Quick answer: Use the exact model/mode’s field of view and delivered pixels to estimate scene width and density at the required target plane. Where FOV is unknown, label any sensor-width calculation as an ideal assumption, not a product specification. Apply the project’s required standard edition and recorded task, then test the actual mounted day/night view. Geometry and historical DORI values do not guarantee identification or plate reading.
On this page
What decides the field of view: focal length and sensor size together
Focal length and the active image area influence field of view, along with the real optical/mode behavior. The simplified rectilinear model below can illustrate the tradeoff; exact horizontal/vertical FOV, active crop and delivered recording data are preferable to guessing active dimensions from an optical-format label. Use the updated pixel-density guide to separate a geometric estimate from the required viewing task.
Ideal horizontal angle = 2 × arctan(active sensor width ÷ (2 × focal length)). Ideal scene width = 2 × perpendicular target-plane distance × tan(horizontal angle ÷ 2). For the same ideal assumptions, that width equals distance × active width ÷ focal length. Keep sensor width and focal length in the same units.
Do not treat 1/3-, 1/2.7- or 1/1.8-inch labels as exact active widths for every model and mode. The examples below assume 5.4 mm solely for geometry. CS-I4KS L-2.8mm lists a 2.8 mm fixed lens and CS-I4KS L-2.8-12mm a motorized 2.8-12 mm lens, but neither lens row establishes the assumed active width. Obtain missing exact-revision FOV/mode data before using the estimate in a proposal.
Angle and density are design inputs, not coverage or evidence promises by themselves. Actual mounting, target pose, occlusions, motion, light, focus, compression and replay conditions affect the usable result across the scene. Agree the task and test it on recording, not only a live preview. Shortlist IP cameras for installers against those viewing requirements.
Identify the required edition before applying a pixel-density guide
The commonly quoted four DORI levels are the 2014-era model. IEC 62676-4:2025 is edition 2 and updates the operational categories; do not present the old values as unqualified current requirements. Record the actual project’s specified edition and human-viewing task. Analytics, thermal and plate tasks need their own criteria. Convert px/m to px/ft by dividing by about 3.28084 when helpful.
| Historical 2014-era label | Common planning density | How to use it |
|---|---|---|
| Detection | 25 px/m ≈7.6 px/ft | A historical guide, not guaranteed target detection |
| Observation | 62.5 px/m ≈19.1 px/ft | Check the real required recorded detail |
| Recognition | 125 px/m ≈38.1 px/ft | Not a guaranteed recognition result |
| Identification | 250 px/m ≈76.2 px/ft | Not an identification or current-edition compliance certificate |
Illumination range does not establish that a face or plate meets its task at that distance. Test the actual recorded target with representative motion, light, exposure, angle and compression; IR range explained covers the distinction. Do not translate a historical DORI number into a plate-reading promise or a conclusion beyond reasonable doubt.
Calculator: sensor, focal length, distance → scene width and pixels per foot
Pixel density = delivered horizontal recording pixels ÷ actual scene width, using consistent distance units. To explore a chosen geometric density, rearrange the ideal equations for distance, but label the result as an assumption-based threshold crossing. It is not a maximum identification distance or a substitute for actual FOV and recorded acceptance.
Worked example: a 4MP dome at an entrance
Illustrative calculation, not a measured CASE result: assume 2560 horizontal recorded pixels, a 2.8 mm lens and a 5.4 mm active width, with a perpendicular target plane 20 ft away. The catalog lists 2560×1440 for CS-I4KS L-2.8mm, but the active-width assumption has not been established for that SKU.
- Ideal scene width = 20×5.4÷2.8 ≈38.57 ft, about 11.76 m.
- Density = 2560÷38.57 ≈66.37 px/ft, about 217.75 px/m. This arithmetic alone does not classify a real image as recognition or identification.
- At the same assumed view, 3840 pixels would give about 99.56 px/ft. A narrower supported FOV or changed mounting may increase density, but each changes coverage and needs a recorded task test.
- Replace assumptions with exact mode/FOV and actual mounted geometry, then assess real lighting, movement and recording/replay before acceptance.
These results depend on assumptions and the delivered mode. CS-I4HS L-2.8mm lists 2688×1520 rather than 2560×1440; that is five percent more horizontal pixels at an identical view, not automatically five percent more usable evidence. Retain the exact product/mode inputs and separate calculated estimates from measured results. Check the IP camera user manual and request any missing exact-mode FOV or recording data.
Fixed 2.8 mm vs 4 mm vs motorized varifocal: which lens for which distance?
