CCTV Camera Selection, Placement, Lighting and Image Quality under the SPF VSS Standard

Choosing a CCTV camera by megapixel count alone is one of the easiest ways to design a surveillance system that looks impressive on paper but fails operationally.

A camera may be:

4MP, 8MP or even higher resolution

and still fail to provide a usable image of the person or activity the building owner actually needs to see.

The Singapore Police Force Video Surveillance System Standard for Buildings approaches the problem differently.

It requires the VSS to be designed around its Operational Requirements, and treats the camera as only one part of a larger system that includes:

lighting, field of view, network transmission, displays and recording.

The Standard recommends cameras with a minimum HD 1080p resolution and true Wide Dynamic Range, but it also places considerable emphasis on:

camera position, field of view, lighting, environmental conditions, tamper protection and interoperability.

The key principle is:

Select the camera according to the image you need to obtain, not simply according to the number printed on the camera datasheet.

What is the minimum CCTV camera resolution under the SPF VSS Standard?

The Standard says cameras should have a minimum resolution of:

HD 1080p, 1920 × 1080 pixels or equivalent.

It also calls for true Wide Dynamic Range (WDR) so the camera can capture colour images under challenging lighting conditions.

This provides a baseline technical requirement.

However, it does not mean:

“Any 1080p camera installed anywhere will automatically achieve the required surveillance performance.”

The camera still needs to be positioned and configured so that the required subject occupies enough of the image.

Is 1080p CCTV enough for facial identification?

It can be, but not automatically.

The Standard separately defines different levels of surveillance detail:

Detect, Observe, Recognise and Identify.

Whether a 1080p camera achieves Identification depends on factors including:

• how large the person’s image is within the frame;
• distance from camera;
• field of view;
• camera angle;
• lighting;
• image compression; and
• overall recorded image quality.

The Standard specifically warns that image-detail guidance must also consider lighting, viewing angle and compression rather than resolution alone.

Consultant’s perspective

A useful way of putting this is:

Resolution tells us how many pixels the camera has. Field of view determines where those pixels are being used.

A 1080p camera tightly framed around a doorway may provide much more useful facial detail than an 8MP camera attempting to cover an entire large lobby.

Does a 4MP or 8MP camera automatically provide better CCTV?

Not necessarily.

Higher resolution can provide more image information, but only if the rest of the system can make practical use of it.

The relevant questions remain:

What area are we covering?

At what distance?

What level of detail is required at that location?

What does the recorded image actually look like?

Higher resolution can also influence:

network bandwidth, storage capacity and processing requirements.

The VSS Standard’s design philosophy is therefore better understood as performance-driven rather than megapixel-driven.

Why is field of view more important than the camera megapixel number alone?

A camera distributes its available pixels across the area it sees.

If its lens is adjusted to cover a very wide area, each individual person or object occupies a smaller part of the image.

If the view is narrowed, that same target occupies more pixels.

That is why the Standard requires defined camera views for locations where specific surveillance outcomes need to be achieved and says cameras used for its Coverage Requirements should be static, with the field of view protected from unauthorised adjustment.

Simple example

Imagine two cameras with the same resolution.

Camera A covers a 4-metre-wide entrance.

Camera B covers a 20-metre-wide lobby.

A person at the same distance may occupy significantly more pixels in Camera A.

So the important design question is not:

“How many megapixels?”

It is:

“How many useful pixels do we have on the target at the location where we need the required level of detail?”

We will go much deeper into that in the dedicated Detect, Observe, Recognise and Identify article.

What is Wide Dynamic Range?

Wide Dynamic Range, or WDR, helps a camera deal with scenes containing both very bright and dark areas.

This is particularly relevant at entrances.

For example, an indoor camera may be looking towards:

a person entering from bright daylight into a relatively dark lobby.

Without suitable dynamic-range performance, the camera may expose correctly for the bright background while the person’s face appears dark.

