Файл: Audio and Radio Engineering.pdf

Video/Audio Networking Systems 8-43

Although Gigabit Ethernet networks can be built using shared media in a broadcast architec-

ture, early implementations are typically full duplex switched (a point-to-point architecture). In a
switched network, the full capacity of the network medium is available at each device. For such a
configuration, the two key switch specifications are the aggregate capacity, which determines
the total amount of throughput for the switch, and the number of frames/s that can be handled.
Ethernet has a variable frame rate and so both parameters are important. Under best-case condi-
tions with large frames, the aggregate capacity defines performance; under-worst case conditions
with small frames, the frame throughput is the limiting factor. In practice, these switches are
specified to achieve wire speed on all their ports simultaneously. For example, an eight port
switch will have 8 Gbits/s aggregate capacity. Many Gigabit Ethernet switches offer backwards
compatibility with existing Ethernet installations, allowing Fast Ethernet or even 10 BASE net-
works to be connected.

Table 8.3.2 Fibre Channel General Performance Specifications 

(

After [4].)

Media

Speed

Distance

Electrical Characteristics

Coax/twinax

ECL

1.0625 Gigabits/s

24 Meters

ECL

266 Megabits/s

47 Meters 

Optical Characteristics

9 micrometer single mode fiber

Longwave laser

1.0625 Gigabits/s

10 Kilometers

50 micrometer multi-mode fiber

Shortwave laser

1.0625 Gigabits/s

300 Meters

Shortwave Laser

266 Megabits/s

2 Kilometer

62.5 micrometer multi-mode fiber

Longwave LED

266 Megabits/s

1 Kilometer

Longwave LED

132 Megabits/s

500 Meters 

Note

: In FC-AL configurations, the distance numbers represents the distance between nodes, not the total dis-

tance around the loop.

Note

: In fabric configurations, the distance numbers represent the distance from the fabric to a node, not the dis-

tance between nodes.

Table 8.3.3 Basic Specifications of Ethernet Performance 

(

After [9].)

Parameter

10 Mbits/s 

100 Mbits/s

1000 Mbits/s 

Frame size

Minimum

Maximum

Minimum

Maximum

Minimum

Maximum

Frames/s

14.8 k

812

148 k

8.1 k

1.48 M

81 k

Data rate

5.5 Mbits/s

9.8 Mbits/s

55 Mbits/s

98 Mbits/s

550 Mbits/s

980 Mbits/s

Frame interval

67 

µ

s

1.2 ms

6.7 

µ

s

120 

µ

s

0.7 

µ

s

12 

µ

s

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2004 The McGraw-Hill Companies. All rights reserved.

Any use is subject to the Terms of Use as given at the website.

Video/Audio Networking Systems

8-44 Audio Networking

8.3.5a

Network Bottlenecks

In many cases the bandwidth available from a switched Gigabit network removes the network
itself as a bottleneck [8]. In fact, the bottleneck moves to the devices themselves that must handle
the Gigabit Ethernet data. It can be seen from Table 8.3.3 that even under best-case conditions,
frames arrive only 12 

µs apart. In this interval, the device must determine if the packet is

addressed to it, verify the checksum, and move the data contents of the frame into memory. Mod-
ern network interfaces use dedicated hardware to take the load off the processor, maximizing the
time it has for user applications. The network interface can verify addresses and checksums, only
interrupting the processor with valid frames. It can write data to discontiguous areas of memory,
removing the need for the host to move different parts of messages around in memory. The net-
work interface can also dynamically manage interrupts so that when traffic is high, several
frames will be handled with only a single interrupt, thus rninimizing processor environment
switching. When traffic is low, the processor will be interrupted for a single frame to minimize
latency.

These measures allow high, real data rates over Gigabit Ethernet of up to 960 Mbits/s. For

systems that are CPU bound, increasing the Ethernet frame size can raise throughput by reducing
the processor overhead.