The table compares ideal angles and density at a 20 ft perpendicular target plane for a 5.4 mm assumed active width. Focal-length entries are geometry, not promises that a specific fixed or varifocal product delivers that angle, supports remote control or meets an identification task. Confirm actual lens/mode and view data.
| Assumed focal length | Ideal horizontal angle | Width at 20 ft | 2560 px: px/ft | 3840 px: px/ft |
|---|---|---|---|---|
| 2.8 mm | 87.9° | 38.57 ft | 66.37 | 99.56 |
| 4 mm | 68.0° | 27.00 ft | 94.81 | 142.22 |
| 12 mm | 25.4° | 9.00 ft | 284.44 | 426.67 |
| 13.5 mm | 22.6° | 8.00 ft | 320.00 | 480.00 |
A wider view can provide context while a narrower view can put more pixels on a target under the same assumptions. Neither focal length has a universal evidence-grade distance. A motorized varifocal can offer supported adjustment where available, but control, focus, range and required access depend on the exact model/client. Varifocal vs fixed lens and the lens angle cheat sheet cover the practical comparison without promising lift-free installation or service.
Actual height, tilt, target plane and pose affect vertical/horizontal coverage and usable target detail. A floor distance or slanted line to a face is not automatically the perpendicular distance in this simple model. Security camera mounting height and angle covers the survey and recorded-test questions; no fixed height alone approves the view.
Examples with CASE lenses: matching the printed lens to the job
These catalog examples identify exact lens/resolution candidates, not a verified generic sensor width, FOV, remote-control workflow or target distance. Obtain current revision/mode documentation and the required recorded test before committing a product to the camera schedule. Confirm the term, exclusions and claim process for the exact model and revision in writing. Low-light and color behavior, exposure and supported lighting modes need exact-model documents and real recorded scene tests, not an assumption from sensor size or illuminator reach.
- CS-I4KS L-2.8mm: 2.8 mm fixed lens and 2560×1440 at 25 fps. Confirm exact view and task, rather than assigning a seventeen-foot identification guarantee.
- CS-I4KS L-2.8-12mm: motorized 2.8-12 mm lens and 2560×1440 at 30 fps. Verify actual FOV and supported adjustment/focus; a generic 12 mm calculation does not establish a twenty-three-metre result.
- CS-I8MS SD-2.7-13.5mm: lens row 2.7-13.5 mm and 3840×2160 at 20 fps. A range printed in the lens row alone does not prove every motor/control behavior; confirm exact operation and recorded motion.
- CS-I4US L-2.8-12mm: motorized 2.8-12 mm lens and 2560×1440 at 30 fps, with an 80 m IR row. That range does not guarantee detection, tracking or identification at its end; test the required lighting/recording task.
What to tell your customer
We choose the view for the recorded detail you need at actual target positions. The calculation helps shortlist a lens, then we confirm exact model/mode FOV, mounting and supported controls. We test representative day/night recordings and replay/export; a density estimate or live picture alone is not the acceptance promise.
FAQ
How do I calculate a security camera’s field of view?
Where exact model/mode horizontal FOV is known, use width = 2×perpendicular target distance×tan(FOV/2). Otherwise an ideal rectilinear estimate uses active sensor width and focal length; mark any assumed dimensions. At 5.4 mm assumed width and 2.8 mm focal length, the ideal angle is about 87.9° and the 20 ft scene about 38.57 ft wide. Verify the actual recorded view.
How many pixels per foot do I need to identify a person?
Determine the specified edition and actual task first. The familiar 2014-era identification guide is 250 px/m, about 76.2 px/ft; it is not an identification guarantee or an unqualified current-edition requirement. IEC 62676-4:2025 updates the model. Assess actual recorded targets, viewing conditions and agreed acceptance, with separate criteria for analytics or plates.
Is a 2.8 mm or 4 mm lens better for security cameras?
Choose from the required view and recorded task. Under identical ideal active-width/resolution assumptions, 4 mm gives about 43% more horizontal density than 2.8 mm at the same target plane and a narrower view. Actual angles and performance depend on the exact model/mode. Do not assign fixed identification distances from those lens names.
When is a motorized varifocal lens worth specifying?
When supported lens adjustment helps meet a defined view/task or resolve approved mounting constraints. Confirm the exact range, FOV, motor/focus and client/control behavior and safe installation/service access. CS-I4KS L-2.8-12mm and CS-I4US L-2.8-12mm list motorized 2.8-12 mm lenses, but neither a fifteen-metre threshold nor a generic seventy-five-foot identification promise follows. Use the varifocal vs fixed lens guide to review the practical tradeoffs.
Does a longer IR range mean I can identify people farther away?
No. An illuminator row is not an evidence or identification distance, and usable detail depends on more than pixel density. CS-I4US L-2.8-12mm lists 80 m IR, but actual target, geometry, exposure, movement and recording determine the result. Do not promise detection, observation or tracking throughout the stated range either; test the required function.
References
- Anabon: exact model/revision documents
- IEC 62676-4:2025, edition 2
- Axis: pixel-density planning in the 2025 model
- Axis: separately identified 2014-era model
