HIK Camera Lens Guide: 2.8mm vs 4mm to 12mm FoV

Two cameras with the same resolution can produce very different results. One shows a wide reception area but cannot identify a face at the door. The other shows a narrow corridor with enough detail to read a badge. The difference is usually the lens, not the sensor.

This guide explains how to choose a HIK IPC lens for the job: how field of view works, what 2.8 mm, 4 mm, 6 mm and 12 mm actually mean in practice, how to calculate pixel density, and how to match the lens to identification, overview, number plates and perimeter coverage. It is written for installers, integrators and buyers who need the camera to deliver the right image at the right distance.

The approach is practical. Lens choice is a geometry problem with a few real-world constraints: sensor size, available light, mounting height, and the distance to the subject. Once those are fixed, the lens follows.

This guide is practical rather than theoretical: it focuses on the decisions that change what the camera can actually see.

1. Why lens choice matters more than resolution

Resolution describes how many pixels the sensor has. The lens decides where those pixels land. A 4 MP camera with a wide lens may put only 60 pixels across a face at 15 m, while a 4 MP camera with a narrow lens may put 200 pixels across the same face. The second camera identifies; the first only observes.

This is why two cameras advertised as "4 MP" can feel like different products. The megapixel number is a property of the sensor; the lens is a property of the view.

This is also why two sites with the same camera model can report different results. The installer's lens and aim turn the same hardware into a wide overview or a close identification view. Standardising lens choice across a project makes results more predictable and support easier.

Lens choice also affects night performance and compression. A narrower lens concentrates the scene on the sensor, which can improve perceived detail, but it also reduces the area covered and may need more IR illumination at distance. A wider lens covers more area but spreads pixels and light across it.

The practical rule is to start with the task. If the task is identification, the lens must put enough pixels on the subject at the expected distance. If the task is overview, a wider lens is appropriate.

There is also a commercial angle. A camera schedule that mixes two or three lens types usually costs less than buying high-resolution cameras everywhere and hoping the wide view is enough. The right lens on a 4 MP camera often beats a wide lens on an 8 MP camera for a specific task.

Finally, lens choice affects how the system is operated. Operators find events faster when each camera has a clear purpose and a predictable view. A wall of wide-angle overviews looks impressive but slows investigation because nothing is close enough to identify.

2. Field of view, distance and the geometry of detail

Field of view (FOV) is the angle the camera sees. It is usually quoted as horizontal, vertical or diagonal; horizontal is the most useful for planning because it determines the width covered at a given distance. A wider FOV covers more width but reduces the number of pixels per metre.

The relationship is geometric. The width covered at distance D is approximately 2 × D × tan(FOV/2). If you know the width you need to cover at a given distance, you can work backwards to the required FOV, and from there to a lens.

Distance is measured from the camera to the subject, not from the camera to the far wall. A camera mounted at 3 m looking at a door 8 m away has a subject distance of about 8 m, not the full depth of the room. Use the distance that matters for the task.

Height also matters. A camera mounted high looking down at a steep angle sees the top of the head and compresses the face. A camera at 2.5–3 m looking slightly down usually produces a more usable identification image than one at 5 m looking steeply down, even with the same lens.

Finally, consider the scene shape. A corridor is long and narrow, so a narrow lens along the corridor may cover the useful area better than a wide lens that includes walls. A yard is wide and open, so a wider lens may be appropriate.

A simple site sketch helps. Mark the camera position, the subject positions, the distances and the width that must be covered. Then calculate the FOV for each subject. This takes a few minutes and prevents the common mistake of choosing a lens for the whole scene when only one part of it matters.

Remember that the subject may move. A person walking through a door passes through a range of distances.

Aspect ratio also affects the calculation. A 16:9 image uses the full horizontal pixel count; a 4:3 crop may use fewer horizontal pixels for the same sensor. Check the camera's output aspect ratio and use the matching horizontal pixel count.

3. The common lens options: 2.8, 4, 6 and 12 mm

Fixed-lens cameras commonly ship with 2.8 mm, 4 mm, 6 mm or 12 mm lenses. The millimetre number describes focal length: smaller numbers give a wider view, larger numbers give a narrower, more magnified view. The exact FOV depends on sensor size, but the pattern is consistent.

