My July editorial, “HDcctv: Dismissible or Disruptive,” discusses how the new kid on the block, HDcctv, is threatening to wrestle market share away from IP megapixel technology in the push toward high definition (HD) video surveillance. The article was based on a recent panel I moderated about this very topic, and featured Scott Schafer of Arecont Vision, Sara Scroggins of Pelco, Steve Surfaro of Axis Communications, and Todd Rockoff, executive director, HDcctv Alliance. Below, as promised in the column, is the extended conversation among these leading HDTV authorities.
Why bother with HD? It comes with costs, complexity and inconvenience, so why pursue it in the first place?
Todd Rockoff: Enough resolution is never enough in surveillance; some applications always need higher resolution than what is readily available, and NTSC/PAL/D1 resolution does not allow clear identification of subjects and other details in all cases. This question predates the availability of HDcctv; HDcctv is the only high-resolution video surveillance camera connection technology that delivers HDTV-compliant signals. For existing CCTV installations, HDcctv is convenient to implement, cost-effective, and not complex to operate.
Scott Schafer: The picture quality improvement, and cost savings vs. standard definition IP and analog camera systems.
Sara Scroggins: While there is a point of diminishing returns regarding increasing resolution levels in, for example, consumer photography, video surveillance applications demand and continuously take advantage of higher resolution. The role of video surveillance is to gather video data that is used in informed decision making to solve business problems. In real-time, that video data provides situational awareness and enables appropriate response to events. In post-event investigation, that video data provides valuable evidence of what actually happened. Better video data provides an input that can yield better, more effective business decisions – whether it’s responding to emergency events, preventing loss of assets, improving workplace safety, understanding customers' traffic flows, etc. Ultimately better decisions, aided by better video data, improve a company’s bottom line and improve any organization’s achievement of operating goals.
Steve Surfaro: HDTV is a standard and therefore guarantees the expectation of video quality, frame rate and color fidelity. HDTV is no more complex than multimegapixel imaging and in fact provides a more convenient and economical solution as video surveillance for observation, forensic review and recognition have a higher probability for successful deployment.
Taking into account the technology, marketplace, standards, reseller and end-user variables, what needs to happen to achieve broader acceptance of HD surveillance, be it megapixel network cameras or HDcctv?
Scroggins: In the video security industry, we are at the early stages of industry-specific standards. Notably in networked video security, our industry is already using and relying on a wide array of IT industry standards — from standard IT architectures and storage to USB, UPnP [Universal Plug and Play] and more. As is natural at the beginning of any new technology deployment, our industry has been using many different standards for video compression, as manufacturers seek different approaches. But as we move forward, there is increasing coalescence around H.264. And the important work of standards bodies such as ONVIF and PSIA will lead the way to broader acceptance of HD video surveillance.
Schafer: End users and systems integrators need to get better educated. They need to understand HD camera technology; system design — pixels on point for facial identification and licence plate regognition, for example; and all about networks, NVR servers and storage. They need to see excellent case studies that show value in the improved images and ones that show a good return on investment.
Rockoff: The key for widespread adoption is that the high-resolution surveillance must work the same as CCTV and cost about the same as CCTV.
Surfaro: Images and video clips from HDTV and quality multimegapixel devices are more usable when they come from devices that meet standards. Public safety professionals are now using HDTV because they have seen demonstrations and sample video clips that demonstrate use cases, like monitoring critical events for first responders, operating room surveillance and transportation security. The deployment of MESH wireless networks and the availability of network infrastructure of all varieties will not only encourage use of HDTV, but make it a device that is normally deployed, right along side a network switch.
Why is acceptance and market penetration taking so long? What might speed up the adoption process?
Surfaro: Adoption of HDTV is already taking place for nonsecurity applications, even more rapidly than security surveillance; it just depends on the industry you work with. Markets with a high degree of penetration include: medical (operating rooms, remote diagnosis); transportation (airport surveillance); entertainment and resort (advertising); and sports, both surveillance and player coaching. Surveillance and security requires operators need better situation awareness and HDTV’s image quality delivers that. As metrics are discovered that illustrate how operators are able to perform their job more effectively with HDTV, this will accelerate its adoption. Automated recognition-based systems that use video analytics perform more effectively on an HDTV platform, and these cameras typically have stronger processors, so they can run analytics and stream video effectively. Video analytics allow a guard force to be more aware of off-normal events by allowing HDTV systems to handle the burden of many routine tasks. For example, an HDTV camera, together with an access control device can admit returning contractors without the need for operator interaction. Security operators can be proactive with other tasks, reducing manpower costs.
