Published in January 2004

Demand for Display Wall Controllers Increases
Conclusion
By John Stark

Presentation and display technologies today are more closely interwoven with the IT environment than ever before.

Editor’s Note: This is the second of two parts. Part 1 appeared in November 2003.

     Presentation and display technologies today are more closely interwoven with the Information Technology (IT) environment than ever before. It has never been more important for AV and IT professionals to work together, or for AV systems integrators to appreciate the overall IT context in which their systems will function.
     For many organizations, this close relationship is most visibly on display in a control room. These facilities come in all sizes and are dedicated to a wide range of functions, but they all have one thing in common: a large display, often an entire wall of video and data displays.
     A client may decide early in the design and construction project that he wants a control room with a display wall. All too often, however, the particulars are left until later, and the client, architect and contractor all discover they cannot just buy a display wall off the shelf. Indeed, the process of choosing equipment and infrastructure for control-room displays often turns into a fundamental review of the room’s mission. This review should address a number of basic questions, whose answers will shape decisions about how the room will be designed, equipped and used.
     Every display wall really is a multi-part system. The most visible, of course, is the screen or screens. The real heart of the system, though—and the key to its success—is the control system that drives those screens. The choice of a display wall controller is critically important, and this choice is influenced by a variety of factors.
     In our examination of the demand for display-all controllers here, I will discuss some of these factors, and highlight a few of the elements that must be weighed in answering them.

What Will be Displayed?
     This is the most fundamental question, and the answer will drive the selection of the appropriate display wall controller and other equipment.
• Video sources: Video sources include television signals, usually in color and typically available as analog signals, either composite or S-video (Y/C). The video may be coming from a TV tuner, a VCR or DVD player, a video teleconferencing system, or closed-circuit video cameras.
Sometimes only one video signal is needed, such as for a weather channel. Other applications, notably traffic-management centers, may have hundreds of video signals.
     Several issues arise regarding the way video is to be displayed: How many video signals must be displayed at one time? How many on a single projector? Will the images be upscaled or downscaled and by how much? What update rate is required? (Video usually refreshes in real time.)
• RGB sources: RGB sources are signals that don’t follow television standards. They usually are scanned progressively. That is, the entire image is refreshed in a single pass, rather than the “interlaced” or alternate-line method used in television. RGB color components are also different from composite or S-video. Typical RGB sources are computer outputs, and the resolution can range from VGA (640x480) to UXGA (1600x1200).
     RGB sources can be “captured” and displayed in a window on the display wall. RGB capture is a good way to pre- sent the display of a single computer on the wall for everyone in the room to see. The signal can be upscaled to make it easier to see (e.g., a 1280x 1024 signal can be zoomed to display over several projectors on the wall covering, maybe, 2560x2048 pixels).
RGB is kind of a “universal interface” because virtually any progressive-scan RGB signal can be captured and displayed. Even an application running on a legacy system with no interface to modern computers can still display on the wall. RGB is also useful if the application, for security reasons, has to be run on a private network and can’t communicate with the wall controller.
Many of the issues that apply to video also apply to RGB, along with questions unique to RGB. Does the RGB processor capture all the VESA standard display formats? Are other display formats also required? Can custom formats be handled? How long does it take to lock onto a new signal that is presented? Can signal acquisition be coordinated with an external switcher?
• Local applications: Many of the display wall controllers available today actually are computers in their own right, with fast Intel processors, lots of main memory, disk-drive storage, networking capability and general-purpose operating systems such as Microsoft Windows or Linux. This not only provides a powerful, intelligent platform for implementing a highly flexible, feature-rich wall controller but also means that many applications can be run locally, on the wall controller itself.
Key questions here are, of course, compatibility with the operating system and whether there is enough CPU power and memory/disk storage for the application. Another question that may not be so obvious is whether the application is well-behaved when given a large window. If your application runs fine on a desktop with a window of 800x600 pixels, what happens when you open a big window on the display wall with 2000x1500 pixels?
• Network applications: Network applications should run fine on the display wall controller because most controllers come with network interfaces. The latest ones have Gigabit Ethernet NICs for high-speed communications. Once again, operating system compatibility is a key issue.
Many control rooms operate in a heterogeneous environment where both Windows and Unix/Linux machines may be present. A good way to provide interoperability here is with the X Window System. If the wall controller uses the Linux operating system, then an X Window user interface will be present and X applications from other computer servers will easily display on the wall. However, this provides limited interoperability with Microsoft Windows applications.
On the other hand, if the display wall controller is based on the Windows operating system, there will be good interoperability with other Windows computers on the network. Furthermore, if the wall controller includes an X Window user interface, you will also have compatibility with all the Linux, Unix, Solaris, VMS and other applications implemented to the X Window standard.
• Network (streaming) video: Streaming video, ubiquitous these days, is used frequently on desktop PCs and is beginning to play a major role in control rooms. This is a good example of an application that has problems scaling to display wall size. Streaming video typically is low resolution, 320x240 pixels or smaller. Although this can be acceptable on a desktop monitor when you are viewing it from 18 inches away, it will not be large enough on a display wall when viewed from 10 or 20 feet. That means the streaming video must be upscaled to a larger window size.
     Decoding compressed video is a computer-intensive task that today’s powerful processors handle with ease. But if you further burden the processor with the task of upscaling the decoded video to a large window size, you won’t achieve real-time operation.
     Such upscaling is beyond the reach of most general-purpose processors and is best handled by an external device that decodes streaming video to a standard analog video signal.
• Network RGB capture: Proprietary programs that allow you to operate one Windows PC from a remote Windows PC via the network have been around for many years. These programs work well on a one-to-one basis with small displays, although they are slow and burden both the network and the processors of both PCs. They have been used in display wall environments with limited success.
     In a display wall environment, one big problem involves the burden on the CPU if the display is upscaled to a large window size. Viable solutions based on a tool called Virtual Network Computing use open software to allow cross-platform operation between Windows, Linux and other operating systems. For the control-room environment, this not only provides convenient display of network-connected computers, but also a network-based method of interacting with the computers themselves. Unfortunately, this position does not provide ubiquitous access to all systems, because the technology cannot display in real time or support the 3D and video standards available.
     All of these critical questions about data types and sources are addressed generally by modern display wall controllers, although their individual capabilities vary.

