A distribution company recently held a series of workshops including one for a new manufacturer’s CCTV product range a few weeks ago. The training room was full, good to see, as a range of questions were raised focusing on networks and cabling, bandwidth (bitrates), storage and remote access.
The questions were fairly basic indicating a lack of real understanding for the IP products presented. At the end of the workshop, a special offer was made for an 8-channel NVR and two cameras at a highly discounted price, for which most attendees purchased one or two packages with one individual purchasing 20 or more. This begged a question: how many of those purchasing a discounted package understood what they were purchasing, how it performed, or did that not matter until the user has a problem?
Let us take a current popular 8-channel NVR and see what it is all about? The recommended retail price for the unit is just over R3 600 including VAT, without hard drive, but with 1 x SATA connection and recommendation for a 4 TB HDD or less. The white box has the capacity for up to eight channels of IP video with eight PoE (Power over Ethernet) ports, all fitted into a mini 1U case with one RJ-45 10/100 Mbps Ethernet interface.
The specification states a maximum incoming bandwidth of 50 Mbps and an outgoing bandwidth of 40 Mbps with live view/playback options of 1080p/720p/VGA, 4CIF down to QCIF. Please note this next comment – capability two channels of 1080p or 4 channels of 720p; that means you can only replay two channels of 2 MP video recording or four of 1 MP recording at one time.
There is a small print at the bottom of the datasheet that is easily overlooked: The formula to calculate the incoming bandwidth and the IP camera connected is: A=B/(C+D), where A refers to the number of IP cameras you connect. B refers to the value of the incoming bandwidth. C refers to the bitrate value of the main stream of the connected IP camera. D refers to the bitrate value of the sub-stream of the connected IP camera. The suggested calculation has B at a maximum of 50 Mbps, C = 6 Mbps, D = 1 Mbps, giving A = 50/(6+1) = 7. You can only connect 7 x 1080p cameras to the box at 49 Mbps and only replay two of the cameras at any one time at 2 MP.
Bitrate – Describes the rate at which bits of data are transferred from one location to another, commonly measured in bits, kilobits (Kbps) or megabits (Mbps) per second. Current IP cameras often have the choice of two streams of bits, a main stream and a sub-stream.
In the example above, and normally within the web interface, these default to a stated maximum bitrate for main stream of 6 Mbps and sub-stream of 1 Mbps, thus giving a 7 Mbps limit per camera (8 x7 = 56 Mbps). You can adjust the bitrate settings as you wish to obtain a higher or lower bitrate maximum setting or change to a variable bitrate option depending upon your requirements. It must be remembered though that your recommended data input is only 50 Mbps for this NVR; reducing or increasing the set bitrate may lose data or cause packet loss (video loss) should it not be managed correctly. Vision Catcher uses online tools to enable the calculation of the bitrate for each camera keeping the total bitrate within specification.
Pixel count – If your field of view and/or camera requirement wants to use a reduced quality of image to achieve recognition only rather than identification, then a lower camera pixel count can be chosen. The 1080p has a pixel count of 1920(H)x1080(V) or 2,07 MP (2.1 Megapixels). A 720p or 1 MP camera has a pixel count of 1280(H) x 720(V) = 921 600 pixels. The difference between the two halves the pixel count and you would see a halved bitrate assuming that frame rate and camera view are the same. So you can opt for cameras of a lower pixel count (lower resolution) such as D1 or 4CIF and combine with 2 MP and 1 MP cameras to ensure that your data rate remains under the NVR limitations. Please note that reducing the pixel count per metre within the field of view may reduce the quality of the image recorded.
Compression standards – M-JPEG compression works in a different way to H.264 (an MPEG 4 variant) and we expect to see M-JPEG use around three times the average bitrate against H.264. If we used M-JPEG to stream to the NVR we would definitely exceed the permitted bitrate and lose video footage (packet loss).
