Specify coverage, not cameras
A tender that says “96 cameras” specifies a budget, not a security outcome. A tender that says “identification-grade coverage at every entry point and monitoring-grade coverage across the car park” specifies an outcome any competent integrator can price and any consultant can verify.
| Purpose | Target pixel density | Typical use |
|---|---|---|
| Monitoring | ~25 px/m | Crowd flow, general area awareness |
| Detection | ~40 px/m | Presence of a person in a zone |
| Observation | ~80 px/m | Behaviour, clothing, direction |
| Recognition | ~125 px/m | Recognising a known individual |
| Identification | ~250 px/m | Evidential identification of a stranger |
| Licence plate (ANPR) | ~150–200 px/m within the capture zone | Lane-specific, with angle and speed limits |
Retention and storage maths
- State retention in days at a defined frame rate, resolution and codec — not in terabytes.
- Size storage on realistic scene complexity; smart-codec savings collapse on moving foliage and rain.
- Add headroom for the phase-two camera count that everyone knows is coming.
- Specify RAID level, rebuild time and failure behaviour, not just capacity.
- Confirm the export workflow produces court-admissible, watermarked evidence with an audit trail.
Network and VMS architecture
- Separate, VLAN-segmented surveillance network with no unmanaged switches in the path.
- Multicast or dual-stream design so the video wall does not saturate the recorder's uplink.
- Hardened VMS: unique credentials per operator, role-based access, no shared admin login.
- Time synchronisation from a single trusted source — evidence hinges on timestamps.
- Documented failover: what happens to recording when the head end, the link or the power dies.
- Analytics licensed with GPU headroom for the next phase, not exactly to today's channel count.
Command centre ergonomics
| Parameter | Benchmark | Note |
|---|---|---|
| Operator viewing distance to video wall | Screen diagonal × 4–6 for situational awareness | Sized after content, not after wall width |
| Desk monitors per operator | 2–3 maximum | More screens reduce, not increase, alarm handling |
| Sustained alarms per operator | Design under 10 actionable alarms per hour | Above this, assessment quality drops sharply |
| Ambient lighting | 150–300 lux, indirect, no reflections on screens | Control rooms are lit wrong more often than not |
| Acoustics | Absorptive finishes; NC 30–35 | Shift-long fatigue is an operational risk |
| Shift design | Break rotation for continuous monitoring roles | Attention degrades well before the shift ends |
SOPs and measurable operations
- Every alarm type maps to a written response, an escalation path and a closure code.
- Monthly report: alarms by type, nuisance rate, mean time to assess, mean time to resolve.
- Quarterly camera-health audit — focus, obstruction, IR, retention verification.
- Annual tabletop exercise with the response team and estate management.
- AMC tied to system availability and health-audit results, not just to visit counts.
Frequently asked
- What pixel density is required for identification-grade CCTV?
- About 250 pixels per metre across the target's width within the capture zone. Recognition of a known person needs roughly 125 px/m, observation about 80 px/m and general monitoring around 25 px/m.
- How should CCTV storage be specified?
- In retention days at a stated resolution, frame rate and codec, sized on realistic scene complexity with headroom for the next phase — never as a bare terabyte figure, which hides all the assumptions that matter.
- How many monitors should a control-room operator have?
- Two to three. Beyond that, alarm handling quality falls. The video wall is for shared situational awareness; the operator works from the desk.
- What makes a command centre operationally successful?
- Written SOPs per alarm type, a nuisance-alarm rate low enough that operators still assess every event, measured mean time to assess, and an AMC contract tied to availability rather than to visit counts.
Cite this reference
Arif Khan, “CCTV & Command Centre Design Standards”, arifkhanglobal.com, 2026.
Free to quote and republish in part with a link to arifkhanglobal.com/resources/cctv-command-centre-standards.