FOV Calculator

HFOV, VFOV and IFOV — the math behind every CCTV camera

A camera's field of view is determined by exactly two physical quantities: the imaging sensor's active dimensions and the lens focal length. Everything else — megapixels, codec, branding, mount type — is irrelevant to the angular coverage. The horizontal field of view (HFOV) is HFOV = 2 × arctan(W / 2f), where W is the sensor active width in millimetres and f is the lens focal length in millimetres. The vertical field of view (VFOV) is the same formula with the sensor height H instead of W. The diagonal FOV uses the sensor diagonal. Most CCTV spec sheets quote only HFOV; the calculator above derives it from the focal length and sensor preset you choose.

Aspect ratio matters because sensor width and sensor height are not independent. A modern 16:9 CMOS like the Sony IMX415 has an active area of 5.6 × 3.1 mm (1/2.8" optical format). With a 4 mm lens that gives HFOV ≈ 70° but VFOV ≈ 42°. A 4:3 sensor of equivalent diagonal would give HFOV ≈ 64° and VFOV ≈ 50°. Specifying "wide angle" without saying which axis is wide is ambiguous: the same lens looks dramatically different on 16:9 vs 4:3 sensors.

Sensor sizes in CCTV are inherited from vidicon-tube nomenclature and almost never match the literal fraction. A "1/3 inch" sensor is roughly 4.8 mm wide, a "1/2.7 inch" is 5.0 mm, a "1/2.8 inch" is 5.4 mm, a "1/2 inch" is 6.4 mm, a "2/3 inch" is 8.8 mm, and a "1 inch" is 12.8 mm. Using the literal inch-fraction in any FOV formula will overestimate angular coverage by 30–60%. Always look up the active mm width from the sensor datasheet, or rely on the well-known presets in the calculator above.

IFOV — the instantaneous field of view, or angular size of a single pixel — is what actually determines whether you can resolve a face or read a plate. IFOV in milliradians is roughly 1000 × pixel_pitch / focal_length. A 1/2.8" 4 MP sensor has a pixel pitch of about 2.0 µm; with a 4 mm lens that is 0.5 mrad/px, or roughly one pixel per 2 mm at 4 m distance. Multiply by the required pixels-on-target (typically 200 px on a face for identification) and you have the maximum identification range without invoking full DORI math.

Corridor mode rotates the sensor 90° so that the long axis runs vertically — useful for hallways, escalators, and narrow aisles. The HFOV and VFOV swap places in the firmware, and the camera produces a portrait-orientation video stream. The VMS must support 9:16 layout to display it correctly. Multi-sensor and panoramic cameras stitch overlapping frames from two to eight sensor-lens pairs to produce a continuous wide field — typical effective HFOV is 180° or 360°, but resolution at the seams drops noticeably and pixel density per metre at distance is no better than a single sensor of the same MP count.

Tilt correction matters for any camera not aimed perfectly horizontal. If a camera at 4 m height looks down at a target on the ground 10 m away, the slant range is √(4² + 10²) = 10.77 m, not 10 m. The vertical FOV captures both close and far ground simultaneously, so the pixel density varies dramatically along the projected footprint. Most planning errors trace back to engineers ignoring this and assuming a clean rectangular ground footprint with uniform PPM.

How to use this FOV calculator

1. Pick the sensor format. The four presets cover almost every fixed-lens camera shipped today: 1/3" for entry-level bullets, 1/2" for mid-tier turrets and most 4 MP cameras, 2/3" for premium box and PTZ models, and 1" for low-light specialist sensors. The mm width is filled in automatically.
2. Set the focal length. Drag the slider for any value between 1 and 50 mm, or click a popular preset (2.8, 3.6, 4, 6, 8, 12, 16, 25, 35, 50). For a varifocal lens, evaluate the calculator at both ends of the zoom range.
3. Set the target distance. This is the horizontal distance from the camera to the plane of interest — the gate, the shelf row, the parking bay edge. Use metres or feet according to your unit preference. Coverage width at that distance is computed live below.
4. Read the two output cards. The first shows the horizontal angular FOV in degrees — useful when comparing against manufacturer marketing. The second shows the linear scene width covered at your chosen distance — useful when comparing against the physical area you need to monitor.

