A 2.8mm lens usually gives a wider view than a 3.6mm lens on the same camera platform. Choose between them by the width you need to cover at the place where the task happens—not by the millimetre label alone. A wide image can show both sides of an entrance while leaving too few pixels on the person crossing it. A narrower image can give that person more pixels while losing an approach route.
This guide turns that trade-off into a selection worksheet for installers and buyers. It uses a real Hector Weyl fixed-focal model to show what its published specifications establish, then uses clearly hypothetical widths for the arithmetic. The distinction matters: the reference specification lists two focal lengths but does not list their horizontal viewing angles. We will not invent those angles or claim a tested identification distance.
1. Define the task before choosing the lens
Start with one sentence describing what the recording must let someone determine. “Cover the entrance” leaves too much open. “Show which route a person takes through the entrance” and “show sufficient facial detail at the access checkpoint” are different requirements. Put the location of the required detail on the plan. That is the task line or target area for this article.
| Question at the site | What to mark on the plan | Lens-selection consequence |
|---|---|---|
| Where did someone enter? | All relevant approach routes | Check coverage before tightening the view |
| What happened at the checkpoint? | The actual crossing or stopping position | Allocate detail at that position |
| Can the operator inspect a small object? | Object size, orientation and working position | Check pixels on that object in recorded video |
| Will an analytic work here? | The analytic's required target region | Obtain its own installation requirements |
Record who will approve the result and how. An operator judging an exported clip may need a different view from a person monitoring a wall display. If two tasks compete for the same pixels, consider separating overview and detail coverage before ordering a single compromise lens.
2. What 2.8mm and 3.6mm actually tell you
Focal length helps describe framing; it does not specify the width of every camera's image at every installation. The sensor and lens combination matters. For a comparable camera configuration, increasing focal length generally narrows the field of view. A fixed-focal option provides one focal length; a varifocal product offers adjustment over a range. Two fixed-focal options printed on a datasheet do not mean the installed camera can switch between them. Axis's lens white paper, under “Lens types” and “Field of view,” explains these distinctions.
Ask for horizontal field of view, often written HFOV, when planning horizontal coverage. A diagonal angle cannot be inserted into a horizontal-width calculation. Confirm the recording mode and any crop as well. A specification copied from a different sensor, resolution mode or similarly named model may describe a different frame.
For an exact order, write the full model and selected focal length on the equipment schedule. “2MP dome” or “wide-angle version” is insufficient when the lens choice changes the view that will be accepted.
3. A real Hector Weyl example—and one missing specification
The HW-IPC-E1241S-IL-G2 is described as a 2MP Advanced Dual Light Fixed-Focal Dome Network Camera. Its reference datasheet, revision 001.001, PDF page 2, and its current product description both list 1920 × 1080 maximum resolution and 2.8mm/3.6mm focal-length options. These are documented model facts, checked for this article on 9 November 2026.
| Reference item | What the sources establish | What still needs confirmation |
|---|---|---|
| Lens type | Fixed-focal | Exact lens option on the offered unit |
| Focal lengths | 2.8mm / 3.6mm listed | Availability and selected order configuration |
| Maximum image size | 1920 × 1080 | The stream actually recorded |
| Horizontal field of view | Not specified in the checked sources | Model-specific HFOV or an agreed measurement |
The sensor's “1/2.9-inch” format label is not a measured active horizontal width in millimetres. We do not convert it into an exact viewing angle here. Equally, the listed illumination distance does not establish the distance at which a face or small object will meet your requirement.
This gap is a useful purchasing checkpoint. Request the missing view information for the exact configuration, or agree a sample framing test. The listing does not, by itself, confirm lens stock or a dispatch date. Hector Weyl's procurement condition is MOQ: 100 units; confirm the proposed equipment schedule before release.
4. Calculate pixels at the task line
A simple first check is horizontal pixel density:
Horizontal pixels per metre = recorded image width in pixels ÷ scene width at the task line in metres.
Consider two hypothetical framings from a 1920-pixel-wide stream. View A covers 12 metres across the task line; View B covers 8 metres. These widths are assumed for illustration. They are not measured HFOV results for the Hector Weyl model, and we do not assign either width to its 2.8mm or 3.6mm option.
| Assumed framing | Recorded width | Scene width at task line | Average horizontal density | Pixels across a 0.5m target |
|---|---|---|---|---|
| View A | 1920 pixels | 12m | 160px/m | 80 pixels |
| View B | 1920 pixels | 8m | 240px/m | 120 pixels |
The narrower assumed framing puts 50% more pixels across the same target: 120 ÷ 80 = 1.5. It also covers one-third less width: 8m instead of 12m. Neither view adds sensor pixels. It changes how the available pixels are distributed across the scene.
That is a budgeting calculation, not a pass certificate. The figures are approximate averages; distortion can change local scale. A 0.5m target is an arbitrary example, not a standard face width. Do not treat 160px/m or 240px/m as an automatic identification or analytics threshold.
5. Measure the right width in the right place
“The camera covers eight metres” needs a location attached to it. Width near the camera differs from width farther away. For a basic rectilinear planning model, a plane perpendicular to the camera's optical axis has this relationship:
Scene width = 2 × distance to that plane × tan(horizontal viewing angle ÷ 2).
Use this only when the horizontal angle is known for the actual configuration. Distance in this formula runs along the optical axis to the plane. A floor-plan distance or slant distance to a person is not automatically interchangeable with it. Camera tilt, an oblique doorway and lens distortion can make a simple top-view sketch inadequate.
