IP cameras

A camera installation is judged on the day something has to be found. That is when the three usual faults appear: the camera did not cover the right angle, the recording had already been overwritten, or nobody knew how to export the clip.

So we design from the question "what will you need to be able to show, and to whom". A camera filming a general view of a walkway is decorative; a camera positioned to read a face at the entrance or a plate at the barrier is useful. The difference is placement and lens, not pixel count.

Retention is the second point, and the most commonly under-specified. An incident is sometimes discovered three weeks later; a system cycling over seven days will have kept nothing. We calculate storage from the retention you want, not the other way round.

Finally, cameras go on a separate network. They consume bandwidth continuously and are a frequent way in when left on factory settings. Isolating them costs little at installation and avoids an entire class of problem.

Power cuts are the leading cause of missing footage. A recorder that restarts during the incident keeps nothing, and nobody notices until the images are needed. Backup power on the recorder costs little and changes everything.

Heat and dust shorten equipment life far faster than the stated figures suggest. A recorder placed in an unventilated room fails after a few summers, and the disk holding the history fails with it.

Finally, parts availability shapes the choice of brand. A camera whose replacement cannot be found here leaves a gap in coverage for weeks. We choose from what can be re-bought locally, as with the rest of the hardware.