After shedding light on the calculation of the bandwidth (BW) of a communications channel to transmit video, now we will talk about how to calculate the capacity of the Hard Drive of a Digital Video Recorder.
by Germán Alexis Cortés*
Many times in the electronic security industry, it happens that we sell CCTV systems without doing the proper analysis. In fact, only a few people know exactly what calculations should be made and I believe that very few of them actually do. The rest are content to choose a brand of equipment and check which model offers a lower price.
There are many analyses that are needed, this involves making some simple calculations and placing common sense at the service of the security of our customers. If we don't, we can put the entire credibility of the electronic security industry at risk, out of simple ignorance or laziness.
One of the frequently asked questions I receive in the courses, which I direct for ALAS, is how much hard drive do I need? Let's see in these short lines an easy and complete way to do the calculation.
The first thing we must understand is that the answer is not immediate, that is, no one, responsibly can say XXX Giga Bytes per day or per hour or per month, without doing the necessary analysis. Some people usually throw a figure in the air, product of their experience, but it is almost always wrong, when compared to the calculated or with the one that is actually used, when verifying a system that is operating satisfactorily.
The appropriate response depends on the following variables:
1. Number of channels or cameras of the VCR.
2. Resolution at which I want to record each camera.
3. Speed at which I want to record every video signal.
4. Compression Algorithm.
5. Quality I want in the stored video signals.
6. Type of complexity in each image.
7. Average size of each stored image.
8. Daily recording time.
9. Activity of the scene I wish to record, according to the operation and schedules of the site.
10. Recording method (continuous, by events, by time lapse)
11. Amount of information to be stored (pre-alarm-postalarm)
12. Importance you want to give to each scene.
All these aspects must be collected, analyzed and used for the final calculation.
Normally manufacturers offer online calculators for this purpose, however these tools do not always take into account all aspects, and especially the common denominator works based on averages, which ultimately make a very important difference; therefore I recommend doing the calculations in an Excel sheet or similar and ensure that the hard drive we buy is the right one.
The initial methodology is very simple:
to. Calculate how much it takes to store one (1) second of video.
b. Multiply this result by the number of seconds you want to store. Remember that an hour has 3,600 seconds, that a week has 168 hours and a month 720 hours.
c. Keep in mind that this should be done for each camera, for each scene and for each segment of the day, because the conditions are different in each case.
The formula for calculating the space for one (1) chamber, in one (1) second, is just as simple:
Hard Drive for 1 second = Average Size of 1 Frame (Bytes) x FPS x % Activity
That is, the Capacity in Bytes for one second of video for one (1) camera, is equal to the average size in Bytes of each Frame, multiplied by the speed (FPS) at which you want to record, multiplied by the percentage of activity of the scene.
Let's begin then to analyze, each of the variables. The average size in Bytes of each Frame must be given to you by the manufacturer of the DVR or NVR you are using. This size is what point seven (7) refers to that we will see later.
The recording speed is determined depending on the application and the scene you have.
The percentage of Activity indicates how static (0%) or changing (100%) the scene is. Remember that to multiply a percentage, we must pass the number to decimal. So multiplying by 100%, it's really multiplying by 1, multiplying by a percentage of 80%, it's really multiplying by 0.8 and so on.
Once we know how many bytes it takes us to store a second of video, it will be very easy to know how much in an hour or day. However, remember that 24 hours a day must be segmented and that in each segment there may be a different average activity or a different scene, which leads us to change the size of each frame.
Let's analyze in detail the other aspects that influence the calculation, in this first installment we will include the first three points:
We must understand that each one is linked to the others and that there is a direct dependence that decisively influences the capacity of the storage that we calculate on the hard disk.
1. Number of Channels
It is obvious that the more cameras (channels) you want to record, the more hard drive you need. The relationship is linear, however remember that not all cameras behave the same, simply because each one is looking at a different scene. The following features should be reviewed for each camera. An important recommendation is to keep in mind that not many cameras should be placed in a DVR.
Even if the equipment has the computing capacity for it, which is quite difficult to fully comply with, by simple common sense applied to security. Notice that in case of failure, it is better to have several units, so as not to lose the information of all the cameras. As the world of computers, is bringing to the mass market, units with multiple processors in parallel, great speed and large memory, we can increase the number of channels in the same DVR, however we already know that for security we should not do it.
