Skip to main content
Reference · Security Infrastructure

CCTV & Command Centre Design Standards

Camera count is the wrong specification. This reference sets out the pixel-density, retention, network and operator-workload standards that decide whether a control room produces evidence or just storage.

Updated 11 August 2026 12 min read
Pixel density, not camera count
Specify by
Alarms per hour, not screens per wall
Operator load
Coverage drawings + retention calculation
Deliverable

Key takeaways

The benchmarks on this page in citable form. Each one links to the table or checklist it comes from and to the FAQ answer that states it in prose — quote either, with attribution to arifkhanglobal.com/resources/cctv-command-centre-standards.

  1. 01~250 px/m for identification

    Identification-grade coverage needs about 250 pixels per metre; recognition about 125 px/m, observation 80 px/m, detection 40 px/m and general monitoring 25 px/m.

  2. 02Retention in days, not terabytes

    State storage as retention days at a defined resolution, frame rate and codec, sized on realistic scene complexity with headroom for the phase-two camera count.

  3. 032–3 desk monitors per operator

    Give each operator two to three desk monitors and design for fewer than ten actionable alarms per hour; more screens and more alarms both reduce assessment quality.

  4. 04150–300 lux, NC 30–35

    Control rooms need indirect ambient light of 150–300 lux with no screen reflections, absorptive finishes at NC 30–35, and break rotation on continuous monitoring shifts.

  5. 05Every alarm maps to a written response

    Operational success is measured: alarms by type, nuisance rate, mean time to assess and resolve, quarterly camera-health audits and an AMC tied to availability.

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.

Pixel-density bands used for coverage drawings (aligned with common industry practice).
PurposeTarget pixel densityTypical use
Monitoring~25 px/mCrowd flow, general area awareness
Detection~40 px/mPresence of a person in a zone
Observation~80 px/mBehaviour, clothing, direction
Recognition~125 px/mRecognising a known individual
Identification~250 px/mEvidential identification of a stranger
Licence plate (ANPR)~150–200 px/m within the capture zoneLane-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

ParameterBenchmarkNote
Operator viewing distance to video wallScreen diagonal × 4–6 for situational awarenessSized after content, not after wall width
Desk monitors per operator2–3 maximumMore screens reduce, not increase, alarm handling
Sustained alarms per operatorDesign under 10 actionable alarms per hourAbove this, assessment quality drops sharply
Ambient lighting150–300 lux, indirect, no reflections on screensControl rooms are lit wrong more often than not
AcousticsAbsorptive finishes; NC 30–35Shift-long fatigue is an operational risk
Shift designBreak rotation for continuous monitoring rolesAttention 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.

Planning a project where these benchmarks matter? Bring us in at design stage.

Let's Talk