At installation, mark a known horizontal span at the relevant task position and inspect the captured frame. Keep the camera position, height, aim and recording mode with that evidence. Check the near and far edges of an area if people can move through several depths. A satisfactory sample at the centre does not establish equal detail everywhere.
If the useful area falls outside the available framing, change position, lens or camera arrangement. Enlarging the drawing's coverage cone without changing the actual equipment does not solve the problem.
6. Check usable recorded detail, including at night
A live preview is a starting point. For acceptance, inspect the stream the recorder saves and a clip exported through the intended workflow. Record its pixel dimensions. A high-resolution main stream cannot be assumed just because a mobile preview looks clear; confirm which stream and settings each workflow uses.
Digital enlargement makes existing pixels larger. It cannot restore details the selected framing never captured. Likewise, a calculated pixel count does not settle motion blur, focus, lighting, compression or an obstructed target. The installation test should include the movement and illumination relevant to the task.
Repeat the agreed scene under required night conditions, with the intended lighting mode. Compare the same target position and viewing direction. A readable daytime label does not establish a readable moving target at night. Keep the original clip and the relevant settings with the result; a selected still frame alone may conceal inconsistent performance through a crossing.
For software analytics, obtain the specific application's requirements separately. The AI camera versus AI NVR guide covers where analysis runs; neither processing location nor a lens label removes the need to validate the input image.
7. Turn the comparison into an order-release decision
Use one equipment-schedule row per distinct view requirement. A project can contain wide approach views and tighter checkpoint views without pretending that every camera has the same job. Ask the supplier to acknowledge the selected lens against each row, along with configuration details that affect framing.
| Release check | Normal | Exception and next action |
|---|---|---|
| Exact configuration | Model and focal length agreed | Unspecified lens: obtain written selection |
| Task-line coverage | Required routes and area remain visible | Clipped route: revise framing or add coverage |
| Required detail | Agreed task demonstrated in recorded clips | Unclear target: investigate framing and image conditions |
| Evidence retained | Day/night clips, settings and plan are traceable | Preview only: complete the recording/export check |
For repeat deployments, preserve the approved mounting conditions as carefully as the model number. The same camera placed higher, aimed differently or set to another stream is not automatically the same tested view. Record substitutions and decide which acceptance checks must be repeated.
If HFOV or an acceptable sample is still missing, mark the lens choice as pending. Do not turn “usually wider” into approval for a quantity order. Separate that technical decision from commercial confirmation of quantity, availability and delivery. Our PoE power-budget guide supplies another independent check before the complete system is commissioned.
8. A commissioning sequence you can hand to the installer
- Freeze the question. Write the required task, target position and acceptance method on the plan.
- Confirm the installed configuration. Record the model, focal length, mounting position, aim and actual recorded resolution.
- Check the boundaries. Walk or place an agreed target through all required routes, including near and far limits.
- Check the detail. Review normal movement and required lighting conditions in recorded and exported clips.
- Close the evidence. Save settings, representative clips, exceptions and the approver's decision together.
Normal: the agreed area is visible and the required task is demonstrated in the retained recording. Abnormal: a route is clipped, detail is insufficient, or the result changes under a required condition. Next: isolate the failing condition, revise framing or settings as appropriate, and repeat the affected check. If detail improves but coverage is lost, return to the task allocation rather than calling the revision complete.
9. Frequently asked questions
Is 2.8mm always better for an entrance?
No. It may preserve more approach coverage on a comparable platform, but the required detail at the crossing can favour a tighter view. Define both coverage and detail before selecting the lens.
Does 3.6mm guarantee identification farther away?
No. A longer focal length can allocate more pixels to a target, but the result still depends on the exact view and recorded image. Confirm the task with relevant clips.
Can I change between these two options in software?
The reference model is listed as fixed-focal. Its two listed focal lengths are configuration options, not evidence of a software zoom range. Confirm the option on the offered unit.
Can digital zoom compensate for an overly wide lens?
It enlarges the captured image. If the required detail was never resolved in the recording, enlargement cannot supply it. Revisit framing, position or an additional detail view.
Are the example widths measured Hector Weyl results?
No. The 12m and 8m widths are explicit assumptions to demonstrate arithmetic. Neither is assigned to the reference model's 2.8mm or 3.6mm lens.
Should I use a DORI chart to approve the order?
Record the chart's standard edition and intended task. Axis's September 2026 explanation of IEC 62676-4:2025 describes changed categories from the 2014 edition. We do not apply an unlabelled threshold table as proof of site performance.
What should I send Hector Weyl for lens selection?
Send the plan, mounting position, task-line distance and width, target task, required day/night conditions and recording requirements. Include the proposed model, focal length and quantity. MOQ: 100 units.
10. Sources and verification boundaries
- Hector Weyl HW-IPC-E1241S-IL-G2 reference datasheet, Rev 001.001: PDF page 2, Camera and Lens sections; checked against the current product description on 9 November 2026. Establishes the cited resolution and focal-length options; HFOV remains unspecified.
- Axis, Lenses in surveillance, January 2025: “Lens types,” “Field of view” and “Lens calculator.” General optical context, not evidence of Hector Weyl lens angles or performance.
- Axis, Pixel density based on IEC 62676-4:2025, September 2026: “Changes in IEC 62676-4” and “Summary.” A manufacturer explanation of the edition change and model limitations; the IEC standard itself was not independently inspected.
The width comparison and release workflow are editorial planning tools. No field measurements, customer outcomes, identification guarantees or lens-stock claims are presented. For a project discussion, contact Hector Weyl with the equipment schedule and the task each view must satisfy.
Share:
PoE Camera Power Budget: Ports, Watts and Night Loads
CCTV Storage for 30 Days: Size by Recorded Bitrate