This article isn't about teaching granny to suck eggs, just notes from my own experience. Despite the prevalence of IP video there are still many legacy CCTV systems scattered throughout the UK and elsewhere. As camera, matrix and telemetry control kit fails or becomes unsupported CCTV managers must take the decision to replace their long-standing installations; we take a stroll through the many challenges of switching one's legacy installation to a modern IP-based system
These systems still work - why upgrade in the first place?
In a word, obsolescence. Many of the original analogue camera heads stopped being manufactured years ago and their replacements are slowly being phased out too - installations increasingly rely on in house spares to replace failed cameras and matrix devices. A more insidious problem is that of support; as time goes by manufacturers want their customers to replace kit with their latest-and-greatest and might force the issue by creating new, astronomically expensive support plans for legacy equipment.
1. Key differences and their implications
Correct configuration of cameras is critical.
IP cameras have an often bewildering array of configuration options, not least for the video encoder. Default encoder profiles often yield undesirable performance for a variety of reasons and must be adjusted to provide optimal streams. Furthermore standards such as ONVIF allow all sorts of configuration operations to take place automatically and understanding exactly what this means for the system in question is critical in installation and troubleshooting. Time and time again cameras are installed without properly setting up the encoder leading to all sorts of issues.
While understanding the full complexity of a compression algorithm or connection standard is unnecessary the installer must have a strong grasp of the key settings and be ale to balance their impact on network performance and video quality.
Data storage and bandwidth requirements are increased.
A D1 PAL stream can happily be encoded into h.264 at a bandwidth of 1Mbit/s while HD cameras will usually require a bandwidth of 4 - 8 Mbit/s and 4K cameras require significantly more than this, often 20MBit/s plus. Storage and transfer requirements can often be an order of magnitude higher than previously and network bandwidth requirements must increase by the same amount.
Storage requirements are drastically increased by switching to IP video. Fortunately hard disk storage costs are low and dropping, but a download stored on a DVD, for example, will contain far fewer hours of footage than what was previously achievable. Furthermore network performance must be suitable to accommodate the requirements of the new installation
Perceived performance by the operator will be different.
Metrics for video performance usually boil down to image quality, video latency and control latency. Aside from the obvious differences in resolution and sensor / lens performance IP video has other differences. Whereas a decent analogue camera's video quality will be compromised by faulty transmission lines and / or termination IP video quality is affected by the encoder's h.264 implementation and configuration, as well as implications of transmitting h.264 over UDP, an inherently lossy protocol. Analogue video latency is the speed of light in the transmission medium, whereas IP video latency is a sum of encoder, network and decoder latencies. Telemetry control is also subject to these software and network latencies resulting in an old RS-485-based PTZ camera often being noticeably more "responsive" than its IP counterpart.
While an analogue camera with poor quality transmission media can still provide a useful image an IP camera experiencing regular packet loss is almost completely useless. Operators may also experience dramatically increased latencies often making high speed operations such as following moving targets more difficult. In addition, due to the GOP structure of h.264, camera switching also has inherent delays, often of a second or more, which is entirely dependent on encoder configuration.
IP systems are a paradigm shift from their analog counterparts.
There's a world of difference between a discretely-cabled analog system and an IP network based system. Installing a new IP CCTV system is essentially two projects in one - a CCTV project, and an IT / network infrastructure project. Video, telemetry and control flows in legacy installations are usually represented by physical wiring whereas IP systems control and view everything over the network, leading to a more opaque system - control and video are hidden in a world of blinking lights. Troubleshooting a bad PAL video signal or dodgy RS485 bus involves tracing the signal physically, whereas troubleshooting inbound packet loss from a camera involves a dive into the management interfaces of network equipment.
The implication of this is a more complex installation, configuration and troubleshooting process requiring a completely different set of skills from the designer / installer / maintainer.
2. Challenges of system selection and migration
After the decision to investigate migration has been made there will be questions galore; these will probably include
What are the key differences between different solutions on the market?What functionality of the old system do we need in the new, and what new functionality is required?Will the new system be commissioned side-by side with the old?What will be used to encode legacy cameras, and how will legacy telemetry control be performed?Is integration to other equipment is necessary?How will the underlying IP network be implemented & maintained, and by whom?How will the CCTV system be monitored and maintained?When and how will the operators be trained?How will the new system impact our control room workflows?What will the project cost?
Approaching the plethora
Systems vary widely in the features they offer and it's critical to carefully analyse one's requirements. In order to choose a new system one must define which capabilities their legacy system has which the new one must also provide, which capabilities of the old system are ultimately unnecessary and which new capabilities are desired. This process will likely be lubricated by employing a consultant who has extensive knowledge of all the various solutions.
Humans, and Status Quo bias
Us humans naturally resist change. While a new system may ultimately provide similar advanced functionality, for example reporting and auditing or video mirroring it is likely that this will work in a different manner to the old system; one will therefore need to change hardware and software in hand with control room procedures. It is tempting to try to specify a system to function in a manner which limits the scope of these operational changes but it can be advantageous to use this as an opportunity to review standard operating procedures for it may well be that certain functionality of a different, previously unconsidered, system allows streamlining of control room workflows.
It follows that system selection and installation are often conflated; in order to install a system the installer must have intimate knowledge of the installation and configuration process and therefore the choice of installer often defines the choice of system, often forcing the hand of the user. If resources are available to manage the project internally then these can be separated but this requires significant time on the part of the project manager to separately define requirements, employ a CCTV consultant, employ a network consultant, define the design, then manage physical installation, system configuration, training and maintenance.
CCTV systems are typically expected to be operational at all times yet, at some point, there will inevitably be some downtime during any major upgrade project. Solutions exist on the market for hybrid analogue and IP systems; these allow all cameras to be moved across to a new recording and control platform without requiring the block replacement of all analogue cameras at the same time. Advantages of this approach are clear as the process and cost of replacing cameras can then be spread out while maintaining a consistent operational environment. Disadvantages could include cost per camera channel, as hybrid NVRs are often more costly per camera channel than their pure IP counterparts. Therefore it may be desirable to install an IP-only system as once migration is complete there will be no redundant functionality in the new system, but in order to provide interim recording of old analogue cameras a large number of PAL to h.264 encoders may be required to bridge the gap, which will then be ultimately redundant.
Operators will of course need training on the new platform. The shift-based environment will usually mean that multiple training sessions will have to be performed as the various shifts come and go, and inevitably questions and problems will crop up as the newly trained operators go about their daily tasks. While installation companies and manufacturers will offer training and support it may be prudent to select a particularly capable operator from each shift to receive a higher level of training such that they can field questions from their colleagues as they get a handle on their new operational environment. Depending on the type of operation this personnel management problem can be rather difficult to solve while maintaining day-to-day functions.
Let's try and add to this list of ins-and-outs - get in touch!