it reaches its steady temperature. Then take several readings over 5 to 10 minutes and use the average, not the single best moment. If you can, check at more than one time of day. Midday in full sun is the hardest test the coating ever faces; late afternoon and after dark show larger differences.

What to expect

What a Healthy Test Reads

Grade your result against conditions, not against the single best number on the label. On a properly built test panel (fully coated, thermally isolated, at least 1 sq ft, about 2 oz/sq ft of paint), here is what a working coating reads:

Condition	Expected reading
Calm (under 5 mph), clear sky, dry air	10°F to 15°F below air temperature
Sustained wind 5 to 10 mph	20 to 30% less, roughly 10°F on an otherwise perfect day
High humidity (roughly 60% and up) or hazy sky	As little as half; humid climates often see 3°F to 7°F
Cloud cover, or the sky view is blocked (fails the Y test)	Near air temperature
A patch on a larger connected object, any weather	At or above air temperature, no matter how good the paint is

At air temperature in full sun is the floor of the coating, not a failure of it. A normal surface would be far above air temperature in the same sun. See the wall test on the How It Works page for what that floor looks like on camera.

Best - Method 1

Calibrated High-Quality Thermal Camera

A properly calibrated thermal camera from a reputable brand is our favorite tool for visualizing and measuring PDRC performance. It gives you a full thermal image of the surface and surrounding area, which makes it easy to see the cooling effect side-by-side with untreated materials.

The catch: the camera must be calibrated correctly and it must be a high-quality unit. In our own field testing we have run into repeated problems with cheap thermal cameras. They often get the relative scale right (cooler areas do look cooler than warmer ones) but the absolute temperature numbers they print on the image are simply wrong, sometimes by 10°F or more.

If you are using a thermal camera, set the emissivity to 0.90, CryoPaint's published emittance, and set the reflected temperature for your scene. Outdoors under a clear sky the reflection is very cold, which pushes an uncorrected reading low; nearby warm walls, equipment, or your own body reflect warm and push it high. When in doubt, check the camera against a contact probe on the same panel.

Method 2

Direct-Contact Thermometer (Thermistor or Thermocouple)

Our second-favorite method is a direct-contact thermometer - a thermistor, thermocouple, or RTD with a wire-based or surface-sensing probe that you can attach to the surface for a sustained reading, or press directly on top of the coating for a brief period.

This works especially well when CryoPaint is applied on top of a conductive substrate like sheet metal. You can place the probe on the underside of the metal and the conductive material will carry the surface temperature through to the probe. It is highly accurate, repeatable, and relatively cheap.

A thermistor we have used successfully: Digital thermistor probe on Amazon.

Pro tip: If your project permits it, tape the thermistor sensor directly to the surface before painting and then paint right over it. The sensor ends up embedded under the coating in perfect thermal contact with the substrate, which gives an extremely accurate, drift-free reading of the actual painted surface temperature.

Watch out for:

The probe or thermometer body itself sitting in direct sun and heating up - this will ruin the reading.
Trapped convective heat. If you are measuring the roof of an enclosed box that is hot on all sides, the rising internal convective heat can warm the underside of that roof faster than the coating can radiate it away. Open the box, add airflow, or test a flat surface in isolation for a clean reading.
Wind during the test. Convection drives the painted surface back toward the air temperature - the same reason blowing on a cold cup of coffee warms it up. Gusts and light breezes are fine, but with sustained wind above 5 mph you should expect a smaller delta, and it shrinks further as speed rises. Record the wind speed alongside every reading, and if you can, test in a sheltered spot or on a calm morning. A "low" number measured in sustained wind is a wind measurement, not a paint measurement.

Free & Easy - Method 3

The Hand Touch Test

The cheapest and easiest way to confirm the product is working is to touch the surface with your hand, ideally comparing it side-by-side with another white surface of the same material - for example, a white-painted piece of metal next to a CryoPaint-coated piece of the same metal.

You can also check by asking yourself a simple question: does the surface feel like the air temperature, or hotter? Say it is a 100°Fday. Touch the coated surface and ask, "Does this feel like 100°F or hotter?" If the honest answer is no - it feels cooler than the air around you - then the surface is sub-ambient and the paint is doing its job.

It is not scientific, but it is a quick, free, no-equipment sanity check when the other tools are not available.

Do Not Use

Infrared Thermometer Guns

We recommend against using IR thermometer guns for PDRC surfaces, even high-end ones. Here is why:

IR thermometer guns infer temperature by measuring infrared radiation coming off a surface. What they actually see is the surface's own emission plus whatever ambient IR the surface reflects back at the gun. On a normal roof that is a non-issue. But a PDRC surface is engineered to be highly emissive only in the 8-13 µm atmospheric window and is also actively sub-ambient, which creates three practical problems:

No way to correct for the scene. Most guns lock emissivity at a fixed value (usually 0.95) and have no reflected-temperature setting at all. CryoPaint's emittance is 0.90, and on a sub-ambient surface the small errors a gun cannot correct for are as large as the signal you are trying to measure.
Reflected longwave IR contamination. The coating reflects longwave IR from whatever is around it - your body, nearby walls, the ground, even the gun itself at close range. Warm surroundings make the surface read warmer than it actually is, which masks the cooling effect.
Real readings get dismissed as errors. Because PDRC surfaces can sit several degrees below air temperature, readings that look "too cold" often get written off as instrument error when they are real - while readings that look "about right" are often reflection-contaminated and falsely warm.

Angle makes it worse: as the gun deviates from perpendicular to the surface, the amount of direct emission reaching the detector decreases and reflected radiation increases at oblique angles.

For defensible numbers, use a contact probe (thermocouple or RTD) bonded to the surface, or a calibrated IR camera with the coating's actual emissivity entered and a clear view of cold sky, not warm objects, in the reflective background.

Troubleshooting

Still Not Seeing the Cooling You Expected?

Work through this list in order. Almost every disappointing test we have ever diagnosed lands on one of these.

The panel reads hotter than the air. This is almost always a patch on a larger connected object, heat conducted in from whatever the panel is sitting on, or a film under 1 oz/sq ft. Re-check isolation (metal on cardboard, nothing hot underneath) and do the thickness arithmetic above, including the roller coat count.
The panel reads at air temperature. Check, in order: sustained wind above 5 mph, a blocked sky view (run the Y test from the panel's position), high humidity or haze, an insulated substrate under the panel, and a film at or below the minimum rate. Each one of these alone can erase the sub-air margin.
The panel is below air but less than the table above. Look at the wind band you tested in, the humidity, the camera's emissivity and reflected-temperature settings, and whether your air reference thermometer was truly shaded. A reference sitting in the sun reads high and exaggerates the gap; one indoors reads low and hides it.

If you have walked the list and the numbers still do not make sense, send us the test and we will diagnose it with you: photos of the setup, ounces used and area covered, coat count if rolled, what the panel sits on, wind, humidity or dew point, time of day, and your readings with how you took them. Every one of those details changes the answer.

Field notes

Test Log

Copy this list into your notes, or print this section, and fill it in for every reading. Most failed tests diagnose themselves the moment the conditions are written next to the numbers.

Date and time
Sky: Y test result and cloud cover
Wind: flag check or measured speed
Humidity or dew point
Panel: material, size, and what it sits on
Paint: ounces used, area covered, method, and coat count
Air temperature: measured in the shade at the site
Panel readings: instrument used, several readings over 5 to 10 minutes
Notes: anything nearby that blocks sky view or changes the wind

Back to all installation resources