Rockoff: HDcctv system adoption is gated by product availability. The adoption of IP network camera-based systems is gated by convincing end customers that there is a significant ROI in using IP networking as a camera connection technology. The availability of prototype HDcctv equipment is already accelerating end-market adoption of high-resolution video.
Schafer: It is taking so long because of the following factors: installed base of coax-based camera products and cost to upgrade to Ethernet networks; higher costs of cameras and NVRs early on; higher bandwidth and storage requirements, although these have been partially solved by H.264; NVR systems supporting H.264 had not available until recently; security resellers are cautious of new technology; and early implementations suffered from installers not properly designing and installing the systems, creating a gun-shy situation. The following factors will speed it up: education of systems integrators and end users; H.264 adoption by NVR platform providers; customers not accepting low image quality or events that cannot be identified; good case studies; consumer requirements for change; and adoption and familarity with 1,080p consumer TV, resulting in the desire to bring that to security departments.
Scroggins: As the cost difference between HD/megapixel cameras and SD cameras continues to shrink, the cost barrier to wider adoption of HD video surveillance will disappear. Notably, the cost difference between SD video and HD video lies not simply in the cost difference between SD and HD cameras. Along with the greater detail and performance of high-definition imaging systems comes the reality of increased bandwidth and storage costs. A typical 720p or 1,080p H.264 stream, for example, requires roughly 3-5 megabits (Mbps) per second of bandwidth, compared to a typical analog camera running through an encoder, which typically requires about 2Mbps or less. Cameras that that utilize high profile H.264 compression are able to deliver high-definition video while minimizing bandwidth and storage costs.
What are the best applications for HDcctv and megapixel network video surveillance?
Schafer: All applications in all vertical markets would benefit from HD/megapixel technology. Tight shots in entrances and small rooms where camera shots are tight would not be as easily benefited as wide field of view applications.
Scroggins: Any application that requires increased image quality and detail makes a good candidate for considering HD video surveillance. Think about the range of applications where detail matters: faces, license plates, playing cards, currency denominations and serial numbers. So HD is a natural for high-security sites (government, military, airports), for money counting operations, for gaming, corrections. HD/megapixel video also offers the ability to see more clearly across a wider field of vision, while providing amazing detail when a camera zooms in on a scene. Even at one of the lowest HD resolution settings (1MP or 720p), an HD camera captures three-and-a-half times more detail than a standard NTSC camera. And HD resolution provides the ability to digitally zoom in on recorded video, far and above what is available through standard-definition video.
Rockoff: Any application wherein the end customer desires high-resolution video is a candidate for an upgrade. HDcctv is ideal for upgrading one or two cameras in a CCTV installation without having to change the cabling infrastructure or involve costly, specialized IT skills.
Surfaro: In addition to those I mentioned earlier, city center surveillance; education (distance learning); and any application requiring the consistent delivery of high quality imaging for both live observation and recorded use.
When is it not appropriate, and what are some of the technologies' shortcomings?
Rockoff: Fundamentally, high-resolution video is not appropriate when the costs outweigh the benefits, or when the high-resolution camera connection technology introduces significant technical compromises. The many shortcomings of IP networking as a camera connection technology are not widely acknowledged by the surveillance equipment industry, although end users seem to be aware of the implications of these shortcomings. Network cameras degrade image quality by applying significant compression before live view and before storage; IP networking relies on sometimes unreliable networks, which drop packets, and therefore frames of surveillance video, or sometimes go down altogether. IP networking is fundamentally incompatible with CCTV as a camera connection technology, and network cameras require the implementation of an IT system.
