Buying a camera by megapixel count alone is the most common way people waste money and still get useless footage. Resolution and lens focal length work together to decide what you can actually recognize on screen. This guide shows you how to match both to the area you want to cover, so you can read a face at the gate or a plate on the driveway instead of just seeing a blurry shape.
Why resolution alone does not solve the problem
Resolution tells you how many pixels the sensor records. But those pixels are spread across the whole scene. A 4K camera pointed at a 20-meter-wide yard puts far fewer pixels on a face than a 2MP camera pointed at a 3-meter doorway. The metric that matters is pixel density on the target, often expressed as pixels per meter (PPM) or pixels per foot.
As a practical reference used across the CCTV industry, you generally need roughly:
- About 25 PPM to detect that a person is present.
- About 60-80 PPM to recognize someone you already know.
- About 120+ PPM to identify a stranger or read a license plate reliably.
These are working guidelines, not laws. Lighting, motion blur, and lens quality all shift the real result.
How focal length changes everything
Focal length (measured in mm) controls the field of view. A short focal length like 2.8mm gives a wide angle, good for covering a whole room but with low detail at distance. A longer focal length like 6mm or 12mm zooms in, giving high detail on a narrow strip far away.
Fixed vs varifocal lenses
Fixed-lens cameras are cheaper and fine when you know the exact distance. Varifocal (adjustable) lenses cost more but let you tune the field of view after mounting. For a long driveway or a specific gate at 10-15 meters, a varifocal or a longer fixed lens usually beats a wide 2.8mm camera at the same resolution.
A simple way to match camera to area
Work backward from the target, not the camera:
- Decide the goal for each spot: detect, recognize, or identify.
- Measure the distance from mount to target and the width you must cover.
- Pick a focal length that frames that width at that distance.
- Confirm the resolution gives enough PPM on the target within that frame.
Real example: a shop entrance and a parking area
A small shop wanted “one camera to see everything.” The single wide 2.8mm 4MP camera showed the whole front, but faces at the door were unreadable and plates in the lot were mush. The fix used two purpose-built cameras instead: a 4MP fixed 4mm unit aimed tightly at the door for face identification, and a 4MP varifocal set near 8mm aimed along the parking row for plates. Same total budget, but now both goals were met because each camera concentrated its pixels on one job.
Common mistakes and how to fix them
- Buying high MP to cover a huge area. Detail still drops with distance. Fix: add a second camera or use a longer lens rather than one wide camera doing everything.
- Ignoring low-light performance. A large sensor with good low-light handling often beats raw megapixels at night. Fix: compare sensor size and night specs, not just MP.
- Mounting too high. Cameras aimed steeply down capture the top of heads, not faces. Fix: mount at roughly 2.5-3m and angle so faces enter the frame level-on.
- Forgetting bandwidth and storage. Higher resolution and frame rate mean bigger files. Fix: budget storage before finalizing specs.
Action checklist before you buy
- List each area and its goal: detect, recognize, or identify.
- Measure distance and required coverage width for each.
- Choose focal length to frame that width; prefer varifocal when distance is uncertain.
- Verify PPM on the target, not just the camera’s headline resolution.
- Check night vision, sensor size, and WDR for backlit spots.
- Confirm your recorder and storage can handle the combined bitrate.
Conclusion and next step
Match pixels to the target, not to the whole scene. Start by writing down the goal for each camera position, then pick focal length first and confirm resolution second. Your next step: walk your site with a tape measure and note distances at each planned mount point before choosing any hardware.
FAQ
Is a 4K camera always better than a 2MP one?
No. 4K helps only if the extra pixels land on your target. For a close doorway, a good 2MP camera can outperform a distant 4K one, and it uses far less storage.
What focal length should I use for a gate 12 meters away?
For recognizing people or reading plates at that range, something around 6-12mm usually works better than a wide 2.8mm lens. A varifocal lens lets you fine-tune after mounting.
Why do my night images look worse than the daytime spec suggests?
Night detail depends on sensor size, aperture, and IR reach, not resolution. Motion blur also increases in low light. Prioritize cameras with strong low-light ratings for dark areas.
Can one camera both identify faces and read plates?
Rarely, unless both targets sit at the same distance and width. Usually two focused cameras beat one compromise camera for these separate goals.
References
General pixels-per-meter guidance in this article aligns with widely used CCTV design conventions, including the IEC 62676-4 standard for video surveillance system design and performance.