Our Man in the Field: The Process of the IP Solution, Part VII

Our man in the field begins his in-depth look at video storage solution with a quick trip back in time


If you break down what is meant by real time according to common use of the terminology, then the first problem you run into is that real time is a matter of interpretation. If you are in the U.S., Canada, Japan, and parts of South America, your understanding of real time is different from the rest of the world. This is because you all work on NTSC standards, while the rest of the world works with PAL standards. This means that your interpretation of real time is 30 images per second, while the rest of the world perceives it as 25 images per second.

What real time really means

The fact is that both sides of the world are wrong. Here's some more history for you. Real time is a slang expression that came from the amount of time it took to use a full reel of tape on a reel-to-reel recorder. It was adapted to the cassette player and somewhere along the way, people slowly forgot the lessons of their grandparents and misspelled the word "reel" as "real". With that new spelling came a new misconception of reel/real time.

Lesson one in digital image storage: What is real time? Real time is relative to the application and should never be confused as a standard. Be it one image per second or 25 images or even 30 images per second, no set number of images per second constitutes real time.

So the next question that should come to your mind is: "How many images per second do I need to store?" And the answer is that I don't know. What is your application? If you are mounting a camera in a bank lobby and the purpose of the camera is to produce an identifiable image of all persons that pass into and out of the lobby, then your math works out as follows:

  1. The average time for a person to walk across the lobby is three seconds while the fastest time that a person can run across the lobby is one second.
  2. The application requires that one image of all persons entering or exiting the lobby is recorded.
  3. There is a slight blind spot directly under the camera that will account for a half second of lost imaging.

All right, based upon the information provided, what is your opinion on how many images, per second you should record of this bank camera Is it just one? Do you need two per second?

Personally, I would recommend between three and five images per second, three images being the minimum and five images being slightly overkill. Three images per second guarantees that I will have at least one image of a person running. Since I always plan for worst case, I am covering my bases. In the event of an alarm of some sort, I could have this image rate increase exponentially. If the lobby is a high-risk area, I could start recording 30 images per second. If the lobby is a "so-so" risk, I could leave it at three or five images per second. Again, the application will drive the design and requirements. This is the process that is used to determine the minimum required images that are stored for each camera in a system.

This is also the first point of consideration when recording with digital systems. If your system design requires flexibility because of multiple, intertwined applications, you will look for features in your storage system that allow for independent programming of images per second, per camera. Additionally, if you are automating your system through integration with alarms and/or other such triggers, you may also want or require that the storage rates of images per second is able to be increased or decreased on a per camera basis during specific events. This whole concept is the strongest advantage of digital storage systems as opposed to VHS.

What part of the video to store

The next logical question that comes to most minds is: "Why can't I just record 30 or more images per second from all my cameras?" This answer also dates back all the way to the reel-to-reel recorder. First, the timing sequence of images that are or were produced by cameras was and still is based primarily upon the frequency of the input power. Those areas that work on the NTSC standards are areas where the locally provided power is 60 hertz. In PAL portions of the world, the power is 50 hertz. Divide either by two and you have your standard images per second as created by the camera for each standard. Do we have cameras that produce more than the standard 25 or 30 images per second? Yes, we have units that can produce 10,000 or more images per second. However, these are not the typical cameras that are found in most security situations.