A 2.8 mm lens is a wide-angle choice. It covers a large area close to the camera and suits small rooms, reception areas, short corridors and overview of a wide scene. It is a poor choice for identifying a person or a plate at distance because the pixels are spread across a wide view.

A 4 mm lens is the common general-purpose choice. It balances coverage and detail and suits entrances, offices, retail floors and medium-distance coverage. Many integrators treat 4 mm as the default and adjust from there.

A 6 mm lens narrows the view further. It suits corridors, gates, loading areas and medium-distance identification where the subject is expected at a known distance. It is a practical choice when the camera cannot be moved closer.

A 12 mm lens is a telephoto choice. It covers a narrow area at distance and suits perimeter fences, parking aisles, gates and plate capture where the subject is far from the camera. It requires careful aiming and a stable mount; small vibrations are magnified.

These FOV values are approximate and assume a common 1/2.8-inch sensor. A larger sensor with the same focal length usually gives a slightly narrower FOV, so

Wide lenses are often overused because they make the scene look complete. In practice, a wide view is useful for context and direction, but it rarely identifies anyone. If the camera must do both, consider a dual-lens model or two cameras rather than forcing one wide lens to do everything.

Narrow lenses are equally easy to overuse. A 12 mm lens on a scene that changes distance can miss activity at the near end.

Choose 4 mm when the subject is close and the scene is moderately wide; choose 6 mm when the subject is at 8–15 m and the scene is narrower. If both near and far subjects matter, a varifocal or a second camera is usually better than compromising.

4. Pixel density: the number that predicts detail

Pixel density is the number of pixels per metre at the subject distance. It is the most useful planning number because it translates resolution and FOV into something you can compare with a task requirement.

To calculate it, divide the horizontal pixel count by the width covered at the subject distance. For a 4 MP camera with 2560 horizontal pixels covering 16.8 m at 10 m distance, the pixel density is about 152 pixels per metre. Cover 5 m with the same camera and the density rises to about 512 pixels per metre.

Industry rules of thumb give a useful starting point. For observation (seeing what is happening), around 60–80 pixels per metre is often enough. For recognition (recognising a known person), around 125 pixels per metre is a common target. For identification (proving who someone is), around 250 pixels per metre or more is a safer target. These are planning guidelines, not guarantees; light, compression and angle still matter.

Use the density target to choose the lens. If the target is 250 px/m at 12 m and the camera has 2560 horizontal pixels, the required coverage width is about 10.2 m. That corresponds to a horizontal FOV of roughly 46°, which points toward a lens between 6 mm and 12 mm depending on the sensor.

Worked calculation: a 4 MP camera with 2560 horizontal pixels, a 6 mm lens and a 10 m subject distance covers about 10.4 m. The pixel density is 2560 ÷ 10.4 ≈ 246 px/m. That is close to identification level. If the subject moves to 20 m, the width doubles to about 20.8 m and the density falls to about 123 px/m, which is recognition level at best.

This is why distance matters so much. The same camera and lens can identify at 10 m and only recognise at 20 m.

Keep the calculation simple and visible. A one-page table with camera, distance, lens, coverage width and pixel density is enough to review a project before ordering and to check it again at commissioning.

5. Matching lens to purpose

Identification at a door. The subject is usually 3–8 m from the camera. A 4 mm or 6 mm lens is common, mounted at 2.5–3 m and aimed slightly down. Check the pixel density at the door, not at the far end of the room.

Overview of a wide area. A 2.8 mm or 4 mm lens covers more scene. Use it to understand activity and direction, and pair it with a narrower camera if identification is also required. One camera rarely does both well.

Corridor. A 6 mm lens along the corridor often covers the useful depth better than a wide lens that includes walls. Aim along the corridor, not across it, and check the far end for pixel density.

Perimeter and parking. A 12 mm lens or a varifocal set to the narrow end gives the reach needed at distance. Stability matters: a long lens magnifies vibration, so use a sturdy mount and avoid poles that move in the wind.

