For integrators and end users, selecting appropriate video surveillance cameras and lenses for a given application throws up serious challenges.
Too often, cameras are installed that more-or-less do the job in reasonable conditions for the lowest price. But when conditions get tough, poorly selected cameras virtually guarantee that an expensive CCTV solution cannot meet its most basic requirements – identifying pedestrians and vehicles.
HOW do you choose the best cameras for your application? Which form factors, focal lengths and resolutions do you need? What are the camera qualities that are most valuable – are they WDR, useful resolution to 15 metres giving plates or faces, lack of motion blur in low light? Then there’s lens choice. Some good camera manufacturers slap on dreadful lenses that are loaded with purple fringing, internal reflections, strong distortions and lack of sharpness.
According to Luke Percy-Dove of Matryx, the secret to selecting the right CCTV camera for an application is understanding exactly what outcome the CCTV camera is expected to provide.
“I do this is by working backwards as if I was going to review CCTV footage post incident,” he says. “What would I want to see on the monitor? Lens selection is very important, critical even. Pick the wrong lens and you can jeopardise the success of camera placement. Focal length is important too, but the lens should complement the camera, not detract from it.When it comes to lenses, a lack of sharpness would be the biggest cause of frustration and why I will commonly reject cameras. Image detail is so important.
“In my opinion there are enormous differences in cameras which many in our industry still do not appreciate. Every application will have a camera to suit, but no one camera can suit every application. For instance, external CCTV is infinitely more difficult than anything you would do internally. For me, WDR is one of the first considerations. After that, everything else will be driven by the application because every application is different.”
“I find calculators to be very helpful with CCTV solutions and use them occasionally to validate my thinking before I document a project’s requirements. I also use them when I am unsure if a standard 4-8mm or 4-12mm lens is going to achieve what I need.”
When it comes to 4K and high-resolution sensors – does Percy-Dove think these are best used with wide angle lenses to deliver situational awareness on large sites? Or does he think there are some jobs that need maximum pixels?
“From my experience, 4K has deficiencies in low light and with motion,” explains Percy-Dove. “So, for stadiums and large open areas where one camera can replace multiple lower resolution cameras or cover very large areas, I think they make sense. For all other applications, I would always select a lower resolution camera with an appropriate lens. The outcome will be better. I am yet to specify a 4K camera anywhere.”
Doug Grant of CCTV Consultants says that camera choice comes down to application.
“For instance, dome or turret cameras are the choice of camera in areas where the public may try and damage the cameras,” Grant says. “These types are compact and aesthetically pleasing for most installations. Bullet cameras are excellent for longer ranges and they offer greater IR range than dome or turret cameras.
“Next come box (full body) cameras – in my opinion, these should only be used in situations which may include extreme distances, where a telephoto lens can be attached. PTZ cameras are a security operator’s choice as the camera can move at the operator’s discretion,” says Grant.
“Panoramic cameras are an exciting development in camera design. They range from 180-270 degrees with multiple image devices built into a single housing using a single Cat6 cable supported by PoE. In my opinion, the single lens 360-degree ceiling mounted panomorphic camera is best used in applications such as reception areas. And finally, there are speciality cameras which might be used for covert surveillance.”
When it comes to camera qualities, Grant has a firm operational focus.
“I value a camera that has useful WDR definitely, as well as firmware support with (open) edge recording algorithms for ANPR and facial recognition that support integration to brand-name VMS options,” he says. “Edge recording was the beginning, now I want to see all these features clients are asking for in a solution that not only record events but is interactive with local and cloud-based databases. Early days I know, but we will get there.”
Grant argues lenses are vital for quality imaging in video surveillance applications.
“Being old school, I cut my teeth on 1-inch image devices used in Newvicon tube type cameras, which processed the refracted light of the lens producing a well-defined image,” he says. “If installers selected a box type camera where they were able to use a C or CS mount type lens with glass optics, then the image quality of every application would be improved. Integrated cameras today give you the vari-focal option with cheap plastic optics and ever shrinking image sensors.”
Grant says he has always wanted more to offer a client in the way of a camera-lens combination.
“I foresee a camera and lens device that can view multiple images with clarity over a 180 deg angle of view and supports AI and has the optical capacity to clearly view objects up to 200 metres away, using segments of the image sensor to multi-view moving objects with a FoV ranging from a few meters,” he says. “With this device, all events that take place in the field of view are digitally magnified and brought into focus. The image sensor will process the information, record and transmit on multiple streams.
“I imagine this camera of the future to be the size of a broadcast camera, mostly lens. I’d really like to see a manufacturer develop this concept for the CCTV market. Imagine viewing multiple images from one camera on a video wall as it happens in real time using an artificial intelligence-based VMS. It’s a big step forward from the present-day panoramic technology which relies on multiple stitched cameras.”
