Camera coverage
Choosing a lens is two questions, not one: how much scene fits in the frame, and whether there are enough pixels on a subject to be worth anything afterwards. Covers focal length against sensor format, and the DORI pixel-density bands that decide whether footage identifies somebody or merely shows that somebody was there.
Covering the room is not the same as identifying a face in it.
The angle from the camera's datasheet. Use this whenever you have it.
Nominal — take it from the datasheet.
Recognise
Enough to say with a high degree of certainty that this is somebody already known to the viewer.
Each band holds out to its distance and no further. This is the figure to design against — put the door inside the identify range, not merely inside the frame.
- Identify≥ 250 px/mto 5.7 m — not reached
- Recognise≥ 125 px/mto 11.4 m — reached at the selected distance
- Observe≥ 62.5 px/mto 22.7 m — reached at the selected distance
- Detect≥ 25 px/mto 56.9 m — reached at the selected distance
Two things this arithmetic cannot tell you. It assumes the subject is square-on and evenly lit — a face at 45° or backlit against a doorway needs considerably more pixels than the band suggests. And it says nothing about frame rate: identification from a still is one problem, following somebody across a lobby is another.
- What focal length do I need for a camera?
- It depends on the sensor and the distance, not on the lens alone. A 4 mm lens on a 1/2.8-inch sensor gives roughly a 65° view; the same lens on a smaller sensor is narrower. Work from the distance to the subject and the pixel density you need at that distance.
- What is DORI?
- Detect, Observe, Recognise, Identify — the pixel-density bands from IEC/EN 62676-4, measured in pixels per metre of scene width. Detect needs 25 px/m, Observe 62.5, Recognise 125, and Identify 250. They are the standard the industry quotes when it claims a camera can identify at a distance.
- Does more megapixels mean I can identify further away?
- Yes, proportionally — pixel density scales directly with horizontal resolution at a given field of view. But it also multiplies the storage and the bandwidth, and it does nothing about a subject that is backlit, at an angle, or moving too fast for the shutter.
- Why does my camera cover the area but the footage is useless?
- Because coverage and pixel density are different measurements. A wide lens fits the whole car park in frame while putting perhaps 20 pixels across a face — enough to say a person was present, and nothing more. Design against identify distance, not against frame coverage.
The rest of the section
- 01RJ45 Wiring: T568A and T568BBoth colour codes, pin by pin, side by side.
- 02RJ11 Wiring and Phone PinoutsSix positions, three lines, two colour systems.
- 03Keystone Jack TerminationStrip, seat, punch, snap — without failing the certification.
- 0466 Block WiringFifty rows, four columns, and where the bridging clips go.
- 05110 Block WiringLay the pairs in, press the connecting block on.
- 06Cat5e vs Cat6 vs Cat6a vs Cat8Speed, distance, and power decide it — not the bigger number.
- 07Fibre Types, Connectors and PolishRead the jacket, read the connector, don't mate blue to green.
- 08Subnetting, ExplainedWhere the numbers come from, and the meaning behind them.
- 09PoE Standards and Why It Won't PowerThe switch reserves the class, not the draw.
- 10Patch Panels and Labelling to TIA-606A label that survives is worth more than a tidy rack.
- 11Rack Planning: Units, Depth, Weight and HeatIt fits vertically. That was never the hard part.
- 12Writing a Switch ConfigurationFrom factory default to a switch you would hand over.
- 13Networking and Cabling GlossaryThe terms, and what they mean on a job.
We do this work
as well as document it
Anyone siting cameras and needing footage that will still be useful when somebody asks who that was.