Number plates. Plate capture is a specialist case. It needs a controlled angle, a fast shutter, enough IR or white light at night, and a lens that puts enough pixels across the plate.

For each purpose, write the target pixel density next to the camera. That single number makes the lens decision objective and gives the installer a clear acceptance criterion at commissioning.

Also consider the direction of travel. A camera looking along a path sees faces and plates at a consistent angle; a camera looking across a path sees profiles and oblique plates.

Retail adds another constraint: the camera often looks across a counter where the subject stands at a predictable distance. Measure that distance rather than the depth of the store. A lens chosen for the far wall may put too few pixels on the customer at the till.

6. Sensor size and lens coverage

The same focal length gives a different FOV on different sensor sizes. A 4 mm lens on a small sensor is wider than a 4 mm lens on a larger sensor. This is why comparing cameras by lens millimetres across models can be misleading.

Check the datasheet's horizontal FOV, not just the focal length. If the datasheet gives the FOV, use it directly in the width calculation. If it gives only the focal length and sensor size, use a lens calculator or the manufacturer's field-of-view table.

Larger sensors generally perform better in low light because each pixel collects more light. That can matter more than a small lens difference at night.

Sensor size also affects depth of field. A larger sensor with a fast lens can produce a shallower depth of field, which may blur the background or the near field. For surveillance, this is usually acceptable, but check that the subject stays in focus across the expected distance range.

If the datasheet gives a diagonal FOV only, convert carefully. Horizontal FOV is smaller than diagonal; using the diagonal value in a horizontal calculation overestimates coverage.

A useful shortcut is crop factor: compare the camera's sensor to a full-frame reference. A smaller sensor multiplies the effective focal length, so the same millimetre lens looks narrower. This is why two 4 mm cameras can differ.

7. Fixed versus varifocal lenses

Fixed lenses are simpler, cheaper and usually smaller. They are a good choice when the required FOV is known and the camera position is fixed. The trade-off is that you cannot adjust the view after installation without changing the camera or the lens.

Varifocal lenses let you adjust the focal length on site, usually over a range such as 2.8–12 mm. They are useful when the exact distance is uncertain, when the camera must cover different scenes, or when the site changes over time. The trade-off is cost, size and the need to set focus carefully after adjusting the zoom.

For a project with a clear camera schedule, fixed lenses are often the better value. For retrofit work or uncertain layouts, varifocal gives flexibility.

Varifocal lenses need a proper focus procedure. Zoom to the required view, then focus at the subject distance, not at infinity. Recheck focus after tightening the lens, because the act of tightening can shift it slightly. Confirm the night image before closing the housing.

Fixed lenses are simpler but less forgiving. If the view is wrong after installation, the options are to move the camera, change the lens or accept the result.

When adjusting a varifocal, use the camera's web interface or a monitor at the camera, not a phone screen at the recorder. Small focus errors are hard to see on a small screen. Tighten the lens, then recheck.

8. Light, aperture and night performance

The lens aperture affects how much light reaches the sensor. A larger aperture (smaller f-number) helps in low light but reduces depth of field. For surveillance, the practical effect is that a fast lens can produce a brighter image at night, but focus becomes more critical.

IR illumination interacts with the lens and the housing. A wide lens may spread IR over a large area and reduce range; a narrow lens concentrates IR but covers less scene. Check the camera's IR range and the actual distance to the subject, not just the datasheet's maximum.

Reflections are a common night problem. A dome close to a wall or ceiling can reflect IR back into the lens, producing a bright halo. A bullet with the IR ring too close to a surface can do the same.

Aperture also affects motion blur. A fast lens helps in low light, but a slow shutter still blurs moving subjects. If the task involves moving people or vehicles, balance aperture, shutter and gain rather than relying on the lens alone.

White light changes the calculation. A camera with white-light illumination may produce colour at night, but the light can disturb neighbours or attract attention.

Backlight is a lens-and-sensor problem. A camera pointed at a bright door may expose for the outside and leave the inside dark. Wide dynamic range helps, but the angle and lens still matter; avoid pointing the camera at the brightest part of the scene.