According to Grant, when it comes to lenses things like distortion, chromatic aberration (seen as purple fringing), ghosting, lack of sharpness are aberrations of mass-produced consumer manufacturing.
“The CCTV industry will continue to suffer if they use cheaper technologies found in low cost consumer cameras, video cams and the like,” he says. “A price-driven industry has made these systems more affordable by pushing manufacturers to use sub-standard lenses in comparison to the glass optics we used to use.”
Grant argues that testing on site is the only way to separate cameras based on performance.
“Attending shoot-outs is one way – or you can undertake tests around the application,” he explains. “There’s difference in some camera solutions based on intended application. Typically, systems built for CBD surveillance will offer the best camera and lens combinations that integrate best into video management systems. Other cameras designed for general viewing or to function as a deterrent will cost less, have fewer features, less in the way of analytics, for instance.
“In my experience there are differences between domes cameras in the 4MP range – the brands I have worked with have their differences between models. Low light capabilities in higher resolution cameras are deficient compared to lower resolution cameras in the same brand. Increasing the lens format and sensor would greatly assist the lower light specification if it were possible.
“We all want high resolution images in every CCTV system but it’s not always appropriate. High resolution sensors work well in stadiums where multiple cameras were once required. These solutions work well with very high illumination and I also see an acceptance of multiple high-resolution cameras in panoramic or single lens 360-degree cameras.
“While the FoV and low light capabilities are much less in these applications, they certainly deliver acceptable general surveillance imagery using multiple image devices,” Grant says. “The panoramic type camera also stitches together 3 to 4 cameras with little distortion. And while single lens 360-degree cameras have specific installation parameters that must be followed, they too have a place in general surveillance applications.”
Grant says software calculators can be useful when planning a CCTV solution.
“I have found them to be a good starting point when designing a system that requires more than general surveillance from wide angle lenses,” he explains. “The only proven method when specific FoV is part of the design, is to set up the camera on site and have your client approve an image – the angle of view, the resolution, the camera settings.”
Meanwhile, Vlado Damjanovski says selecting the right CCTV camera is about education.
“I have studied, worked with and used cameras, sensors and lenses all my life, and there is no simple answer as to their selection,” he says. “There’s a variety of uses and expectations from a CCTV camera, and the only secret I can offer is to stay updated with technology so you can select the right CCTV camera based on your requirements.”
When it comes to desirable camera qualities, Damjanovski goes right to the heart of the camera system, pointing out that every CCTV solution is designed with certain expectations – almost always a demand for a good picture in very low light or in difficult lighting conditions.
“When it comes to the lens, if it’s no good it can be changed on some cameras. WDR can be turned on and off, or adjusted. But the sensor’s ability to gathering light is directly proportional to pixel size so the quality I always start with is the pixel size defined by sensor size and pixel count,” he explains. “Overall pixel count is secondary, although it’s important to have a certain standard format HD or UHD (4K) to match the system.
“The next most valuable quality would be the lens – is the lens matching or exceeding the sensor quality? I regret to say this, but the CCTV industry doesn’t pay much attention to camera optics,” Damjanovski says. “I have been writing about the importance of optics for over 20 years, but it still needs to be discussed. The industry needs to understand the optical principles, features and qualities.
“And the third quality would be the camera image processing electronics, which includes WDR and compression. Motion blur is controllable by tweaking electronic exposure so it can be minimised. Many users don’t know that they can control the electronic exposure to minimise the motion blur. This is typically done by shortening the electronic exposure from the typical 1/25s or 1/30s, to 1/100s, 1/250s or maybe even 1/500s while still producing live video.”
According to Damjanovski, calculators can be a big help when planning and commissioning CCTV applications, but he says they are not all are created equally.
“The ViDiLabs calculator available for both iOS and Android smart devices is not like anything else available on the market,” he says. “There’s nothing which calculates all the things we do – a reflected light lux-meter will soon be included. You can find it by searching under ViDiLabs calc or if you don’t want to buy it without seeing it, visit our web site (www.vidilabs.com/theapp) and see some of the examples I explain in my videos.”
Damjanovski argues that all CCTV cameras are not the same.
“CCTV cameras are definitely not created equal,” he says. “There is huge difference depending between them including the sensor used, the lens quality, the electronic processing quality, which includes compression. One thing I should also mention here, since we are on this topic, at our labs we are equipped to do various tests and we can evaluate cameras, lenses and systems. If anybody is interested in a scientific and objective test comparison of cameras, I would be happy to help.”
When it comes to higher resolution sensors, Damjanovski says the selected application depends on the pixel density at the target object distance.