Surfaro: There used to be a larger price difference between non-HDTV and HDTV cameras, but this has been greatly reduced. With the use of H.264 video compression, bandwidth consumption is also reduced. Packaging of HDTV cameras is now available in fixed box style, fixed dome and p/t/z camera form factors, so this has also been accommodated. The only area that HDTV cameras have not been able to impact is the OEM camera market, such as ATMs, automotive and marine safety, but this is expected to change in the next couple of years.
Schafer: When is HD digital television not appropriate?
Scroggins: HD/megapixel cameras still face some performance challenges in low-light applications, when compared with SD cameras. That’s why Pelco took special effort in developing Sarix Technology to overcome the challenges of low-light performance for megapixel imaging. Scenes with wide dynamic range may pose a challenge for some HD/megapixel cameras. But manufacturers are working hard to improve performance in these areas.
How do you see the market share between 720p and 1080p cameras given the tradeoff between sensitivity and resolution?
Schafer: With the price point so minimal, not sure there is a reason to not go at least 1,080p.
Surfaro: Simply put, if the application requires higher resolution or more pixels on target then market share will increase for that application.
Rockoff: It will be interesting to see how the market decides. Note that 720p and 1,080p are HDTV formats that are reproduced in the HDcctv standard, but no network camera transmits 720p or 1,080p because compression is not included in the SMPTE definitions of those television formats.
What is the benefit of transmitting uncompressed HDTV signals in CCTV applications?
Surfaro: Public safety, law enforcement and medical applications all benefit from the use of uncompressed video; however, with more efficient compression technologies like H.264, the application where uncompressed video is beneficial is decreasing.
Rockoff: The benefits include zero transmission delay, which is especially useful for controlling p/t/z, zero image artifacts from transmission, direct reuse of existing infrastructure, and the direct extension of the CCTV evidence gathering and management paradigm to resolutions of 720p, 1,080p, and beyond.
How will the network camera industry overcome the challenge of incompatibility among compression engines?
Schafer: Strong efforts among manufacturers of megapixel cameras and NVR players, with improved SDKs to improve integration time and support. Also key is standards adoption, such as PSIA and ONVIF.
Scroggins: As noted before, our industry has been using many different standards for video compression, as manufacturers seek different approaches. But as we move forward, there is increasing coalescence around H.264. And the important work of standards bodies such as ONVIF and PSIA will lead the way to broader acceptance of HD video surveillance.
Surfaro: It is not incompatibility of compression engines but the use of standardized API and interoperability that is the challenge, currently being met by standardized API or interoperability like ONVIF.
What is the process of adding one HDcctv camera to a CCTV installation? By contrast, what is the process of adding one megapixel network camera to a CCTV installation?
Surfaro: Deployment of HDTV or IP megapixel cameras is simple indeed with the use of Ethernet cables and a network switch. The use of HDcctv cameras may required more specialized DVR and video capture cards to accommodate the video sources’ increased resolution, not to mention the variable nature of coaxial infrastructure.
Scroggins: Adding a megapixel camera to a video security system can be easy or complicated, depending upon the manufacturer of the camera and the video management system. In some systems that are ‘HD ready,’ and in which camera and video management system are from the same vendor, adding a camera can be as easy as plugging a camera into a network and configuring the camera via a Web tool. If the camera is from one manufacturer and the VMS is from a different vendor, the process can get more complicated, involving multiple lookups of camera IP or MAC address, sending that info to VMS provider, awaiting VMS camera license for that IP or MAC address, then entering license info into configuration tool, etc. You can see how it can get very complicated very quickly.
Rockoff: To add one HDcctv camera to a CCTV system, upgrade the DVR to accept both analog and HDcctv inputs, selectively replace analog cameras with HDcctv cameras, or add new ones. By contrast, to add one megapixel network camera to a CCTV system, step one is to implement an IT network. The never-ending series of Webinars on this topic from vendors of network cameras highlights the complexity involved. The end user is subsequently tasked with maintaining the IT network.
Schafer: For megapixel cameras, if the CCTV installation has a network video recorder or a hybrid NVR/DVR, it is quite easy. Pick the camera location, connect the camera to the network that the NVR is using, set the camera for proper distance/lens use, focus the camera, and then collect fabulous images vs. the low resolution images being provided by the low definition cameras in the system.