2. Resolution
In the digital world it is the size of the image, expressed in horizontal pixels by vertical pixels. The larger the resolution, the better sharpness and detail I will achieve, it is easy to understand that a greater amount of information needs a greater storage capacity. The native (original) resolution of many cameras (analog or IP) is close to VGA (640x480) or higher, so if you record at a lower resolution, for example CIF (320x240), you are making a serious mistake, because resolution is wasted.
There are those who claim to record at higher resolutions, by showing the image in a larger size than the native (for example increasing it with the mouse when dragging it from a corner), but this is a deception because the really important thing is not how big the image is on the monitor, but how many real pixels it has horizontally and vertically.
Some IP cameras of recent technologies have a high-resolution sensor (almost always in the order of MegaPixels), therefore these units must be recorded at that native resolution, otherwise again the opportunity to store with excellent resolution and detail is wasted. The world of IP, so fashionable today, brings precisely units with very high resolution that can be used to the maximum by digital zoom, however we must be aware that they will occupy much more hard disk space.
It is important to understand that as long as it is recorded at the native resolution offered by each camera, the stored video will be the best possible. Only when we are aware that the image of a channel can be of lower quality, because we do not care about the details or sharpness of the image, we can decrease the recording resolution. For security applications this is not recommended.
The concept of pixels per meter, in conjunction with the calculation of the lens, are fundamental to obtain the specification of the camera that is required. The idea is that the VCR respects that resolution.
3. Speed
The more frames per second we record of a scene, the better the video will look. The traditional maximum limit is in the commercial standard (NTSC, PAL, SECAM) that applies. Many people in electronic security, come accustomed to recording for periods of time, that is, to record few frames per second, losing valuable information, however there is nothing better in the area of security, than to see an image stored at good resolution and at excellent speed, that is, 30 FPS (NTSC).
This makes the difference between using the recorded video efficiently in the investigation of an accident, or having the typical feeling of frustration knowing that something unreadable was recorded that is of no use to us. With today's advanced image analysis systems, you can decide to always record at maximum speed and resolution, but only when necessary. Even so, I must mention that maximum speed is not always necessary, it all depends on the need and risk that each scene has.
When the application of the CCTV system is different from security (for example control or only supervision), it is common to record at low speed, however I warn that in these videos it is not possible to identify a stranger, nor are scenes obtained that can serve as evidence in a legal process.
I must remember that a digitally stored video scene is valid at the legal level in North America, Great Britain and some Latin nations, when it meets three (3) requirements at the recording level: a. That has the native resolution of the camera that captured the image. b. That it is recorded at the native speed of the standard, that is, 30 FPS and C. That it is guaranteed that the scene has not been manipulated or altered by a third party (Watermark).
At what speed to record then? Always record at 30 FPS in times of alarms or critical situations. In the rest of the time, it depends on the application and the movement that exists in each scene. The higher the number of FPS, the better. However, for some applications 2 or 4 FPS will be enough, for example to control if employees are in the office. It should be understood that employees are known and in many cases just seeing a vague silhouette is enough.
In applications where a visitor (stranger we do not know) enters the scene, it will already be necessary to climb to 15 FPS minimum, with the aim of clearly seeing the actions of unknown characters. In scenes where critical and important processes are handled, it is not advisable to lower from 20 to 24 FPS and in scenes where there is money, casino games, very important personnel, detailed processes, always at the maximum speed (30 FPS in NTSC).
When the images do not come from fixed cameras, but from PTZ mobiles, we must always record at least 10-12 FPS, in scenes of very low importance, at 20 FPS in normal scenes and again at 30 FPS in critical and important processes.
Remember that the speed in the recorded video will influence the sharpness of the image (not its resolution); which is after all what allows us to see details in the scene. Do not throw away the high investment required by a CCTV system in the trash. Always record at the highest resolution and speed you can, I assure you that you will not regret it.
Wait in the next edition of SECURITY SALES, the second part of this article.
*Germán Alexis Cortés H. is a Colombian Electronic Engineer, with a postgraduate degree in Engineering Management Systems. Consultant of the National Institute of Standards and Technology – NIST, in Electronic Security and Building Automation. CCP of ASIS, and CISSP of ISC2. He has over 20 years of experience in the electronic security industry. Renowned lecturer and university professor on high-tech issues at the Latino level. He has been a director of several companies in the sector and is currently a partner and directs Insetrón Ltda, an engineering and technical consulting company in electronic network projects for security, communications and automation. He has successfully participated in more than 150 projects at the Latino level.
Leave your comment