The VSS Standard therefore calls for true WDR capability as part of its general camera requirement.

Why is WDR particularly important at entrances?

Entrance cameras frequently face difficult lighting conditions.

Examples include:

• bright sunlight outside and darker interior lighting;
• glass entrance façades;
• reflections;
• changing sunlight throughout the day; and
• vehicle headlights at certain locations.

A camera can technically provide the correct field of view but still fail the surveillance requirement if the target is badly exposed.

Consultant’s perspective

This is why entrance testing should ideally include realistic lighting conditions, not just commissioning the camera in one convenient daytime condition.

Does the Standard require colour CCTV images?

The camera guidance calls for colour cameras and requires the camera to capture colour images under appropriate conditions.

For recorded video, the Standard later says images should be sharply defined with accurate colour reproduction under normal lighting, with sharply defined monochrome images continuing under reduced or emergency lighting.

The objective is therefore useful image information rather than colour at any cost in extremely low light.

Fixed camera or PTZ camera: which is better?

They perform different roles.

A fixed or static camera continuously looks at a predefined area.

A Pan-Tilt-Zoom (PTZ) camera allows an operator or system to move the camera and zoom into different areas.

The VSS Standard recognises PTZ cameras as useful for monitoring larger areas and following ongoing events.

However, it specifically warns that PTZ cameras may not be suitable for ingress and egress points, because the camera could intentionally or unintentionally be pointing away from the required view when someone enters or exits.

Can one PTZ camera replace several fixed cameras?

Sometimes a PTZ can provide operational coverage over a large area, but it cannot necessarily replace cameras required to maintain continuous defined views.

The Standard says cameras used to achieve its Coverage Requirements should be static.

Consider a PTZ covering four entrances.

It can look closely at:

Entrance A, B, C or D.

But unless specifically designed otherwise, it cannot maintain the required detailed view of all four simultaneously.

Consultant’s perspective

A useful distinction is:

Fixed camera: guaranteed view.

PTZ camera: operator-controlled or automated investigative view.

PTZ can supplement fixed coverage very effectively.

It should not automatically replace a fixed camera whose view must always be available.

What features should a PTZ camera have?

The Standard says PTZ cameras should allow predetermined:

positions, schedules, preset sequences and dwell times where needed.

It also calls for a default function that can automatically return the PTZ to its original position after a predetermined period.

The camera should also be able to pan and tilt quickly enough to capture fast-moving targets when suspicious activity is detected.

This addresses an important operational problem:

What happens after somebody manually moves the camera?

Without a return-to-home function, the camera could remain pointed away from its normal surveillance area indefinitely.

Are infrared cameras better for night surveillance?

Infrared-sensitive cameras with built-in IR illumination can improve images where visible light is poor.

The VSS Standard recognises this benefit.

However, it also states that increasing ambient lighting is preferred to relying on IR cameras, and notes that IR cameras can provide poorer colour rendition during daytime unless suitable filtering is used.

That is an important design point.

The solution to bad nighttime CCTV is not always:

“Buy a camera with stronger IR.”

Sometimes the better solution is:

improve the lighting.

Why might visible lighting be preferable to infrared?

Visible lighting benefits more than the camera.

It can also support:

security personnel, pedestrians, patrols and deterrence.

The Standard explains that appropriate security lighting can improve VSS effectiveness and security patrols while making criminal activity less attractive in properly illuminated areas.

Visible light can also preserve useful colour information that may help describe:

clothing, vehicles or other characteristics.

Does more lighting always improve CCTV?

No.

Badly designed lighting can actually make the CCTV image worse.

The Standard warns that poor lighting deployment can create glare, hinder vision and reduce image quality. It recommends considering glare reduction, light-source selection and protection against tampering or sabotage.

Consultant’s perspective

A bright floodlight pointing directly into the camera may create a worse image than a lower-power light positioned correctly.

Security lighting therefore needs to be designed with the cameras, not separately.

