Video management systems, image quality, compression control, bandwidth--these are all crucial factors video solutions manufacturers must keep at the top of their list when creating a product for their customers. This month, we’re giving you a glimpse of what leading industry professionals are doing to keep up with changing video solutions technology.
What are the most important criteria for effectively managing video on the network?
Steve Surfaro, strategic channel manager, Axis Communications, Chelmsford, Mass.: Match available bandwidth with primary camera function, resolution requirements and frame rate. Typical camera functions can include observation, forensic review and recognition. Use network protocols to manage, route and distribute video streams as required to multiple users, prioritizing their use as required. Use well-designed and deployed physical infrastructure to distribute power to network video sources.
Dr. Bob Banerjee, product manager IP video, Bosch, Lancaster, Pa.: Respect bandwidth as a finite resource. Many system designs assume they have a lot more bandwidth than they really do while grossly under-estimating the amount of bandwidth required for a particular solution they are putting in place. Good compression technology is helpful for transmission; so too is transmitting video only when necessary.
Eli Gorovici, president/CEO, DVTel, Ridgefield Park, N.J.: There are two critical components to look at--bandwidth and storage. They are the key factors in keeping costs down and allowing video to share the IP infrastructure investment.
Rob Blasofsel, access/video integration manager, Honeywell Systems Group, Melville, N.Y.: It’s crucial to collaborate with the end-user’s IT department early in the installation process to focus on meeting the facility’s security policies. It’s also important to incorporate the manufacturer’s recommended best practices for products involved in the installation, such as factory recommended configurations and network bandwidth calculation/storage calculation tools.
Mickey Lavery, account executive, I2C Technologies Ltd., Uniontown, Ohio: To effectively manage video on the network the most important criteria is the choice of a video management software which minimizes file sizes, allows you to adjust different frame rates for different cameras, incorporates motion sensors to activate recordings and provides user-definable filters for fast and efficient archive retrieval.
John Monti, vice president of Marketing and Business Development, Pixim, Mountain View, Calif.: The most important criteria for network video is the quality of the images. Without good pictures, all of the other expensive equipment purchases, installation and maintenance are meaningless. Since video capture quality must be maintained over a wide range of applications – from drug stores to sports stadiums, airports, casinos, retail and banking – a state-of-the-art all-digital solution is preferred.
Rich Anderson, chief technology officer, SAMSUNG GVI Security, Carrollton, Texas: Integrators must be aware of where the customer will be storing the video. Trying to put a camera on the edge of the network and bringing it all the way back to the central station is doing the customer no favors. You are simply overloading the network with unnecessary video. Instead, store video as close to the camera as possible. With multiple sites and in a single facility, put a recorder in the closest possible closet to a camera. This lets you isolate video streams to keep them off the general network.
In 4-6 steps, how does your company effectively manage video on the network?
Surfaro: 1. Establish network camera viewing and recording requirements, 2. maintain QoS and provision the video traffic, 3. use port authentication (802.1x), MAC filtering or other methods, appliances and applications to maintain logical security on the network, 4. align video storage location with bandwidth availability, 5. provide remote diagnostic tools to monitor, diagnose and troubleshoot network issues.
Banerjee: 1. Use MPEG-4 (including H.264) in preference to Motion JPEG, 2. watch bandwidth and storage needs, and record at the edge if that helps, 3. move the intelligence to the edge through embedded video analytics so you don’t have to stream large amounts of video 24/7, 4. multichannel encoders with two network ports allow you to record at the edge by going Direct-to-iSCSI, which means zero bandwidth utilization, 5. keep all devices multicast-enabled which can reduce the amount of traffic on your network, but increases complexity of installation.
Gorovici: 1. Choose the right compression standard, 2. use advanced multicast technology (a network standard), 3. boost the quality of the video to increase efficiency in its management, 4. use Adaptive Visualization Technology (AVT) which provides the ability to more effectively process algorithms, 5. use camera sequences so that many cameras can be viewed over the same bandwidth as a single camera viewed full-time.
Blasofsel: 1. CODEC selection based on user requirements, 2. utilize VLAN, 3. minimize routers, 4. isolate IP recording traffic on layer edge three switches.
Lavery: 1. Determine the proper placement and amount of cameras needed, 2. determine whether cameras will record continuously or by specific events, 3. find out how many users will be viewing live or archiving video on a daily basis, 4. define whether the company’s existing network infrastructure can support the video, 5. set the frames-per-second (FPS) that is required without affecting overall performance, 6. calculate the amount of server storage that is required to accomplish the archived video requirements.
Monti: 1.capture accurate images in all lighting conditions, 2. have a global snap shutter to remove motion artifacts versus other cameras using analog image sensors, 3. support both CCTV and IP infrastructure with identical image capture to ensure consistency, 4. reduce network bandwidth and compression.
Do you think the industry is moving to separate networks for security versus using the existing data network? What’s the value in that approach? What skill-sets might it require for an integrator to build their own network?
Surfaro: Not really. The sophistication of virtually separating networks, rather than physical separation allows radically different content, like voice, data and video to co-exist on the same network. The value of physical network separation is ease of designing, provisioning and commissioning the video application on the dedicated network. An integrator will need to understand the guidelines for deploying a physically separate network within a corporation, even if it is not shared by other applications.
