A thermal camera module purchase almost always starts with a quick scan of nine headline specs; the teams who present these clearly—and honestly—win RFQs faster, cut back-and-forth emails, and prevent costly mismatches in the field. This guide explains what each spec really means in use, the trade-offs behind the numbers, and how to present them so engineers, procurement, and operators can make a confident choice.
How decision-makers actually read a thermal camera module spec sheet
Before we dive into the list, remember how buyers consume specs in B2B:
- They skim for fit/no-fit in under a minute (FOV/resolution, frame rate, interfaces, power).
- They return for risk (NETD at the real fps/f/#, shutter policy, latency, thermal behavior).
- They finally ask for proof (sample clips at target ranges, MRTD/MTF plots, acceptance tests).
Structure your page for all three moments. Put the nine items below into a single, predictable block and link out once to your deeper engineering notes.
1) Resolution & Sensor Format
What buyers think: “How many pixels do I get—and do they land on my target at my distances?”
What it really controls: Sampling opportunity. Resolution (e.g., 256×192, 384×288, 640×512) plus sensor width determines how wide the scene is for a given focal length and how many pixels span a target at range. Resolution alone doesn’t guarantee recognition; it must be paired with the right IFOV and MTF.
What to publish (clean & useful):
- Array (e.g., 640×512, LWIR).
- Pixel pitch (e.g., 12 µm or 17 µm).
- Active area (mm × mm) and diagonal.
- Suggested HFOV families with common lenses (e.g., 12°, 18°, 32° with 50/35/19 mm at 12 µm).
- One line of pixels-on-target examples (e.g., “0.5 m torso at 350 m ≈ 8–10 px with 12° HFOV”).
How to present: a tiny table labeled Sampling at Range (illustrative) with two target sizes and three distances. Keep numbers round; your goal is orientation, not simulation.
2) Pixel Pitch & IFOV
What buyers think: “Is this 12 µm or 17 µm, and what does it do to my lens and gimbal?”
What it really controls: Angular detail per pixel. IFOV ≈ pixel pitch ÷ focal length (radians). For the same FOV, smaller pitch yields a shorter focal length (smaller, lighter, cheaper lens)—a huge win on UAVs and PTZ heads. For the same focal length, smaller pitch increases pixels on target.
What to publish:
- IFOV values for your standard lenses (mrad/px).
- A one-line optics impact note: “12 µm at 25 mm ≈ 17 µm at 35 mm HFOV—smaller germanium, lighter gimbal load.”
How to present: a one-row comparison bar: Lens size/weight implication for 12 µm vs 17 µm at the same HFOV.
3) NETD (at Your FPS and f/#)
What buyers think: “Will faint contrast survive noise on hot nights and long paths?”
What it really controls: Scalar sensitivity—how small a ΔT rises above noise. The context (frame rate, optics f/#, FPA temperature) decides whether your advertised NETD is reproducible in their use.
What to publish (no wiggle):
- System NETD (e.g., ≤40 mK @30 fps, f/1.0, 25 °C).
- If you also publish detector-only NETD, label it clearly.
- A thumbnail MRTD curve screenshot or link in your docs (system-level is best).
- One sentence on NUC cadence that holds that NETD in flight/field.
How to present: a single bold line with the “@” conditions; avoid vague “typical” icons.
4) Lens, FOV & f/# (with MTF and Focus Behavior)
What buyers think: “Which FOVs exist, how fast is the glass, and will focus stay put?”
What it really controls: Contrast at the spatial frequencies where recognition/ID live. f/# affects irradiance and effective SNR; MTF tells how much of that contrast survives; focus discipline determines whether you keep it.
What to publish:
- Standard HFOV choices and matching focal lengths.
- f/# per lens and whether focus is fixed, manual, or motorized.
- A small MTF at f/# plot or a promise to supply it on request.
- Optional: DLC coatings / window notes (durability).
How to present: one tidy table: FOV → focal length → f/# → focus method → mass.
5) Frame Rate, Pipeline Latency & Encoder
What buyers think: “Can I steer a gimbal in real time? Will low-contrast details survive the codec?”
What it really controls: Motion handling and operator trust. UAVs and PTZ users care about glass-to-glass latency and bitrate floors; security VMS teams care about RTSP/ONVIF and GOP settings.
What to publish:
- Sensor frame rates (e.g., 30/50/60 fps) and any export-class variants.
- End-to-end latency (e.g., 120–180 ms @60 fps, H.265 main).
- Encoder profiles (H.264/H.265, selectable bitrate ranges), plus a conservative recommended bitrate for long-range tasks.