9. Distortion, IR and edge performance

Wide lenses often show barrel distortion at the edges: straight lines curve. This is normal and can be acceptable for overview, but it complicates measurement and plate reading. Some cameras offer distortion correction; check whether it affects the effective FOV.

Edge performance can be softer than the centre, especially on wide lenses. If the subject of interest is at the edge of the frame, aim the camera so the subject falls closer to the centre. Test the actual scene rather than assuming the whole frame is equally sharp.

IR focus can shift slightly compared with daylight focus. After setting focus in daylight, check the night image and adjust if necessary.

Check the corners of the image with a test target or a detailed scene. A lens that looks acceptable on a plain wall may show softness on text or faces at the edge. If the subject is likely to appear at the edge, re-aim the camera.

Compression also affects perceived sharpness. A high bitrate preserves detail; a low bitrate smears it. When comparing lenses, compare images at the bitrate the system will actually use, not at a laboratory maximum.

10. Worked examples

Retail entrance. A 4 MP camera at 3 m height, subject at 6 m, 4 mm lens. The horizontal FOV is about 80°, so the width at 6 m is about 10 m. Pixel density is about 256 px/m, which supports identification at the door. A 2.8 mm lens would cover more of the street but reduce density to roughly 160 px/m.

Warehouse aisle. A 4 MP camera at 4 m height looking down a 40 m aisle. A 6 mm lens covers about 10.4 m at 10 m and about 20.8 m at 20 m, which is still wider than the aisle. The far end may fall below the density target; consider a second camera or a varifocal set narrower.

Parking perimeter. A 4 MP camera covering a gate at 25 m. A 12 mm lens covers about 12.5 m at 25 m, giving roughly 205 px/m, which is close to recognition level. For plate capture, a dedicated plate camera with the correct angle and shutter is a better choice.

Small office. A 2.8 mm lens at 2.5 m covers a wide view of a small room. Pixel density is high because the distances are short, so a wide lens is appropriate.

Loading dock. A 6 mm camera at 5 m height covering a 15 m deep dock. At 10 m the width is about 10.4 m, which fits the dock width. At 15 m the density falls; if identification is needed at the far end, add a second camera rather than choosing a narrower lens that misses the near field.

Long perimeter. A 12 mm camera covering a fence line at 30 m. The width at 30 m is about 15 m, giving roughly 170 px/m. That supports recognition but not necessarily identification.

Warehouse aisle at 30 m. A 6 mm lens covers about 31 m at 30 m, which is wider than a typical aisle. The pixel density falls to about 82 px/m, enough for observation but not identification. Add a camera at the far end if identification is required.

11. Common mistakes

Choosing by megapixels alone is the most common mistake. It ignores where the pixels land. A 4 MP camera with the wrong lens can underperform a 2 MP camera with the right one for a specific task.

Using a wide lens for identification at distance is the second. It covers the area but spreads pixels too thinly. If identification is required, choose the lens for the subject distance, even if the overview is less complete.

Ignoring mounting height is the third. A high camera looking steeply down compresses the face and reduces usable detail. Mount at a height that supports the task, not at the highest convenient point.

Forgetting night performance is the fourth. A lens that works in daylight may be too slow at night, or the IR may not reach the subject. Test the night image before handover.

Not recording the final setting is the fifth.

Assuming the datasheet FOV is the horizontal FOV is another common error. Some datasheets quote diagonal FOV, which is larger. Using it in a horizontal calculation overestimates coverage and leads to a lens that is too wide.

Ignoring the scene at night is the sixth. A camera that covers the area in daylight may be unusable after dark if the IR range is short or the lens is slow.

Commissioning is the final check. After the camera is mounted and focused, measure or estimate the distance to the subject and confirm the pixel density against the target. If the density is short, adjust the lens or the position before handover. A five-minute check prevents months of unusable footage.

12. Selection checklist

• State the task for each camera: observation, recognition, identification or plate capture.
• Measure the subject distance and the width to cover.
• Set a pixel-density target for the task.
• Choose a lens (or varifocal range) that meets the target at that distance.
• Check the datasheet FOV for the actual sensor, not just the focal length.
• Confirm mounting height and angle support the task.
• Check night performance and IR range at the real distance.
• Test for distortion and edge softness if the subject is off-centre.
• Record the final lens, focus and aim in the handover document.
• Confirm the aspect ratio and horizontal pixel count used in the calculation.
• Check the near field and far field, not just the mid-point.
• Verify night focus and IR range at the subject distance.