Legacy systems
     Many control rooms must draw data from older legacy systems as well as from more recently implemented networks and applications. The most common pairing is Windows and UNIX, particularly in organizations that have moved from a UNIX-based workstation environment to a more economical Windows PC strategy.
     If the installation is predominantly UNIX based, it’s best to craft a solution that provides native UNIX tools and interfaces. Windows applications can be integrated into the X Window environment using a simple and free tool called a Visual Network Computer (VNC). VNC, a client/server application that sends graphical screen content over a network connection, has some drawbacks in performance and scalability, but may be appropriate for many installations.
In a Windows setting, it’s important to choose a Windows-centric solution and augment it with UNIX tools. X Window content can be displayed in a Windows environment by using a Windows-based X server.
     Blending these distinct environments is a crucial job, and one that an experienced systems integrator is ideally qualified to execute.

From Desktop to the Big Wall
     Few Windows-based PCs support screen resolutions of more than 1600x1200 pixels, with most operat ing at a level below that, typically 1280x1024. In contrast, a 4x2 array of 1600x1200 pixel screens in a display wall offers usable “pixel real estate” of 6400x2400. Many display walls go even larger, sometimes up to 10,000 pixels of horizontal resolution.
Any solution that simply enlarges the desktop display will actually lose clarity and readability as it gets larger rather than benefiting from the additional resolution. Moreover, when a display designed for the desktop is transported to such a large canvas, it may exhibit artifacts, become unstable or even crash the computer system—hardly acceptable in an environment requiring 24/7 availability.
     The real benefit of increasing screen size and resolution is that it becomes possible to put more detail on the screen. Rather than the same display eight times larger, the user can see eight different displays simultaneously, or some other combination of windows and data flows.
It is also important to understand the tradeoffs in moving to a very large display. While a single monitor at 1280x1024 requires about 2.6 megabytes of data for its display, an eight- or 10-screen wall requires eight or 10 times that much. Handling all of this data affects drawing speeds, the ease with which windows can be moved or resized and other characteristics.

How Interactive?
     One of the ancillary but important missions for many control-room installations today is marketing. Telecoms and other companies often use their Network Operations Center (NOC) to demonstrate the scope of their resources and the sophistication of their command and control operations. An impressive appearance is a prime consideration in these installations.
At the opposite extreme are facilities such as those operated by electric-power companies, where displays provide real-time data to support critical decisions.
     In these settings, the display wall must be a completely interactive tool, pulling data quickly from dozens or hundreds of sources, displaying it in meaningful ways and enabling managers to make fast, sound decisions—whose results are also displayed quickly and usefully on the wall.

Security Issues
     Control rooms present obvious security concerns because they so often display critical information that an organization wants to keep to itself. A large display wall magnifies these concerns through its simple visibility. Thus, the first line of defense for the client is to control access to the room itself. Who is allowed to be there?
     A second, related issue, is that of identifying and tracing changes. A display wall is a powerful “anonymizer,” i.e., when a change appears on the screen, it can be difficult to tell immediately who made the change, and at which workstation. Part of the solution, clearly, lies with the organ- ization’s system of access and editing permissions.
     Another option is to make the display wall itself read-only. Thus, all changes to data must be done on the appropriate workstations by the authorized individuals, and receive the correct management review and approval before being displayed. It is also possible to allocate certain portions of the wall to specific operators.
Although display technology provides some flexible options, these critical decisions must be driven by organizational policy.