Where should security lighting be provided?

The Standard identifies areas such as:

building surroundings, perimeter lines, sensitive locations, pedestrian walkways, back lanes, publicly accessible routes, entrances and exits

as locations where adequate lighting can support VSS performance.

The lighting requirement should be established early so that good image quality can be achieved under the expected operating conditions.

Should security lighting be downward facing?

The Standard says that with appropriate fixtures, lighting should be directed downwards to support security operations while avoiding poorly deployed glare.

The exact lighting design still depends on:

the space, camera position, required view and surrounding environment.

Should security lighting be protected against tampering?

Yes.

The Standard recognises that security lighting itself may be deliberately damaged or disabled before an intrusion attempt.

It therefore discusses mounting lights high enough to make tampering difficult or protecting them with vandal-resistant measures such as mesh or tough polycarbonate housings.

For higher-risk installations, it also discusses standby power for security lighting.

Should the camera itself be visible?

For public areas, yes.

The Standard says cameras should be situated where they cannot easily be:

evaded, damaged or obscured

and should be clearly visible to the public.

Visibility can support the deterrence function of the VSS.

But visibility should not mean the camera becomes easy to reach and tamper with.

How high should CCTV cameras be mounted?

In its camera tamper-protection section, the Standard says cameras should be installed so an intruder cannot easily tamper with them or change their field of view.

It gives a minimum height of 2 metres from floor level as part of this protection approach.

However, this should not be interpreted as:

“Every camera should be installed as high as possible.”

Camera height also affects the viewing angle.

A very high camera looking steeply downwards may provide good top-of-head images but poor facial identification.

Consultant’s perspective

The correct position needs to balance:

tamper resistance + required facial/object view + physical constraints.

Is higher camera mounting always more secure?

Physically, a higher camera may be harder to reach.

Operationally, however, excessively high mounting can create poor viewing angles.

At an entrance where facial identification is required, the Standard says cameras should look towards the doorway or driver rather than down at them.

So the design should not sacrifice the actual surveillance objective simply to maximise mounting height.

Why is camera angle so important?

Imagine a person wearing a baseball cap.

A camera mounted directly overhead may capture:

the top of the cap.

A camera positioned to obtain a more frontal view may capture:

the person’s face.

The VSS Standard’s entrance requirements reflect this issue by specifically calling for frontal views at main entrances and exits.

This is why a floor plan showing camera cones is only part of CCTV design.

The vertical angle of view matters too.

Can one ceiling camera cover both general surveillance and facial identification?

Potentially, but this needs to be demonstrated.

A camera may provide excellent general Observation coverage but insufficient facial detail at the entrance.

In many designs, the two objectives may justify different views:

overview camera for understanding activity

and

targeted camera for identification at a specific chokepoint.

That is an application of the Standard’s different Field of View requirements, rather than a rule that two cameras must always be used.

Should CCTV camera views be protected from obstruction?

Yes.

The Standard says the required field of view should remain clear of:

temporary and permanent structures, vegetation and anti-climb elements associated with perimeter fencing.

This has implications beyond the original installation.

A camera view can deteriorate later because of:

new signage, furniture, partitions, landscaping, construction or stored materials.

Can a CCTV camera technically remain “healthy” while its coverage has failed?

Yes.

A camera could remain:

online, streaming and recording

while a newly installed sign completely blocks the target area.

Consultant’s perspective

This is why health monitoring and image-quality auditing are different things.

System health: Is the camera functioning?

Operational performance: Can the camera still see what it was designed to see?

Both matter.

The Standard requires agreed camera views and image quality to be documented at commissioning and periodically reviewed afterwards.

What is camera tamper protection?

Tamper protection aims to prevent somebody from deliberately:

moving, obscuring, damaging or disconnecting the camera.

The Standard says cameras should be mounted where tampering is difficult and calls for:

secured openings, inaccessible interconnections, protected cabling and antennae, and vandal-resistant/tamper-resistant housings with suitable viewing ports.