Worked example: parking lot ANPR

A retail park manager wants automatic number-plate recognition (ANPR) at the single-lane vehicle entrance. The lane is 3.5 m wide, and the camera will be mounted at 4 m on a pole positioned 12 m from the read line. The vehicle has to be travelling slow enough to be identifiable — assume 10 km/h or less, which gives a shutter budget of about 1/250 s.

Start with a 4 MP camera (2560 horizontal pixels) on a 1/2.8" sensor. To read a European plate (520 mm wide) reliably, you need at least 250 PPM on the plate plane — equivalent to about 130 pixels across the plate itself. Plug a 4 mm lens into the calculator: HFOV ≈ 68°, scene width at 12 m ≈ 16.2 m. That spreads 2560 pixels across 16.2 m, giving only 158 PPM — well short of the 250 PPM needed.

Step up to an 8 mm lens. HFOV drops to 37.4°, scene width at 12 m becomes 8.1 m, and pixel density rises to 316 PPM — comfortably above the 250 PPM identification floor. The 8.1 m horizontal coverage easily contains the 3.5 m lane plus margin. The slant range from the 4 m mounting height to the 12 m read line is √(4² + 12²) = 12.65 m, so the effective PPM at the slant target plane is closer to 300, still well above threshold.

A 12 mm lens would give 474 PPM — overkill for a single lane and too narrow to catch plates if a vehicle stops slightly to one side. The 8 mm lens is the right choice. The same calculation also reveals why "any 4 MP camera" is not enough: a 4 mm lens at 12 m simply does not put enough pixels on the plate, regardless of how the camera is marketed.

Common FOV mistakes

• Using the literal inch fraction as sensor width. A 1/2.8" sensor is not 1/2.8 inch (9 mm) wide — it is 5.4 mm. Using the wrong width makes every FOV value 30–60% too wide and every distance estimate too optimistic.
• Quoting HFOV when the install needs VFOV. Hallways and corridors care about vertical coverage, not horizontal. Either rotate to corridor mode or compute VFOV explicitly. The default spec-sheet HFOV value is irrelevant for vertical-axis applications.
• Ignoring tilt and slant range. A camera at 4 m height aiming at a target on the ground 10 m away has a slant range of 10.77 m and the floor footprint is a trapezoid, not a rectangle. The simple horizontal-FOV math is exact only at the optical axis.
• Forgetting aspect ratio when mixing 16:9 and 4:3. A 4 mm lens on a 16:9 1/2.8" sensor gives 70° HFOV but only 42° VFOV. The same lens on a 4:3 sensor of equivalent diagonal gives 64° HFOV and 50° VFOV. Hardware mixed across formats produces inconsistent coverage even when "the lens is the same".
• Treating panoramic FOV as additive. A 4-sensor 360° camera does not give 4× the pixels at distance — it gives 1× the pixel density of a single sensor at any given range, just stitched across a wider azimuth. Use a panoramic camera for situational awareness, not for long-range identification.

Standards and compliance references

• EN 62676-4:2015 — Application guideline for video surveillance systems. Defines the DORI pixel-density framework that converts FOV into operational performance categories. EN 62676-4 calculator →
• IEC 62676-4:2025 (OODPCVS) — The 2025 international refresh that introduces corridor-mode pixel density (PPM_v) and AI-analytics-aware sub-tiers.
• NATO STANAG 4347 / Johnson Criteria — Cycles-on-target metric for thermal sensors, with 1.5 / 6 / 12 cycles for Detect / Recognize / Identify. Uses angular metrics rather than pixel counts. Johnson Criteria calculator →
• NDAA Section 889 — US procurement restriction on covered video equipment from listed manufacturers; orthogonal to FOV math but typically a tender prerequisite. NDAA compliance reference →
• IEC 61146-1 — Methods of measurement for video cameras: defines the formal procedures for measuring resolution, sensitivity, and angular coverage at the laboratory level.

Continue reading