Ergonomics
     The critical questions in planning a control room do not all focus on technology. The people who will use the room are not mere generators and consumers of data: They are people.
Will they be able to read a certain size type from their desk 20 feet away? If you make the type larger, will people up close be able to read it? Will someone in a corner of the room, viewing the display wall at an angle of 50 or 60 degrees, be able to see everything clearly? Does your choice of color enhance readability or detract from it?
     Will users interface with the displays using such familiar tools such as a keyboard and mouse? Will they see their inputs displayed immediately?
     Some of the factors that help determine the success of a control room, such as color schemes for the furniture, carpets and walls, are esthetic, and it may be tempting to call them matters of individual taste. Yet it is possible to quantify the impact of room esthetics on functionality and productivity, and an experienced systems integrator can provide valuable advice in this area.

A Sound Approach
     Much of the success or failure of a project takes shape during the development of the Request for Proposals. All too often, the first indication that an RFP is flawed comes at a bidders’ information meeting, when questions outnumber answers and the project appears beset by ambiguities and knowledge gaps. Contractors may decline to bid on poorly defined projects, and clients often realize how much additional fundamental analysis they need to do.
     This disaster can be largely avoided if the RFP writing process includes someone who thoroughly understands the entire control-room environment, and can give adequate attention to all the types of concerns we’ve outlined here. As a result, when you do issue your RFP, it will be comprehensive, detailed and realistic.
     Clients often visit operational control rooms suggested by integrators. On such a visit, the issues we’ve outlined should be a checklist. Find out what kinds of data and data sources the room accommodates, and try to see a variety of displays: surveillance video, broadcast and cable TV, graphics, spreadsheets, videoconferences and so on.
     Try to sit at workstations at varying angles and distances from the screen and judge text readability for yourself. Spend enough time to get a feel for what it is really like to work in this room. Then apply your evaluation to your own project.

The Correct Controller
Four basic types of display controllers are available today:
• Multiple-screen PC
• Multiple-screen PC with specialized control room software
• Multiple-screen, RGB-input-only processor
• Built-to-purpose display wall controller with RGB and/or video inputs and control-room software.
     The simplest of these options is an off-the-shelf PC with a multiple-screen graphics controller or multiple graphics cards. Often these systems are adequate for small installations, but quickly lose relevance when needs become more complex. Overlooked requirements in the control room include longevity and vendor support. Systems installed in control rooms must remain operational, often for some seven to 10 years. However, this is well beyond the lifespan of most PC technologies and the interest or capability of most vendors catering to the off-the-shelf market. As a result, customers are left with systems that are not supported, who have nowhere to turn except to discard the old system for a new one.
     By adding a layer of control-room software, users can enable their PCs to bring in applications and displays over a network. This approach can also meet many needs, but lacks the speed and performance necessary for really critical tasks. Imagine a combat commander waiting for his screen to refresh before finding out whether an incoming aircraft is friend or foe! Ad- ditionally, this solution still suffers from a lack of long-term vendor support, a risk many users are unwilling to take.
     The RGB-input-only processor typically is a standalone device that displays inputs from video, TV and computers. It doesn’t provide the ability to process local applications requiring input from such devices as workstations and PCs and, depending on the number of RGB sources that must be displayed, plus the up or down scaling required, it may be unable to respond in real-time. Apart from being costly for large complex installations, these systems do not provide the real-time input and manageability that a PC infrastructure allows. Instead, users must interact with custom software that often is designed for the entertainment industry, rather than a control room.
The fourth option—the “augmented PC” controller—provides the greatest range of capabilities. Essentially, these are purpose-built to meet the client’s specific needs, and they offer the greatest performance, reliability, ease-of-use and maintenance over the broadest range of requirements. They can pull data from the greatest number of sources and display them in movable, scalable windows in near real-time.
     The best of these processors offer 24/7 operation with redundant components and hot-swappable parts. They are self-monitoring, and alert users to conditions that may require intervention. Indeed, these display wall controllers integrate the capabilities of the other categories while carefully addressing the specific requirements of the control room. These high performance controllers provide flexibility, ease-of-use, high-quality display, and long-term maintenance and support. Moreover, these controllers are available from a number of reputable manufacturers, giving integrators and clients a great deal of choice.

Selecting the Best Controller
Today’s AV and presentation technologies, coupled with increasingly powerful and flexible IT tools, offer an unprecedented array of options in designing a control room. The facility truly can be and do whatever the client wants. Today’s display wall controllers are the key to these capabilities, and selection of the best one is critical to the most effective implementation.
An astute integrator will be attuned to the client’s operational needs, budget or physical space constraints. Often these decisions will differentiate the exceptional systems integrator and his or her value added.