This protects both:

the camera itself

and

its intended field of view.

Should CCTV cables be accessible to the public?

No.

The Standard says camera interconnections, including cabling and antennae, should be secured and not accessible to the public.

There is little value in installing a well-protected camera if its communication cable can simply be disconnected nearby.

What IP rating should an outdoor camera have?

For outdoor cameras exposed to adverse weather, the Standard specifies a minimum IP65 enclosure requirement.

It explains that IP65 means the enclosure is dust-tight and water resistant.

This should be understood as the minimum guidance in this Standard for cameras exposed to those environmental conditions.

Different locations may justify higher environmental protection.

Does IP65 mean the camera is vandal resistant?

No.

IP ratings relate to ingress protection, such as dust and water.

Vandal resistance is a different issue.

The Standard separately discusses vandal-resistant and tamper-resistant camera enclosures.

So:

Weather resistance and impact/tamper resistance are not the same rating.

What is ONVIF and why does the VSS Standard mention it?

The Standard says network IP-based VSS should comply with prevailing Open Network Video Interface Forum (ONVIF) standards or an equivalent approach to support interoperability between compatible IP physical-security products.

In practical terms, interoperability helps reduce the risk of a surveillance system being completely dependent on proprietary interfaces between:

cameras, recorders and management platforms.

Does ONVIF guarantee that every feature works between every camera and VMS?

The VSS source does not make that claim.

It recommends ONVIF or equivalent to promote interoperability.

Consultant’s perspective

Actual compatibility should still be checked for the functions required by the project.

Basic live video interoperability does not necessarily prove that every advanced feature, analytic or configuration option will operate identically across different manufacturers.

That is implementation advice rather than a statement from the VSS Standard.

Should cameras be suitable for their environment?

Yes.

The Standard says cameras should be suitable for their intended indoor or outdoor location and continue to provide the specified picture and view across the expected:

weather, environmental conditions and temperatures.

This is another reason not to select cameras solely by resolution.

The environment may require consideration of:

weather exposure, low light, glare, temperature and physical tampering.

What should we look at when selecting a CCTV camera?

A practical selection process should begin with the required image, not the product.

This is a consultant’s design framework based on the VSS Standard rather than a table reproduced from the document.

A practical example: choosing a camera for a main entrance

Assume a building wants to identify people entering through its main door.

The Operational Requirement is not:

“Install a 4MP dome camera.”

It is:

“Obtain a frontal image of people entering the main entrance at Identification level.”

The design process then considers:

Position:
Can the camera obtain a frontal rather than steep overhead view?

Field of View:
How wide can the scene be while still providing sufficient image detail?

Lighting:
Will daylight behind the entrant create strong backlighting?

WDR:
Can the camera handle the bright external and darker internal areas?

Resolution:
Is sufficient image information available at the target point?

Tamper protection:
Can the camera be reached or redirected?

Recording:
Does the recorded footage preserve the required detail?

Only after answering those questions does the actual camera model become meaningful.

A practical example: choosing a camera for a car park

A general car-park camera may primarily require Observation coverage of activity.

A separate vehicle entrance camera may need to capture:

number plates.

Those are different requirements.

The general camera may have a relatively broad field of view.

The number-plate camera may require a much more controlled:

angle, viewing width, exposure and target position.

Consultant’s perspective

Attempting to make one very wide camera perform every surveillance task is often where image quality starts to fail.

Why should live and recorded image quality both be checked?

Because the footage ultimately used for investigation normally comes from the recording system.

Compression and recording settings can reduce detail.

The Standard therefore requires agreed live and recorded image quality to be documented at commissioning.

It also specifically warns that compression can reduce recorded image detail.

A camera’s live view may look excellent during configuration while the stored footage provides a lower-quality result.

What are some common CCTV camera-design mistakes?

