Surveillance Recording Mode Statistics 2011By Ethan Ace, Published on Nov 28, 2011
The question of what recording mode to use is an ongoing debate in the surveillance industry. Some maintain that motion based recording is the only way to go, as continuous recording is a waste of hard disk space, and can add significantly to investigation time. Others are skeptical of using motion based recording, because of incidents where motion recording missed important events.
To see how integrators are actually deploying systems, we asked "For recording type, what do you use most often?" Here's the breakdown:
A strong majority of integrators (63%) preferred motion-based recording about three times the number choosing continuous recording. The remaining ~15.5% using boost up/speed up recording, recording continously at a low frame rate, and speeding recording on motion.
Most integrators record on motion only, with the number one reason by far being storage savings. It is not uncommon for scenes, especially in hallways and stairwells, to only experience motion 30% of the time or less. Recording on motion in these environments has the potential benefits of reducing the number of hard disks required or increasing the retention period for archived video proportionately. Scenes with 50% motion, for example, may theoretically reduce hard disk requirements by half, or double the retention period. Some users may prefer to have this extended retention period, while others prefer to save costs up front, by using fewer drives. This savings may be substantial in systems with large camera counts.
Another advantage of motion-based recording, cited multiple times, was the reduction in the amount of video operators need to review in case of an incident. Unlike continuous recording, operators searching archived video will only see motion events, which may drastically reduce the amount of video they need to review. This assumes that motion detection is set up and working properly, a common concern cited by those using continous recording.
Integrators in favor of motion-based recording had the following to say:
- "Typically it's a matter of cost of storage. It also depends on the environment. Most applications don't require continuous recording of a stairwell that is rarely used."
- "Most of our customers deploy numerous cameras and want to store their video for at least 30 days AND have high video quality. Thus, we use motion based recording to help them achieve their goals AND stay within their budget for their data storage."
- "We feel that it makes more sense to record activity and not open fields with nothing to view."
- "To save storage and find events quicker"
- "The majority of our customers are interested in utilizing video for forensic purposes. They are also concerned about storage costs and file sizes. Motion based recording offers a good blend of recording when needed while saving space when not needed."
- "Storage space efficiency, mostly. Makes searching easier, too, when you don't have to scan through times with nothing going on, especially for sites that are only open 12 hours per day."
- "Motion Based Recording is the efficient way to optimize video records. It's very important feature when we looking for interesting events records - easy and fast."
- "Why record an empty room? With a suitable pre and post record setting, it will cover all applications."
- "Why would you want to waste hard drive space and view a bunch of video where nothing is happening?"
Proponents of continuous recording cited customer demand as the most common reason for choosing this recording mode. Many industries, such as gaming, municipal, and critical infrastructure demand 24/7 recording. In some cases, it is due to regulations, and in others simply because these larger systems can make the demand, with the budget to back it up. Additionally, many municipal systems use mostly PTZs, making motion-based recording a moot point, since cameras are constantly in motion.
- "Our customers generally require 24x7 recording for evidential purposes"
- "It varies by customer. We are astarting to use more boost recording, but for airports, water plants, and large municipal accounts, they demand continuous recording."
- "Storage is cheap, for the most part, and most of our customers specify it that way."
- "Most clients want that feature of everything that goes on in front of the camera. Some are required by law to have at least 30 days so motion based is not an option. Also since storage has become so cheap, clients don't mind pay a little more for extra storage so they can that extra week of record time."
- "It varies by customer. We are astarting to use more boost recording, but for airports, water plants, and large municpal accounts, they demand continuous recording."
Unreliability of video motion detection and fear of missed events as the most common reasons for choosing this recording mode. Some of this, we expect, is due to earlier implementations of video motion detection which did not perform well. However, with motion detection often being performed on cameras, now integrators have the added challenge of knowing how well each model of camera performs, which may vary between manufacturers, and even in a manufacturer's own line. Motion detection may also be performed server-side, at the expense of higher CPU utilization.
- "Don't want to take a chance on missing something at a really low frame rate if activity is not properly detected. Storage is cheaper than ever so storage is not so much of an issue as in the past."
- "I am afraid that I seem to have a fear of missing an event from motion based recording, although this should not be the case. We have had problems in the past with motion based being done by the VMS and as the PC server was underspeced then the loading on the CPU was very high. By using the motion detection in the cameras the loading was reduced. However at that time the built in motion detection in the cameras was not brilliant, particularly in low light and so it was deemed sensible to do continuous recording at a reduced frame rate."
- "Because we think it is more reliable."
Boost Up/Speed Up Recording
Lastly, integrators choosing boost up/speed up recording typically saw it as a best of both worlds approach. Low frame rate continuous recording assured that at least some video of an incident is captured if motion detection should not trigger. Yet, motion events are captured at a better frame rate, for better evidence of incidents. Boost up recording does have some challenges, however, as we discussed in our Motion Boost Recording test. Integrators had the following to say about boost up recording:
- "We generally use a relatively low quality continuously recorded image, ramped up for motion to assure admissable evidence in a potential prosecution or civil matter."
- "I like to see everything. Even with the Sony's that we are using I notice that we can sometimes miss parts of a recording."
- "I like 1 IPS plus motion based speed up save space yet still have a time lapse."
- "It's just the right thing to do. Motion detecftion is never accurate and loses key video or just records constantly anyway."
- "We've found that it's best to have a baseline of video in case motion doesn't trigger. There was also a law suit because of pure motion recording as there was some arguement about what happened before and after."
Small Systems Continuous Recording
Though these results were generally the same across users of different VMSs, those deploying more megapixel than SD cameras, and other factors, one noticeable difference was evident in those deploying 16 or fewer cameras per site. Among this segment of integrators, those using motion-based recording decreases by 13%, from 63% to 50%, with those using continuous recording increasing by 5.5%, and motion boost by 7.5%. We suspect that this may be due to a tendency to use extra hard drive space in small systems. Larger systems are usually built with slimmer storage margins, normally only equipping systems with the storage that is absolutely required. Small systems, on the other hand, are often deployed using prepackaged NVRs or smaller PC hardware, with standard 1TB or 2TB drives, which may not necessarily be required. Instead of simply not using this extra storage space, integrators may configure the system for continuous recording simply because they can.