Banerjee: I agree that for peace of mind, it is moving towards independent networks which cross over at a certain point (gateway), with people who are on the data network able to view the video on demand.
Gorovici: Absolutely not. The only real value in a separate video network is the ability to more easily provide isolation in the event of a power failure so that battery backup can be provided for those critical cameras that are supported by Power over Ethernet (PoE). Today, networks are fairly simple to build and maintain if you follow the standards. It is always important to have staff members who have IT and computer skills in order to support both the network and video solution.
Blasofsel: Separate networks certainly have their place in the mission critical market segmentation based on internal and external threat analysis. We see users today utilizing existing network infrastructure by creating VLANS in the small- to mid-size commercial market space. Key skill sets include: network topology fundamentals, understanding of IP, firewalls, secure network devices and sources of internal/external threats.
Lavery: As prices of fiber optic cabling continue to decrease businesses will be pulling multiple strands of fiber throughout their facilities. It makes sense to separate the video network on its own dedicated network if it is available, eliminating any bandwidth concerns. The skill-sets required by the integrator include installing fiber optics, or in some cases, wireless radios. Integrators needs to be familiar with switches, routers, VLANs, firewalls and Virtual Private Networks (VPN) for wider access to video archives.
Monti: Systems can use either existing or new network hardware. In the case of legacy systems, it is common for a new IP network to be constructed alongside the existing CCTV network. This way previous equipment investment is not lost. Network video transmission, display, and storage requirements vary radically from data networks, so existing systems may not have the capacity or security infrastructure required for video security.
Anderson: In most cases, using separate networks makes little sense. When installed and maintained properly, a single network should be able to support a considerable amount of video. In the majority of applications, you want to strive for one cable, one set of equipment and one group of people supporting the network.
What’s the biggest no-no in managing video on the network?
Surfaro: Not using tools provided by the logical infrastructure manufacturer or other software supplier to manage the network. Sniffers are an example of one of these tools that analyze traffic and see patterns. A poor network design can create bottlenecks, even if you have a substantial amount of bandwidth available.
Banerjee: The assumption that bandwidth is free – it may be on a LAN but not a WAN, and that is a major application area for IP video. Also, the notion that multicast and assembling multicast is easy because it requires a lot of switch set-up. Many systems fail because installers were not network certified.
Gorovici: The most important element is to follow standards. The IT world follows known and published standards and most manufacturers build their products to those standards. The biggest no-no would be to use proprietary equipment followed by using older compression technology that was not built for a security environment.
Blasofsel: Not involving the end-user’s IT department in system deployments, which may potentially leave the system vulnerable to violations of security polices that enable destructive behavior.
Lavery: It is giving video network unmanaged access to multiple users. Determine who can access live and archive video and from what computers they can access this information from.
Monti: The biggest mistake with video management is forgetting to consider the importance of capturing useful video in the first place. Significant care must be taken to specify the right cameras and video encoders for the job. Many assume megapixel cameras have better images, most megapixel cameras have much poorer signal-to-noise ratio (SNR) in dim lighting. So performance of high pixel count cameras in low light tends to be poor. The noise will clog networks with high video bit rates and storage requirement containing no extra benefit.
Anderson: IP video must be implemented by an installer with considerable experience, using good design practices. We encourage our channel partners not to spec everything on their own.
What’s coming down the pike in technology or software that might make a big difference in effectively managing video on the network?
Surfaro: Treat the video as data, and realize that there’s a good deal of software to manage performance on your network and provide analysis and early identification of potential issues. Dark screen monitoring comes to network management with tools like nGenius that provides for more meaningful alerts that there are network issues.
Banerjee: There will be a realization in the marketplace that bandwidth is limited and how do I reduce my dependency on it? There are two mindsets that support this going with either IP cameras or DVRs; one is totally dependent on bandwidth, the other, is not. The marketplace also consists of other options including an encoder with storage built in, encoders which can be attached to an external storage source such as USB hard drives or iSCSI RAIDs and IP cameras with internal storage.
Gorovici: Two major areas are emerging for effectively managing video on the network: The new compression standard H.264 and the improvement of video quality through high compression MPEG -4 cameras, both of which enable better video quality and compression. Each of these image processing algorithms keep the video burden to a minimum on the network.
Blasofsel: Higher compression CODECs with more efficient workstation decoding is an evolving technology that captures the interest of most IT stakeholders. The output is lower bandwidth consumption and a smaller storage footprint while maintaining a high quality experience for end users.
Lavery: Higher compression rate and higher capacity medium are being introduced to the industry. This includes the use of H.264 compression which will increase the use of megapixel cameras that are too burdensome to most networks at this time. Also, the increased use of fiber, media converters and 802.11n wireless technology will increase network capacity.
Monti: Embedded, distributed intelligence will have the largest impact on the video security industry over the next ten years.
Anderson: Video storage and management should be done on a software platform on a conventional server. Most VMS platforms, software is run on a box, which adds a lot of complexity, costly especially for smaller applications.