“This can easily be calculated with a ViDiLabs calculator but let’s use an example,” he explains. “Sydney Trains wants to cover customers passing multiple ticket-gates at a train station across a field of view of 20 metres. We will say the customer wants to be able to identify faces with 350pix/m. If we use a 4K camera, let’s assume is has a ½-inch sensor – remember the issue of pixel size? 4K has twice the resolution so should have a larger sensor so that we don’t shrink the pixels and kill the light.
“We could use 3.6mm lens for example (assuming we have none, or minimal, optical barrel distortions). According to the ViDiLabs calculator, this would give us 350pix/m at 6m away from the boom-gates, covering a field of view about 11m wide. So, with two 4K cameras, and an optically good 3.6mm lens, we could cover all 20m with good facial identification. Comparatively, we could get almost the same quality coverage using four 1/3-inch HD cameras fitted with 5.2mm lenses installed 6m away from the gates.
“When we get 8K cameras (and they will come) – and assuming they have not shrunk pixel sizes too much – I would suggest a 1-inch sensor, with a 3.6mm non-distorting lens, which would give us coverage of a field of view of 22m from 6m distance, at the required 350pix/m.”
According to Scott Myles of ICS-Group, selecting the right CCTV camera is an interesting and open question aligned with the overall requirements of the client. Myles argues that the selection of cameras should be driven by the client’s overall security strategy.
“ICS-Group receives many requests for electronic security and CCTV from our clients, however in most cases these requests are not supported by a comprehensive strategy, as such, leaving it open to a one size fits all approach that is neither qualitative or meets the functional and operational needs of the client,” he says.
“These theoretical recommendations without the benefit of a defined strategy, results in the offering of a ‘canned solution’ which often leads to significant loss of value for the client – the intent of the security system may be lost, or in the worst case, completely compromised. Prior to any selection or recommendation, it is our practice to work with all stakeholders to clearly define the requirements so that any recommendations are fit for purpose, meeting the needs of the defined application.”
When it comes to qualities of the best cameras, Myles says camera applications are wide and varied and the qualities that are most valuable differ between applications and depend on the ultimate goals of the solution.
“We steer away from a one-size-fits-all approach and look at each scenario that drives the requirements for our recommendations,” Myles explains. “This approach ensures that the product chosen is effective and sustainable meeting the specific needs of the application. In our opinion it’s pointless deploying high megapixel cameras in an application where it they are not required, such as close capture views, where the intended subjects make up a large percentage of the captured area. There’s also no reason to deploy cameras with superior WDR performance in internal applications that will never see the light of day.
“However, in more challenging applications such as external or low light applications where properties such as depth of field and WDR is needed, it is our practice to appraise products that not only provide good day and night images but also proved low amounts of motion blur and have good dynamic range dealing with harsh light from environmental factors and vehicles. From our experience when evaluating cameras for more challenging environments is the tipping point that differentiates cameras from the vast array of products that are available in the marketplace.”
How important are lenses and what should end users and installers be looking for?
“As you would expect, lenses are a very important element of the camera and its ability to provide the desired outcome,” explains Myles. “It’s important to note that the past decade has seen the integration of CCTV cameras both fixed and PTZ, with the lens included as the delivered package – the good, old days of the integrator having to choose the best lens for the application is no longer a requirement.
“There is still a requirement for an understanding of focal lengths and targeted areas. During any evaluation process, lens quality such as distortion, depth of field and f-stop must be considered as part of the overall assessment of the product. Tier 1 manufacturers usually put effort into the quality of lenses, and this is reflected in the overall camera image output, as well as a camera’s ability to respond to rapid light fluctuations and the handling of motion blur,” Myles explains.
“Unfortunately for clients who do not do their due diligence and instead go for the one-size-fits-all low-cost option, we find such solutions ultimately run into trouble in low light or harsh conditions where the camera is exposed to all the issues consultants try to avoid during design. These include a lack of depth of field, image sharpness and distortion and the overall inability to provide a usable outcome when most needed.
“In an internal application with good lighting and no interference from environmental factors, most ‘in-range reputable security cameras’ provide good quality images, however, harsh environments are the big differentiator between products,” Myles explains.
“ICS-Group has often been called to remediate the situations which can arise from the ‘canned approach’ when investing in technology, which is why we are keen to engage new clients as early as possible in the security solution life cycle.”
When it comes to the quest for high resolution, Myles says less is sometimes more.
“The application for 4K cameras covering wide angle views for situational awareness are valid, however, it’s important to consider how these cameras perform in applications with lower operational light levels, where higher resolution and megapixel cameras often have limitations.
“Although we have found that 1080p has delivered the desired outcome without impacting overly on storage, we are not averse to recommending the use of 4K or megapixel cameras as the application requires, however, we don’t subscribe to the blanket use of 4K or high MP cameras in every application.” Ω