13. FAQ

Is a 4 mm lens always the best default? It is a good general-purpose choice, but not universal. Identification at distance may need 6 mm or 12 mm; a small room may need 2.8 mm.

Does a higher megapixel camera always give more detail? Only if the lens puts those pixels on the subject. A wide lens on a high-resolution camera can still produce low pixel density at distance.

What is pixel density? Pixels per metre at the subject distance. It combines resolution and field of view into a single planning number.

Can I change the lens later? On some cameras, yes; on many fixed-lens models, no. A varifocal camera gives more flexibility if the layout may change.

Why does my wide camera look distorted at the edges? Wide lenses often show barrel distortion. It is normal, but it can matter for measurement and plates. Consider a narrower lens or distortion correction.

Why is the night image soft? IR focus can differ from daylight focus, and IR may reflect from nearby surfaces. Check focus at night and adjust the angle.

Can one camera do overview and identification? Rarely well. Use two cameras or accept a compromise. The lens that covers a wide area usually does not put enough pixels on a distant face.

How do I choose between 6 mm and 12 mm? Use the distance and the required width. A 6 mm lens covers about 10 m at 10 m; a 12 mm covers about 5 m at 10 m. Match that to the scene.

What is the difference between horizontal and diagonal FOV? Diagonal is the largest angle across the image; horizontal is the width you usually plan with. Use horizontal for coverage calculations.

Should I choose a varifocal for every camera? No. Fixed lenses are cheaper and simpler when the view is known. Varifocal is useful for uncertain layouts and retrofits.

How do I test a lens before buying many cameras? Install one camera at the real position and check the image at the real distances, day and night. A single test is worth more than a spreadsheet.

Does a wider lens reduce night range? Yes, because IR is spread over a larger area. A narrower lens concentrates IR but covers less scene.

Can I use a lens calculator? Yes, if you use the correct sensor size and aspect ratio. Verify the result on site.

Is a 2.8 mm lens ever right for identification? Only at very short distances, such as a small room or a door right beside the camera. At longer distances the pixel density is too low.

How does mounting height change the effective distance? A higher camera looks down at a steeper angle, which foreshortens the face and changes the distance to the subject. Measure the actual line of sight, not the floor distance.

14. Glossary

• Focal length: the lens property, in millimetres, that determines the view angle.
• Field of view (FOV): the angle the camera sees, usually quoted horizontally.
• Pixel density: pixels per metre at the subject distance.
• Varifocal: a lens with adjustable focal length.
• Aperture: the lens opening; a smaller f-number lets in more light.
• Depth of field: the range of distances that appear in focus.
• Barrel distortion: curvature of straight lines at the edges of a wide image.
• IR: infrared illumination used for night vision.
• Identification: detail level that supports proving who a person is.
• Recognition: detail level that supports recognising a known person.
• Horizontal FOV: the angle across the width of the image; the main planning value.
• Diagonal FOV: the angle across the image diagonal; larger than horizontal.
• f-number: the ratio describing aperture size; smaller numbers admit more light.
• Crop factor: a multiplier comparing a sensor to a full-frame reference.
• Line of sight: the straight-line distance and angle from camera to subject.

15. Sources

• Angle of view — geometric background for field-of-view calculations.
• Focal length — relationship between focal length, sensor size and view angle.
• ONVIF — interoperability profiles for IP video products.

Shop HIK IP cameras

In-stock and priced models for this topic. Confirm the exact suffix and quantity before ordering.

DS-2CD1043G2-LIU(F) 4 MP Bullet

$98.00

View product →

DS-2CD1083G2-LIU(F) 8 MP Bullet

$116.00

View product →

DS-2CD2023G0-I 2 MP Bullet

$119.00

View product →

DS-2CD2023G2-I(U) 2 MP Bullet

$127.00

View product →