- Recording options (native stream + radiometric stills if applicable).
How to present: bullets with one crisp “Search” preset and one “Overwatch” preset (bitrate, GOP, tone curve), so procurement can copy/paste into acceptance.
6) NUC Policy (Shuttered/Shutterless/Hybrid) & Uniformity
What buyers think: “Will the picture freeze mid-mission? Will striping creep in after temperature swings?”
What it really controls: Continuity, operator workload, and radiometric trust.
What to publish:
- Default NUC mode (e.g., “Hybrid: shutterless continuous + timed shutter at take-off/landing or on command”).
- Freeze-time budget if shutters exist (e.g., ≤0.8 s, no more than once per 20 min in flight).
- Residual FPN metric on a flat-field test (counts or mK).
- Telemetry/API:
nuc_mode,nuc_trigger(),fpa_temp.
How to present: a two-row table—Continuity and Radiometry—with a check mark beside your default profile.
7) Interfaces, SDK & Protocols
What buyers think: “Will it drop into my avionics/robot stack without an adapter circus?”
What it really controls: Integration time and field reliability.
What to publish:
- Physical: USB 3, GigE/PoE, MIPI CSI-2, LVDS, UART, I²C; connector models/pinouts; video formats.
- Network/Video: RTSP/ONVIF, RTP multicast, MJPEG, raw/YUV/BT.656 if applicable.
- Control SDKs: C/C++, Python; OS support; example code for gain/NUC/palette/focus/encoder.
- Time sync: PPS / PTP / MAVLink messages (if UAV).
- Security: user auth, TLS/HTTPS for control endpoints, signed firmware.
How to present: a clean two-column list—Video out and Control—with one code snippet (10–12 lines) showing how to set NUC mode and encoder bitrate.
8) Size, Weight, Power & Thermal Design (SWaP-T)
What buyers think: “Will my gimbal and battery like this?”
What it really controls: Platform compatibility, endurance, and wind performance.
What to publish:
- Module envelope (L×W×H), mass (with typical lens).
- Power: typical/peak draw (W), input range, inrush profile.
- Heat: steady-state case temperature rise at 25 °C ambient, airflow needed (if any).
- Gimbal class compatibility.
- Accessory mass (lens A/B/C, window options, mounts).
How to present: a one-line SWaP-T summary plus a small diagram call-out with bolt pattern and CG location; UAV teams will thank you.
9) Compliance, Reliability & Operating Conditions
What buyers think: “Will my QA and lawyers sign this?”
What it really controls: Time-to-deploy and total cost of ownership.
What to publish:
- Electrical/EMC/Safety: CE, FCC, RoHS/REACH; surge/ESD notes.
- Environmental: operating/storage temp ranges; IP rating when housed; vibration/shock test methods (IEC/DO-160 where relevant).
- Radiometric: calibration traceability (blackbody reference, ΔT drift over temp).
- Firmware: versioning, rollback, long-term support policy.
- Export: HTS code, EAR/dual-use notes (if applicable).
How to present: a tight compliance grid with cert numbers and test methods. Add a “Quality & lifecycle” paragraph: spare parts policy, lens availability, and typical lead times.
The quick-glance table buyers actually use
| Spec (the “9”) | What buyers check first | What to show on page | Common pitfalls |
|---|---|---|---|
| Resolution & format | 256/384/640 and aspect | Array, pitch, active area, HFOV families | Publishing resolution without FOV context |
| Pixel pitch & IFOV | 12 µm vs 17 µm | IFOV per standard lenses; lens size impact | Hiding lens mass/cost effects |
| NETD @ fps/f/# | ≤50 mK? at their fps | System NETD with conditions; MRTD snapshot | Detector-only numbers; undefined conditions |
| Lens/FOV/f/#/MTF | Narrow vs wide; f/1.0 vs f/1.2 | FOV→focal length→f/#→focus→mass; MTF note | Ignoring focus and window impact on MTF |
| Frame rate & latency | 30/50/60 fps; ms to glass | Latency @fps/codec; Search/Overwatch presets | Bitrate too low for long-range detail |
| NUC policy & uniformity | Freeze? Striping? | Default mode; freeze-time; residual FPN; API | No guardrails for uniform scenes |
| Interfaces & SDK | Plug-in effort | Physical + protocols; example code | Exotic connectors; missing pinout |
| SWaP-T | Mass/power budget | L×W×H, mass (with lens), W; CG/bolt pattern | Only “module-only” mass without lens |
| Compliance & reliability | Certs & temp | CE/FCC/RoHS; temp; IP; shock/vibe; LTS | Vague claims; no test method IDs |
How to write spec copy that closes RFQs
Bad: “NETD 35 mK (typ.).”
