Doubling the pixel count of a camera sounds like doubling its usefulness. On a plan it moves the DORI rings outward by a factor of about 1.4, since density scales with the width of the image and width scales with the square root of the pixel count. Whether that shows up in the recording depends on everything downstream of the sensor.

Pixels per metre, again

A 4 MP camera is 2560 pixels wide; an 8 MP camera is 3840. With the same lens the 8 MP image is 1.5 times denser, so a doorway that sat at recognise moves to identify, or the same identify range is reached with a lens 1.5 times wider. That is the whole benefit, and it is real.

Where it leaks away

  • The lens. Cheap lenses resolve less than the sensor behind them. An 8 MP sensor behind a lens made for 4 MP delivers 4 MP of detail in an 8 MP file.
  • Pixel size. The same sensor area divided into more pixels means smaller pixels, each catching less light. At night the 8 MP camera runs higher gain and the extra detail drowns in noise. The vendor's low-light figure tells you how much.
  • Compression. A recorder set to a fixed bit rate spends the same bits on the bigger frame; each pixel gets fewer. Faces at the edge of identify are exactly the fine detail the encoder discards first. An 8 MP stream needs roughly twice the bit rate of a 4 MP stream to keep the same quality.
  • Frame rate. Some cameras and most budget recorders drop to a lower frame rate at their top resolution. A number plate seen for one frame at 8 fps is a guess.

Storage and bandwidth

Bit rate scales with pixels, roughly. A site of thirty 4 MP cameras at 4 Mbit/s each stores about 1.3 TB a day; swapping them for 8 MP at 8 Mbit/s doubles that and doubles the switch load. If retention is fixed by policy, the extra resolution costs disks, and the client should choose that knowingly.

How to decide

Work backward from the rings. Draw the plan with the cheaper camera and see which zones fall short of the quality the client needs at the distances that matter. If the shortfall is at one doorway, a narrower lens or a second camera fixes it for less than upgrading thirty. If the shortfall is everywhere, the higher resolution earns its bit rate. Either way the drawing shows the decision, and the datasheet's megapixel figure is a starting point rather than an answer.