A compression type such as H.264 has a number of variables that can affect its bitrate including key frame settings, differing levels of resolution, wide dynamic range, movement, lighting and noise. They all add to the required understanding of the camera, its field of view and basic requirement down to the actual bitrate it produces over a 24-hour period; this can vary from less than 1 Mbps to well over 50 Mbps, hence the option for a maximum bitrate setting to avoid the possibility of losing video footage.
Live viewing, replay and remote communication – Our white box has a recommended output bitrate of 40 Mbps, an input bitrate of 50 Mbps on a 10/100 Mbps Ethernet interface. This allows for a 10% bitrate deviation owing to data flow, cabling and connectors. If I have a live stream for viewing on a monitor with all cameras viewed, this will probably exceed the allowed bitrate. To get around this problem, our white box restricts live viewing on two or more channels to a 4CIF stream even for 1 MP and 2 MP cameras. If I want to replay or remotely link into the NVR, this will further reduce viewing or replay bitrate capabilities.
Populating the box and hooking it up to the network
Now we have purchased our white CCTV box and we need to purchase a suitable hard drive. Most camera and recording device manufacturers should provide you with access to a storage and data rate calculator so that you can confirm each of your cameras average bitrate and how much storage they will require. The largest hard drive you can use is a 4 TB hard drive as stated, at a current cost of around R3250 incl VAT.
We are going to choose a 4 TB hard drive, as we know that our chosen 4 x 1080p and 4 x 720p cameras have mid-level bitrates of approximately 7 Mbps and 3.5 Mbps respectively. Using our storage calculator at 25% motion for all eight cameras during the period per day gives us a total likely daily storage of 116 GB, over the 30-day period. This gives us a total monthly storage of 3.45 TB. We can now hook in the Cat5e cabling providing PoE to our cameras and stream the eight video feeds, hook into our router/switch from the single non-PoE network port and stream video to our local or mobile network, finally hook up via VGA or HDMI to view live or recorded detail.
Our CCTV white box is ready to go and like all CCTV devices with live, recording and playback options, it needs video management software (VMS) of some sort that may well be referred to as firmware; the VMS is licensed to you as the end user within the terms and conditions. The cost of the VMS is factored with the cost of the hardware to give you a full working solution for the retail price paid. There are any number of manufacturers providing their own VMS within a ready built COTS (Customisable off the Shelf) solution all at a single cost and all containing licensed VMS software.
Your white box VMS should be fully compatible with your camera options; they are after all manufactured by the same company to ensure that you are locked in. Older cameras and those of a different manufacturer may not be compatible or will have very limited functionality through an ONVIF option. This is what keeps the VMS licensing cost down and ensures that you stay loyal to the manufacturer. What about adding or changing a camera six months later, well your choice is still probably limited to that same manufacturer. Do you ask questions about licensing and upgrades of software/hardware when purchasing your NVR/camera security solution?
The manufacturer will update and make changes to their NVR VMS on a regular basis, the same for the camera firmware as bugs or issues are fixed and amended. Do you actually know where you can find the VMS or camera firmware upgrade and what is more important, do you know how to upgrade this firmware? Many resellers often struggle to upgrade a camera let alone the NVR.
This is often the difference between a camera manufacturer providing the NVR or VMS, compared to a company such as exacqVision that specialises in VMS development. Vision Catcher is a distributor and fully trained reseller for exacqVision enabling you to scale the VMS to the workstation needed for the project. You can then choose from a long list of camera brands and models best suited for your requirement and change brands according to best of breed choice for expansion at a later date. Vision Catcher will ensure that bitrate, storage, pixel count and resolution are all factored to ensure that your solution has the longevity required to outperform and outlast the scaled down products at low prices.
There is nothing wrong with purchasing a CCTV NVR box, it may be just right for purpose. However, Vision Catcher focuses on your preferences, provide recommendations for hard drive space, storage capabilities, bandwidth requirements, viewing of live or recorded footage, camera brands and the options for newer models. All of these questions should be asked prior to purchasing your low cost NVR option; you might find that for a little more money you have a fully scalable solution with so many more enterprise level features.
|+27 11 463 9797
|More information and articles about Vision Catcher
© Technews Publishing (Pty) Ltd. | All Rights Reserved.