Choosing cameras by megapixels alone.
Resolution does not prove that the required target detail will be achieved.

Installing cameras too high.
Tamper resistance improves, but the resulting angle may undermine facial identification.

Using PTZ as the only camera for a critical fixed view.
The camera may be pointing somewhere else when the event occurs.

Ignoring lighting.
A high-specification camera can still produce poor evidence if the lighting creates glare, silhouette or insufficient illumination.

Using IR to compensate for every low-light problem.
The Standard actually prefers increasing ambient illumination where possible.

Allowing vegetation or new construction to block the view.
The Standard specifically expects required fields of view to remain unobstructed.

Ignoring the camera’s environment.
Indoor and outdoor cameras need to remain suitable for the conditions in which they operate.

Protecting the camera but leaving its cable exposed.
The Standard expects camera interconnections to be secured from public access.

Checking only the live image.
The recorded image should also be validated.

Practical CCTV Camera Design Checklist

Before finalising each camera, the designer should be able to answer:

1. What security or operational problem is this camera addressing?
2. What exactly must be visible?
3. At what point in the scene must that detail be achieved?
4. Is the required view Detect, Observe, Recognise or Identify?
5. Is the camera resolution sufficient for that field of view?
6. Is the lens or viewing width appropriate?
7. Is the horizontal camera position appropriate?
8. Is the vertical viewing angle appropriate?
9. Does the camera obtain a frontal view where identification is required?
10. What happens during bright daylight?
11. What happens after dark?
12. Is true WDR needed?
13. Is additional visible lighting preferable to IR?
14. Could lighting create glare?
15. Is the required view permanent?
16. If so, should the camera be fixed rather than PTZ?
17. Can the camera be easily reached or redirected?
18. Are cables and connections protected?
19. Is the camera suitable for the environmental conditions?
20. Does an exposed outdoor camera meet the relevant ingress-protection requirement?
21. Can temporary structures, vegetation or future works obstruct the view?
22. Does the camera integrate appropriately with the VMS/NVR?
23. Does the recorded image achieve the same intended surveillance outcome?
24. Has the agreed view been documented during commissioning?
25. How will the view be checked again if the environment changes?

Common misunderstandings

“The camera is 8MP, so facial identification is guaranteed.”
No. Target size, field of view, angle, lighting and compression still determine usable detail.

“1080p is too low because modern cameras are 4MP or 8MP.”
The Standard sets 1080p as a minimum baseline, but the correct resolution depends on the required operational result.

“PTZ is better than fixed because it can see everywhere.”
It can look at different locations, but it cannot necessarily guarantee that a required fixed view is being captured when an incident occurs.

“The higher the camera, the better.”
Not necessarily. Height can improve tamper resistance while worsening the target viewing angle.

“IR means I don’t need lighting.”
The Standard expressly prefers increasing ambient lighting where possible.

“IP65 means vandal proof.”
No. IP65 addresses dust and water ingress; tamper and vandal resistance are addressed separately.

“ONVIF means every manufacturer’s camera will support every function on every VMS.”
The Standard recommends ONVIF or equivalent for interoperability, but it does not make that broader compatibility guarantee.

“If the installer can see the camera live, the camera works.”
That proves connectivity, not that the Operational Requirement has been achieved.

Conclusion

The SPF VSS Standard makes camera selection a performance question rather than a product question.

It establishes useful technical baselines such as:

HD 1080p minimum resolution, true WDR, suitable environmental protection, protected installation and IP interoperability.

But those specifications alone do not produce an effective CCTV system.

The camera must also have:

the correct position, the correct field of view, appropriate lighting, sufficient target detail and a stable view that remains available when needed.

PTZ, IR, higher megapixel counts and other camera capabilities can all be useful, but they only have value when they solve the particular surveillance problem.

The better design question is therefore not:

“Which camera has the best specification?”

It is:

“At the exact location where the incident matters, will this camera produce the image we need, under the conditions in which the building actually operates?”