Today, there are many options for video storage, from DVDs to hard drives and SATA storage. That leaves dealers and integrators who install surveillance systems of 16 or fewer cameras wondering which video storage approach they should recommend to their clients, based on considerations of cost-effectiveness and practicality.

For Dennis Smith, director of integrated services at Duncan, S.C.-based SFI Electronics, the answer is simple. “It’s better to just have larger hard drives,” he says. “It’s functional, economical, and in terms of quality, it’s just about equal. You should be able to offload an exact copy of video you have on your machine, with no degradation in quality. But once you get this information into storage media, playing it back may require reloading it onto the DVR, which can be time-consuming and cumbersome.”

Smith says his preference for hard drives is based on their affordability, with 1-terabyte drives available for about $300. Moving from a 500-gigabyte to a 1-terabyte recorder adds about $1,000 to the cost of a DVR, he says.

Some integrators can upgrade their DVRs by installing larger hard drives. That suggests integrators might be best off buying smaller DVRs and installing larger hard drives. “They run the risk of voiding the warranty, and not having support from the manufacturers,” Smith says. ”But often, that’s an acceptable risk.”

Todd Flowers, president of Surveillance Systems Integration, Roseville, Calif., believes that for video surveillance systems with 16 or fewer cameras, a DVR’s onboard storage will be sufficient in 95 percent of the cases. After all, he says, most retail applications only record about seven images a second, and those images need be saved only seven to 14 days.

“A lot of integrators will take a consumer-grade PC and create a recorder themselves, by installing a hard drive for extra storage,” Flowers says. “But usually, those are Windows-based and they’re very sensitive. You’re getting a system with no warranty. The warranty is the integrator himself.”

His statements are based on the assumption that such systems are using analog cameras. If Internet protocol (IP) cameras or extremely high-resolution megapixel cameras are utilized, the storage requirements grow vastly different. “That’s when you can easily get into the need for external storage,” he says. “Now you’re recording video not onto a digital recorder but a server, and there’s only so much storage you can record on that.”

Speaking from the manufacturers’ standpoint, director of technical services Ted Brahms of Carson, Calif.-based Samsung Techwin America says most dealers and integrators offering smaller systems will find the simplest and most cost-effective solution to be the current generation of embedded DVR systems.

Older video recorders using MJPEG and wavelet-based compressions were often limited by the size of internal storage available, lack of external expandability and inefficient video compression requiring massive storage.

But new generation DVRs feature multiple internal serial advanced technology attachment (SATA) hard drive bays. SATA drives are inexpensive today, at a cost Brahms estimates at around $150 a terabyte. As a result, adding greater storage is comparatively inexpensive. And when newer video compressions, such as MPEG4 and H.264, are used, one terabyte can store about a week of 24-7 recording at real time image rates, he points out.

“Simplicity of set up and operation for the installer and end user, the one-box solution, the reliability of an embedded Linux device over a PC-based system ­â€” these are all good reasons to recommend this type of system,” he adds.

“The reliability of the hard drives themselves has greatly improved over the last few years, as several hard drive manufacturers are now producing hard drives specifically designed for digital security video applications,” Brahms says.

A number of ways exist to gain more from available storage media. Smith urges using motion detection as an effective means of limiting storage requirements. “Lower frame rates in most cases are acceptable,” he says. “Down to one frame a second is probably acceptable in most applications.”

Brahms also advocates lowered frame rate. Reducing the frame rate to as low as five frames per second per channel extends storage of a one-terabyte system to a month or more, depending on image size and compression, he says. And that reduction can be achieved without impacting most applications, he asserts.

Using higher-quality cameras equipped with digital noise control can also help limit storage requirements, says Wayne Hurd, executive vice-president of sales and marketing with Digimerge Technologies, Toronto. Eliminating a grainy image in low light levels can be an efficient means of limiting storage requirements, he says.

Digimerge’s manager of technical services, Rick Buchanan, adds: “The digital noise control helps a lot with compression. The camera will cut down on the irregulars coming into the camera. And good quality cabling helps a lot. If you have any device causing interference, you can get wavy lines. You want to run a good, shielded copper-braided cable. That will cut down on interference and you’ll waste fewer pixels on distortion.”