Video Servers: Select the design that best suits your needs

It is not a surprise that many of the terms used in the computer industry are becoming very common in the lexicon of the television engineer With the conversion from analog to digital raises the need for today's video system designers to become familiar with the terminology and architecture of systems. of computers, added to those of traditional analog video. While the computer industry has written about the use of servers, only in the last two years have serious video applications become commonplace. The term server comes from the client-server relationship in a computer network.

The current design or architecture of a video server is dictated by the applications for which it was made and the specific types of mass storage devices employed. Any basic design server can be 'transformed' to maximize its performance in a variety of applications. For example, the speed and bandwidth of the I/O (input/output) required for a PPV application is very different from what is needed for a streaming-to-air server in playback. In all cases, video compression/decoding techniques, data transfer range, capacity and reliability are key elements for the total success with which the server meets the needs of the application.

The most common application for servers in a TV streaming environment is in over-the-air playback and recording. A second level of application usually involves the storage and playback of notes that are to be integrated into news stories or presentations. Because of the cost of storage, few stations rely on servers to save long programs and then play them.

Editing systems also use server technology, and their applications vary greatly. For many editing systems, the disc video recorder (VDR) provides sufficient storage capacity. A mass storage capacity is required from time to time on a Desktop editing system. However, the concept of a central server becomes not only possible, but desirable in many applications.

Some transmitters and some cable applications require the server to provide multiple program outputs for pay per-view or near-video-on-demand ( NVOD) videos. Keep in mind that the performance characteristics for PPV or NVOD are different from those of the graphics or on-air playback system. When it comes to video servers the size is not universal.

Engineers should be aware of the different possibilities in the market to select the most appropriate video server for their needs. Not all engineers have the option of building fully digital facilities from the ground up, so the server must be able to easily integrate with the existing computer network, the automation system, or with the fundamental network of signal distribution.

The first video servers for television transmission were actually the longitudinal scanning quadruplex head machines developed in the early seventies. While the first Ampex (AR) and RCA (TCR) systems were limited in storage capacity and were nightmares due to their mechanical maintenance, it provided a level of automation to the air never before achieved. These systems soon gave way to more sophisticated devices that focused on multiple recording/playback machines and cassette handling robots.

Around the same time that digital tape formats became popular, hard drive technology began to offer interesting proposals for on-air applications.

Then, high-density (HD) storage became common in editing rooms. Although early products offered quick access to any point in a recording, they often suffered from limited storage capacity and quality was always questioned.

Many automation systems used in the air are actually a union between the two types of technologies, with tapes used for long-term mass storage (files) and hard drive devices with quick access to a limited amount of frequently used video information.

The reliability of a video storage system becomes an important point and depends a lot on the individual reliability of the components used to deploy the server.

We normally think of server technology as HD-based technology, but that's not necessarily the case. Optical storage can provide many of the same benefits of HD storage, often with a lower cost per minute of storage.

Optical discs offer high reliability, and few moving parts, but recording media (media) is almost always written only once and read many times (WORM–Write Once,Read Many– discs), which may exclude their use in some applications that require frequent changes to the recorded video.

Digital tape storage systems provide the lowest cost for media, but need the highest level of maintenance to function properly. Robots for handling tapes and/or optical discs are of medium reliability and may also require a lot of maintenance work. All these factors make choosing a server a difficult task.

The breadth or capacity of a storage system and bit reduction techniques are also important factors in a video server. A 10-bit 4:2:2 uncompressed video serial signal requires 270 Mb/s (or nearly 34 Mb/s) of continuous flow of information. This is faster than a single hard drive can do, so a variety of techniques must be used to achieve that speed of information.

A frequently used solution is to distribute or fragment video information files, which is based on the grouping of individual storage systems using multiple internal Small Computer System Interface (SCSI) buses and disks. This technique speeds up the total rate of information on the storage unit.

The use of bit reduction techniques – which constitutes information compression – (MPEG, JPEG and others), can also reduce the need for the desired capacity to achieve video quality and storage capacity.

Some optical and hard drive server systems offer the option of compression algorithms allowing the user to mold system storage requirements to a specific application with the desired level of video quality. However, physical laws still work and you don't get something for nothing. Barter is usually image integrity by system storage capacity.

System control and integration is another important consideration. Many servers are controlled by external appliances via an R-422 or an Ethernet port. These protocols are also used to hook the computer's internal control system to individual storage units, digital video routers, and input/output subsystems.

However, broadcast-quality video requires a much higher digital bandwidth than the 10 Mb/s that Ethernet can give. This usually requires a high-speed information exchange technology, such as SCI, FDDI, or ATM, to move video file information internally between storage subsystems and input/output drives. Although 100 Base T or Ethernet (100 Mbp/s) is already available, it is still not widely used for servers.

A current control system for a server usually involves the use of a reliable multi-purpose operating model, such as UNIX. Input/output systems handle video (and audio) inputs and outputs with respect to the server and usually all compression techniques present on the server are applied at this level.

Depending on the application, the server can interface with an external equipment via serial digital, component analog video, or composite analog. For servers designed to operate in an all-digital studio environment, server input and output can use FDDI, AIM, or some other standard high-information-breadth workstation. Many servers can be configured to provide an input and output type at any given time and then be changed in the future. This means that one can implement an interface. It must be modified to digital WEGO series if needed. As this option is a bit expensive make sure with the providers.

It is important to remember that orderly storage and video program material usually require more than just collection. An effective file management system is essential in the proper and successful use of any server.

Some servers are capable of handling files internally, while others are designed to be a little more than devices guided by an external control or automation system.

Some manufacturers use commercial common or open technology (OMF) platforms to develop their video archiving and organization systems, but many others develop technologies of their own.

In short, while there are many solutions to mass video organized storage problems, each solution should be unique to your application.