CCTV bandwidth and storage requirements

With the availability and adoption of 4K and even higher resolution correctly calculating camera storage and network bandwidth requirements can make significant differences to the cost and performance of your video surveillance system. Under specification may mean a reduction in frame rate or quality to meet your days of archive storage or become costly as more servers are required if quality cannot be compromised this goes hand in hand in cost effective large capacity hard drives becoming readily available and at affordable pricing.

Calculating CCTV network bandwidth and storage requirements looks to be a straightforward calculation on a few fixed variables such as camera resolution, frames per second, video compression algorithm used and for storage the number of days archiving required.

Frame Size = Resolution * Colour Depth (Chrominance/Luminance information)
Bit Rate = Frame Size * Frame Rate (fps)

Some of the most accurate calculators are supplied by the camera manufacturers and are based upon the actual performance of their devices in various situations. Using these can demonstrate a wide variance with other calculators and illustrate the difficulties in making anything other than a prediction. Using an established installer such as Check Your Security to design your surveillance system means you additionally get the experience of our engineers and technicians that have installed literally thousands of such cameras and have gained a reasonable idea of how devices perform under different conditions.

Storage is also affected by the number of hours per day you may expect images of interest to be generated. So, in retail establishments this is likely to be during opening hours and when staff are on site around these times in contrast to premises open 24 hours that have people or vehicles in motion at all times.

Where the calculations tend to let us down are where there are complex scenes with a lot of variation and movement such as outdoors where the movement of clouds, trees, grass, vehicles and people create ever changing scenes. On top of this we have lighting effects which to a camera are seen as changes in luminance (brightness) and chrominance (colour). Audio channels available on a number of cameras also add to the bandwidth but do not make a significant difference in bandwidth calculations.

Video compression codecs compress video by removing data, an unchanging scene in a warehouse for example will compress extremely well as many similar frames are discarded, whilst a complex outdoor scene may not as the codec must send more data to preserve the integrity and quality of the video.

The typical compression algorithms used in CCTV are lossy that is they lose data and as such the images can never be restored to their original resolution but like so many of our IT systems, we use today it is a trade-off. If you want superfast internet all of the time for example you will have to be prepared to pay considerably more than you do for the shared bandwidth you enjoy for your home broadband. Compressed CCTV images therefore require less storage space, are quicker to read or write and take up less network bandwidth. Full resolution uncompressed video and storage is still unrealistic in a typical security installation budget

Modern IP cameras and video encoders additionally could have motion detection or advanced video analytics, that identify people or vehicles, configured such that they only stream images when a particular threshold of movement is exceeded. This is often achieved by throttling the frame rate so a low frame rate is used until the motion trigger is activated then they start delivering frames at a faster rate. The cameras maintain a buffer, so the stored footage not only shows the event but also an increased frame rate up to several seconds prior to and after the event trigger.

Advanced video analytics using AI can make a significant difference to both storage requirements and network bandwidth. This can be particularly important with Cloud based systems. Cloud storage can be very flexible and infinitely scalable, but costs can run away if you are paying for bandwidth and storage by volume, do you really need to be storing that 30 FPS video of a high-resolution camera in a corridor 24 hours per day that only has any scenes of interest for less than 1 hour per day? Many cloud connectors, the onsite interface between your video management system and your cloud storage, can make these decisions for you.

Ultimately there can be trade off and with a security system with a large number of cameras in terms of system design and the difference in cost is significant. What needs to be ascertained are the system’s functional requirements, so if you need to detect faces or read vehicle registrations you can use high resolution video at great expense, position the camera closer to the subject so that more of the subject fills the image or use a high-quality telephoto lens to bring the subject closer. The same outcome may be achieved in different ways, but the storage and bandwidth implications may vary greatly.

Today bandwidth and storage are considerably cheaper than they were a few years ago but when your system with storage grade drives uses RAID to offer a level of redundancy and you have hundreds of cameras getting the calculations and options right still makes a difference.

Check Your Security can help you make the right choices and assist you in defining your functional and non-functional requirements for your CCTV Video management system and integrated security operations.

Kevin King Head of Technical Services Check Your Security Ltd