Evaluation Metrics for Storage Savings

Published Sep 15, 2009 00:00 AM
PUBLIC - This article does not require an IPVM subscription. Feel free to share.

Properly evaluating TimeSight's extreme storage savings claims depend on carefully analyzing TimeSight's assumptions and contrasting that to competitive technology and common deployment scenarios. We examine this herein.

TimeSight has to be evaluated as a VMS

TimeSight is frequently but unrealistically evaluated as a storage appliance or a storage system. However, unlike true storage systems such as Pivot3 or Intransa, TimeSight is not compatible with any VMS. As such you cannot feasibly connect TimeSight appliances to Gentec, Milestone, OnSSI, etc., etc. TimeSight has to used as your VMS if you want to use TimeSight for storage management.

Storage savings are, therefore, only one component. Any storage savings need to be contrasted against other deficiencies of TimeSight as a VMS - User Interface, Access Control/3rd party integration, scalability, etc.

MJPEG Comparisons

TimeSight's marketing material [link no longer available] assumes the use of MJPEG as the CODEC choice for the alternative system.

This was never realistic for SD cameras, and is increasingly unrealistic for HD cameras. When you assume H.264, the correct assumption for 2009 deployments, the storage savings drop dramatically (simply because H.264 reduces storage consumption by 50-90% relative to MJPEG - my test results, not marketing material).

HD Comparisons

TimeSight's marketing material assumes the use of all HD cameras. Today, this is rarely the case. Even in the future, while H.264 HD cameras use will increase, the majority of cameras will likely remain SD for years to come.

Even if the percentage decrease is significant, in dollar terms the savings for SD cameras will be small, simply because SD cameras do not consume that much storage.

Large Camera Counts

For the savings to be financially significant, TimeSight has to assume 100+ cameras. For sites with 16 or 32 cameras, even if the percentage savings are significant, the dollar value will not be. Adding an extra 2 hard drives to a PC/server or adding a NAS array with 4GB is simple and inexpensive.

Long Term Storage Comparisons

TimeSight's marketing material assumes 3 months to 2 years storage duration. This generates a significant portion of the total claimed savings - simply because they are increasing total storage duration by 2 to 10x more than commonly done in deployments.

For those that really value long term extended storage, this is important. For others, it is just misleading.

Competitors Assumed to Have No Storage Optimization

TimeSight generally assumes that competitors are recording continuously and that they have no analytics or storage optimizations of their own. However, almost any VMS system offers motion based recording, which can reduce storage consumption by 40-70% relative to continuous recording.

Also a number of manufacturers offer frame pruning or reduction of storage over time - Pelco Endura, March Networks and most recently ipConfigure come to mind. Additionally, numerous manufacturers are using analytics to significantly reduce storage costs including 3VR, Aimetis and VideoIQ

Doing Comparisons

TimeSight could likley make the following claim, "For large scale megapixel only systems using MJPEG for long term storage compared to a VMS recording continuously and with no storage optimization features, TimeSight has dramatically better storage efficiency."

This is probably true but it's also a niche case.

The comparsions that industy professionals need are for more general deployments with mixed cameras types, using H.264, for shorter time frames and compared to the many systems that do offer storage optimizations. Those will certainly be smaller but more realistic savings.

[Update 2012: TimeSight is out of business]