What are the most common mistakes end-users do when purchasing a CCTV system? What should the end-users consider before purchasing a CCTV system?

Going the right way to make an effective, stable CCTV system in your project could be a tough task. Especially, if the budget is not skyrocketed. That's the main reason why customers sometimes make mistakes which lead to incapacity of the system. Yet, with the help of qualified installators, presale-engineers and technical support teams, and, we hope, with this article, you will able to increase the efficiency of your future CCTV system not necessary Increasing the cost of it. Even though it can take more time to calculate the project, we recommend to avoid the following mistakes when designing the system.
We have separated all the mistakes into 4 categories:

Not considering important hardware parameters and software restrictions to accomplish the technical design of the project. And overpayment for not-needed features.

We strongly recommend to consider and check all the parameters which could be important in a particular project. Whether that be camera sensitivity, or NVR maximum available bitrate, or maximum object distance for some software detector, it will take more time to figure out all that stuff, but it will save the budget in future when one will find out that some part of the hardware or software are not in their place. The same will go for paying for not-needed qualities and features. If to consider most common mistakes, we can highlight:
A) Choosing higher resolution cameras with smaller sensor or/and aperture.
Very popular on low-budget cameras, manufacturer of which likes to indicate in first place the resolution and leaving not less important characteristics in "smaller font". Those 3 characteristics work in unison, and each of them could decrease the quality of the image drastically. The 2 Mp camera with 1/2.8 sensor will have a better picture than a 4 Mp camera with 1/4 sensor.
B) Choosing too-high resolution camera. And too-high FPS.
The choice of resolution has to be based on how-detailed picture you want to get, how much pixel density you need on which distance. For example, if you only want to detect or/and count a human crossing a border on entering a zone, and you don't plan to check his signs from afar, you will not need high resolution of 5 or 8 Mp, most likely 2 Mp or lesser will do. Considering the quality of neural webs designed to detect various objects, some of them working already good on sub-stream resolutions. For Trassir Neuro detector, for example, minimum required resolution is 416*416 . Also we have to consider the backside of the higher resolution and FPS - the need for more space on HDDs, more network capacity, more load on the recording device. If it is planned to place the camera in not-that-populated zone, probably 30 or 25 FPS is not needed, and 8~15 will be enough.
C) Choosing wrong focal length
There always should be a balance between the camera's angle of view and pixel density at the distance. For example, 2.8 mm focal length will give larger angle of view, but 3.6 mm focal length will give more pixel density at the same distance.
D) Not paying attention for camera characteristics for various software detectors.
Reading the face attributes or text on fast-moving vehicle's license plate could be a tough task for CCTV system. But you can make it more easier for it, if you'll consider the recommendations of Recognition Module's manufacturer. That could be minimal sensitivity, minimal pixel density, recommended minimal shutter speed, etc.
E) Overpaying for not-needed characteristics.
That's the back side of the coin. If you don't plan to read the license plate of moving vehicle in your office, you might not need the 1/30000 shutter speed. Same for most-popular -resolution - what we talked about above and many other things. Same goes for the software. We think that the model for using license bundles are not that correct. If you don't need a particular detector - why to pay for it?
F) Choosing the wrong NVR.
As an example - all have heard of the ARM processors. Many NVRs built on those processors will likely to have different restrictions, like: maximum HDD capacity, limitation of resolution of connected cameras, restrictions in functionality.
G) Not paying attention to VMS software restrictions.
The VMS products are very different from each other. One could have the Failover function for 1200 channels maximum, the other one - only for 600. It might happen, that you will not be able to find such kind of information if you do not dive deep enough into the pool of documentation. This is why that part is probably is most difficult to find out without the help of the employees of the VMS manufacturer, but they will be ready to help.

Lack of relationship between CCTV and ACS components.

Many projects contain the hardware and software from different manufacturers, and bringing CCTV and ACS in one Software is becoming more and more popular. The task to link them all together could be a tough one. We recommend to contact technical specialists who will help you to figure out, based on your technical task - can one device work another, or with software from different vendor. If to consider most common troubles, we can highlight:
A) Device's lack of integration in the software.
Even if the camera is integrated "natively" in a particular software system - there could be the lack of some functionality. For example, the video settings setup will work, but the onboard detector's events won't go to the VMS. Same can go for the Access Control or Point Of Sale device - for example one or another type authentication or event could not be supported In the software. The specialists can help to examine at which level of integration particular device is.
B) Lack of the Onvif profiles in the VMS.
There are different Onvif profiles, and some of them could not be supported by the VMS. For example, you could see the video stream from a camera, and set up some settings, but PTZ or Motion detector will not be available. The specialists can help to examine which Onvif profiles are supported in the VMS
C) Codecs support in the VMS.
At the point of project design and codecs choice - you should be aware that the software or even part of it will not support some of them, like h.265 or h.265+, for example it can be supported in VMS but not in the Cloud of the same manufacturer.
D) Different Wiegand interfaces.
Occurs in joint use of VMS and ACS. Contact the technical team of the manufacturer to examine whether there will be support of a particular Wiegand interface.
E) API/SDK hidden restrictions.
If planned to integrate the third-party software functionality into your system, it is recommended to examine the API/SDK documentation more deeply. It could be that the manufacturer declares that the output of the module is fully accessible, but in reality, part of the functionality will not be provided.
F) PoE standards compatibility
If planned to use PoE in your projects, pay attention at what IEEE standard the camera is using and the capability of the PoE switch.

Making wrong calculations of network bandwidth and archive depth. Including the absence of the reserve.

We recommend to always use the recommended bitrates to calculate the archive depth and network bandwidth to correctly choose the hardware part - NVR/Server, HDDs, network devices - switches, routers. You should also consider the 10-20% reserve both in capacity and slots/ports.
Be free to contact technical specialists - they will not only help you to calculate but also will tell you which part of their HW/SW will affect (and in which way) the archive and network. Also they will be ready to advice how reduce it, if necessary.

Choice of low-quality accessories.

The importance of accessories and their influence on the reliability of the CCTV system is not that much less as the cameras, NVRs, network devices and so on. Most popular mistakes here:
A) Poor quality cabling.
For example one should meet the recommendations for the ethernet cables for PoE outdoor - CAT5e (or better), 24AWG, higher temperature rating.
B) Poor quality connectors.
Not less important as the cables themselves. The same requirements as for cabling should be met. We would not recommend to use standard plastic ones.
C) Unsuitable mounting boxes.
Common mounting boxes may not provide the same protection levels as the manufacturer's ones, which we would recommend to use.
D) Unsuitable lens for box form-factor cameras.
Focal length, aperture, resolution, and, most importantly, sensor size - all those parameters should met the characteristics of the box camera
E) Lightning protection.
All parts should be protected form lighting attacks. TVS diodes, grounding, network devices protection - all should be implemented in the CCTV design of the project.

Vasily Selivantsev