What do resellers need to know to get started with megapixel network surveillance or HDcctv? In other words, what actions do you recommend for resellers as far as selling, proposing, delivering and supporting HD solutions?
Scroggins: Resellers need a fundamental knowledge of IP networking, because any significant scalable HD video surveillance system is going to be running on a network. They need knowledge of bandwidth and storage requirements for the cameras they’re specifying, and they need to be able to understand and plan for the demands the video application places on the network and the implications for switching infrastructure.
Schafer: Understand the network and the importance of network-based systems. Get educated on megapixel camera technology. Use the tools that megapixel manufacturers provide to determine the number of pixels on point you require, the distance from the subject you require, proper lens selection, when to select jpeg or H.264, how to maximize coverage and minimize storage time length, good understanding of the customers, etc.
Rockoff: No new learning is required for CCTV installers and technicians to take advantage of HDcctv. For resellers, it’s a simple matter of showing the end customer what kind of evidence can be expected with the higher resolution video. By contrast, getting started with megapixel network cameras requires extensive training in IT systems.
Surfaro: There are many classes and education opportunities for learning about IP solution deployment like the Axis Academy and the IP Institute.
What challenges are resellers likely to encounter, and how can they minimize the impact of these challenges?
Schafer: Use the manufacturer to assist in calls; go to training classes to improve knowledge; learn to demonstrate the products. Get the customer to try one camera, and the other 299 will follow. Learn to quickly qualify a prospect. If they don’t get it in 20 minutes, leave. They will call you back. Where it comes to system configuration: use manufacturer’s tools for configuration of cameras, network and server/storage design. Make sure to pick the proper optics package. Use manufacturer’s tools for set-up. Understand lighting; don’t go with too many pixels for the area as more pixels reduces image clarity in low light situations. Finally, regarding installation and support: again, understand the network; utilize a dedicated network; and provide a lifecycle maintainance program for customers.
Scroggins: Predicting network bandwidth and storage requirements is quite a complex problem due to the many variables that go into the calculation. Some manufacturers provide software tools to help estimate these requirements. Pelco provides such tools to its customers, which allow the consideration of a great number of variables. We mentioned the potential complications of adding cameras to a system. Some manufacturers are working hard to make the installation experience easier than ever, by developing new configuration tools, installation wizards, and more.
Surfaro: Matching the video source to the surveillance function, as well as performing a good video site survey are ways of minimizing design challenges.
How can they best make the ROI and TCO cases for their existing customers and prospects?
Scroggins: There are no shortcuts to understanding and articulating ROI and TCO benefits, because it comes down to each customer’s situation. The best way to make the ROI and TCO cases for HD/megapixel video is to start by understanding the customer’s core business. Where is their pain? What drives their profit margin? How is the customer using video currently to solve problems and improve profit or operations? How could better video help the customer make better decisions, improve operations, mitigate risk, eliminate loss, or use resources more wisely? That’s where it starts. The investment in HD video is an investment in business information. Ask the customer, ‘How much is it costing you to not have this information right now?’
Surfaro: Demonstrate the cost savings of the use of existing infrastructure. Demonstrate use of the HDTV system for other purposes than security, like safety, operations management and education. Demonstrate the measurable increase of useable video when HDTV devices are deployed. For example: If we even consider the split of observation/forensic review/recognition as 40/40/20 percent of a customer’s applications, then a standards-based 720p HDTV system will deliver images that may be four times the quality and improve 60 percent of a customer’s images.
Schafer: Determine the picture qualtiy or pixels on point that you need in your system. Determine the number of cameras that will cover the location. Note the difference in number of cameras because megapixel will be fewer to cover the same area, perhaps by 50 percent or more. Remember 300,000 pixels vs. 3 million. Figure out the total costs of each of the systems including fewer cameras, housings, mounts, installation costs, lenses and NVR license fees. Articulate the benefits of digital zooming into the recored image. Reduce the number of security guards. We had one school system where that resulted in a $480,000 cost reduction. Factor in the reduction in incidents, including graffiti. In one school system, that came to $200,000-$300,000 per year.
So where do you stand on the HD debate, and more specifically IP megapixel vs. HDcctv? As always, thanks for reading ...
SECURITY SALES & INTEGRATION