A 640×512 thermal sensor roughly doubles the range at which you can detect and identify a target compared with 256×192 — but for close-range roof, facade and equipment work, the lighter 256-class payload is often the smarter buy.
- ~4xPixels on target at 640
- 20 gWeight difference, MV-2M to MV-2P
- ~100 mWhere the decision flips
- 12 µmPitch on both
Key takeaways
- More pixels on target at the same distance is the whole difference — a person at 300 m fills roughly four times the pixels on a 640 sensor.
- That converts a warm blob into a recognisable human, which is the difference between a contact and a finding.
- The MV-2M pairs 256×192 with a 4K camera at 110 g; the MV-2P brings full 640×512 at 130 g. On sub-250 g builds those 20 grams matter.
- Standoff distance decides. Under roughly 100 m, 256×192 saves weight and budget. At range, 640×512 with no debate.
On this page
What Resolution Actually Changes
More pixels on target at the same distance. That is the entire mechanism, and everything else follows from it.
A person at 300 m fills roughly four times the pixels on a 640×512 sensor as on a 256×192 one — the pixel count is 6.7 times greater overall, and the linear dimension roughly 2.5 times, so the target subtends proportionally more of the frame in each direction. Operationally that is the difference between a warm blob and “human, walking, carrying something”.
The DRI thresholds formalise this: detection needs a handful of pixels across the target, recognition roughly double that, identification double again. Doubling linear resolution moves you a full step up that ladder at any given range — see DRI ranges explained for the model and the two competing standards.
Note that pixel pitch is 12 µm on both across the current line, so this is a pure array-size difference rather than a generational one — why 12 µm replaced 17 µm covers the pitch axis separately.
Range Comparison for a Person-Sized Target
With a 25 mm lens, a 640×512 core detects a person at roughly 1,000 m. A 256×192 core with typical micro-payload optics works comfortably inside 200–300 m. Numbers shift with lens choice and conditions, and both figures are clear-air laboratory-style estimates rather than field guarantees.
The lens caveat is not a footnote. Focal length is the other half of the equation, and a 640 core behind a wide lens can easily underperform a 256 core behind a narrow one at distance — thermal lens focal length covers that trade. Resolution and optics have to be read together.
Conditions apply the second discount. Humidity, rain and fog attenuate, and real-world planning applies a 30–50% margin to laboratory figures. For a worked example against actual hardware, how far a drone thermal camera detects a person runs the numbers across six payloads under both DRI standards.
Compare like with like. A resolution comparison only means something at a stated lens and a stated DRI criterion. Two payloads quoted at different focal lengths under different standards are not comparable, whatever their sensor sizes.
Weight and Platform Reality
The published figures make the trade concrete, and the gap is smaller than most people expect.
| Payload | Weight | Thermal | Published NETD | Visible |
|---|---|---|---|---|
| MV-2M | 110 g | 256×192, 12 µm | <50 mK @ 25 °C | 3840×2160, 8.29 MP |
| MV-2P | 130 g | 640×512, 12 µm | <40 mK @ 25 °C | 3840×2160, 8.29 MP |
Twenty grams separates them, and the MV-2P is also the more sensitive core at under 40 mK against under 50 mK. On most aircraft that difference is irrelevant and the MV-2P is simply the better payload. On sub-250 g builds those 20 grams compete directly with battery, and the calculus changes — see payload weight versus flight time.
Cost moves too, and by more than weight. Stepping from 256×192 to 640×512 multiplies core cost, since array size drives yield — what drives payload price covers the six drivers behind a quote.
The Decision Rule
Standoff distance decides. Everything else is secondary.
| Working distance | Choice | Typical missions |
|---|---|---|
| Under ~100 m | 256×192 | Roof surveys, facade QA, close equipment checks, small-site work |
| 100–300 m | Either — check the lens | Building envelope, construction QA, distribution substations |
| Above ~300 m | 640×512 | Search, patrol, corridor inspection, large sites, anything at range |
Search, patrol, corridor inspection or anything at range: 640×512, no debate. The extra pixels are the mission. SAR sweeps and pipeline patrol both live entirely in this band.
Under roughly 100 m working distance, 256×192 saves weight and budget without costing findings, because the target already fills plenty of pixels. Roof moisture surveys and close-range envelope work are the archetypes.
Two Honest Caveats
Resolution is not sensitivity. A 640 core with poor NETD behind slow optics can find less than a well-specified 256 one, because the pixels may be there while the temperature difference is below the noise floor. For marginal-contrast work — a barely-warm connection, a person on sun-warmed ground — NETD is the binding constraint rather than pixel count, as thermal sensitivity explained covers.
Resolution does not fix the mount. Above modest zoom, gimbal stability rather than sensor quality caps what you can use, and a high-resolution core on a mediocre gimbal delivers blurred pixels at exactly the range you bought it for — what ±0.01° means at 1,000 metres has the arithmetic.
The practical reading: resolution is the first thing to specify and rarely the only thing that decides the outcome. What buyers should ignore first covers the specifications that consume attention without changing results.
Related reading
- From NETD to DRI
- How Uncooled VOx Microbolometers Work
- DRI Ranges Explained
- How Far Can a Drone Thermal Camera Detect a Person?
- Thermal Lens Focal Length on Drones
- Thermal Sensitivity (NETD) Explained
- Thermal Pixel Pitch: Why 12 µm Replaced 17 µm
- Sub-250 g Drones, Real Thermal
- MV-2M — 110 g, 256×192 thermal
- MV-2P — 130 g, 640×512 thermal
- OP-90A — 640×512 with zoom and ranging
- Thermal Drone Payloads for Search and Rescue
FAQ
Is 640×512 always better?
Optically yes, operationally no. If your platform cannot lift it or the mission never leaves close range, the extra resolution is unused mass and cost. Under roughly 100 m working distance a 256×192 core already puts plenty of pixels on the target, and the 20 grams and the price difference buy more useful capability elsewhere. Above 300 m the answer flips completely.
Does digital zoom fix low resolution?
No. Digital zoom crops and enlarges pixels that already exist — it adds no information the sensor did not capture. Enlarging a 256×192 frame produces a bigger image of the same detail, which helps framing on a small screen and does nothing for detection range. Only optical magnification or a larger array adds real pixels on target.
Should I choose resolution or sensitivity if I can only improve one?
It depends on contrast. For high-contrast targets — a person against cool ground, a hot connector against ambient metal — resolution dominates, because the signal is strong and the limit is how many pixels land on it. For marginal contrast, sensitivity binds first: pixels that cannot resolve the temperature difference do not help. Match the choice to whether your typical finding is a strong or weak thermal signal.
Questions about the technology? Talk to our engineers — we reply within 2 business days.