Good: “System NETD ≤35 mK @30 fps, f/1.0, 25 °C; MRTD curve available; hybrid NUC holds uniformity across ±10 °C FPA drift.”
Bad: “Supports RTSP.”
Good: “RTSP (H.265/H.264) with bitrate presets: Search 8–12 Mb/s, Overwatch 14–20 Mb/s; ONVIF Profile-S; SDK exposes encoder profile and GOP.”
Bad: “Shutterless.”
Good: “Shutterless default with motion-registered NUC; freeze-free in flight; user Calibrate button for uniform scenes; logged NUC events.”
Bad: “Lightweight.”
Good: “210 g module with 19 mm lens (12 µm pitch); 320 g with 35 mm; CG 22 mm above base; 6-hole pattern 25 mm square.”
Mini case studies: which spec actually decided the deal?
A. Municipal UAV (search & rescue).
They skimmed resolution first, but the decider was latency + NUC policy. The winning module posted ≤150 ms glass-to-glass and a freeze-free shutterless profile with an operator Calibrate button. Sample clips sealed it.
B. Refinery perimeter (corridors).
Resolution was a tie; the decider was lens/MTF/focus at long focal lengths, and the vendor who published IFOV math and a simple recognition acceptance test (bar target at 350 m, windy day) closed the PO.
C. Maritime patrol (uniform horizons).
Everyone claimed shutterless. Only one vendor documented the fallback when scene diversity drops (single-image NUC + low-frequency shutter in hover), plus a small residual FPN chart. They won on risk control.
Buyer-ready acceptance tests you can paste into a PRD
- System NETD: ≤X mK @30 fps, f/1.0, 25 °C; verify with calibrated source after 30 min warm-up.
- MRTD: deliver curve at 20/30/40 °C FPA, measured with shipping lens and display pipeline.
- IFOV & FOV: provide drawing + measured HFOV (±0.3°) and a pixels-on-target check at 300 m.
- MTF & focus: confirm MTF50 at specified spatial frequency; no more than Q% drop over 10 °C ambient change.
- Latency: glass-to-glass ≤N ms @60 fps; encoder Search and Overwatch presets fixed.
- NUC policy: no more than one freeze ≤0.8 s per 20 min flight; residual FPN ≤R mK on uniform panel.
- SWaP-T: mass with 25/35 mm lenses; power ≤P W; case rise ≤T °C at 25 °C ambient without forced air.
- Compliance: CE/FCC pass, operating −20…+60 °C (or your spec), shock/vibration per method ID.
Page architecture: where each of the nine lives
- Hero: one-line promise + Search/Overwatch encoder presets.
- Spec Block: these 9 specs in the same order every time.
- Clips: 3 short videos—wide search, narrow clarify, humid night.
- Engineering Notes: link to deeper docs once.
- CTA: schedule a spec workshop with internal links.
To keep your internal linking tidy and unique within the article, use each link exactly once:
- Explore modules: <a href=”/thermal-camera-module/”>Thermal camera module</a>
- Build details: <a href=”/thermal-camera-module-integration/”>Thermal camera module integration</a>
- Mixed sensing: <a href=”/laser-rangefinder-module/”>Laser Rangefinder Modules</a>
- Partnership path: <a href=”/oem-odm-partner-program/”>OEM/ODM Partner Program</a>
- Talk to an engineer: <a href=”/contact/”>contact us</a>
FAQs (concise, buyer-oriented)
Is 640 always better than 384 on a thermal camera module?
Only when IFOV and MTF support it and your gimbal/bitrate preserve detail. A tuned 384 can beat a shaky 640.
Why do vendors’ NETD numbers feel incomparable?
Because conditions differ. Always read “@fps/f/#/temp” and ask for system NETD plus a MRTD snapshot.
What single mistake causes the most field disappointment?
Publishing resolution without FOV/IFOV and MTF/focus context—operators get pixels but not usable detail.
Do I need shutterless for UAVs?
Usually yes (continuity), but a hybrid with rare, timed shutters handles uniform scenes and radiometric drift better.
Call to Action
If you’re shortlisting a thermal camera module and want specs that hold up in flight tests and design reviews, we’ll map your targets and distances into IFOV math, fix encoder presets, and propose lens/NUC options with acceptance tests you can paste into the PO. Start with our Thermal camera module, dive into Thermal camera module integration, consider ranging overlays via Laser Rangefinder Modules, align with the OEM/ODM Partner Program, then contact us to schedule a 30-minute spec workshop.