8.3.5b

A Network Solution

Ethernet on its own does not provide complete network solutions; it simply provides the lower
two layers of the Open Systems Interconnect (OSI) model, specifically [8]:

•

Layer 7, Application

•

Layer 6, Presentation

•

Layer 5, Session

•

Layer 4, Transport

•

Layer 3, Network

•

Layer 2, Data Link: logical link—framing and flow control; media access—controls access
to the medium

•

Layer 1, Physical: the cable and/or fiber

The most widely used protocols for the Transport and Network layers are the Transmission

Control Protocol (TCP) and Internet Protocol (IP), more commonly referred to as TCP/IP. These
layers provide for reliable transmission of messages between a given source and destination over
the network. Using TCP/IP on Ethernet is sufficiently common that most network hardware, for
example a network interface card (NIC) or switch, typically has built-in support for key aspects
of layers three and four.

How messages are interfaced to user applications is the function of the higher OSI layers.

Here again, there are many choices depending upon the application. The Network File System
(NFS) is a collection of protocols (developed by Sun Microsystems) with multiplatform support
that presents devices on the network as disk drives. The advantage of this approach is that appli-
cations do not need to be specially coded to take advantage of the network. After a network
device is mounted, network access looks to the application exactly the same as accessing a local

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2004 The McGraw-Hill Companies. All rights reserved.

Any use is subject to the Terms of Use as given at the website.

Video/Audio Networking Systems

Video/Audio Networking Systems 8-45

drive. Familiar techniques such as “drag and drop” can continue to be used, only now working
with media data over a network.

8.3.5c

Quality of Service

A guaranteed quality of service (QoS) transfers allocate network bandwidth in advance and
maintains it for the duration of a session [8]. After set up, such transfers are deterministic in that
the time taken to transfer a given amount of data can be predicted. They can also be wasteful
because the bandwidth is reserved, even if it is not being used, preventing it from being used by
others. In some cases, a QoS transfer to a device can lock out other nodes communicating to the
same device.

For established television practitioners, QoS brings familiarity to the world of data network-

ing, however, its implications for complete system design may not be fully appreciated. One key
issue is that of the capabilities of devices connected to the network, many of which have video
and network interfaces. The allocation of device bandwidth among these different interfaces is
an important consideration. Video transfers must be real-time and so bandwidth must be able to
service them when needed. However, QoS transfers that are also deterministic need guaranteed
device attention. Resolving this conflict is not a trivial matter, and different operational scenarios
require different—sometimes creative—solutions.

8.3.6

References

1.

Piercy, John: “ATM Networked Video: Moving From Leased-Lines to Packetized Trans-
mission,” Proceedings of the Transition to Digital Conference, Intertec Publishing, Over-
land Park, Kan., 1996.

2.

Wu, Tsong-Ho: “Network Switching Concepts,” The Electronics Handbook, Jerry C. Whi-
taker (ed.), CRC Press, Boca Raton, Fla., p. 1513, 1996.

3.

ATSC, “Guide to the Use of the Digital Television Standard,” Advanced Television Sys-
tems Committee, Washington, D.C., Doc. A/54, Oct. 4, 1995.

4.

“Technology Brief—Networking and Storage Strategies,” Omneon Video Networks,
Campbell, Calif., 1999.

5.

“Networking and Internet Broadcasting,” Omneon Video Networks, Campbell, Calif, 1999.

6.

“Networking and Production,” Omneon Video Networks, Campbell, Calif., 1999.

7.

Craig, Donald: “Network Architectures: What does Isochronous Mean?,” IBC Daily News,
IBC, Amsterdam, September 1999. 

8.

Owen, Peter: “Gigabit Ethernet for Broadcast and Beyond,” Proceedings of DTV99, Inter-
tec Publishing, Overland Park, Kan., November 1999.

9.

Gallo and Hancock: Networking Explained, Digital Press, pp. 191–235, 1999.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2004 The McGraw-Hill Companies. All rights reserved.

Any use is subject to the Terms of Use as given at the website.

Video/Audio Networking Systems

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2004 The McGraw-Hill Companies. All rights reserved.

Any use is subject to the Terms of Use as given at the website.

Video/Audio Networking Systems

9-1

Section

9

Audio Recording Systems

The audio tape recorders that first went on the air in the late 1940s were notable in their ability to
store programs for later broadcast with a sound quality not available from disc recording. Never-
theless, the recorders then were extremely simple in comparison with today’s machines. Even in
this age of high technology, the progress that has taken place in audio tape recording is quite
remarkable and may be attributed to two factors. One was the evolution in the field of electron-
ics: transistor technology; integrated circuits and then large-scale integrated circuits—first in
digital and then in linear devices; digital signal processing; and microprocessors. The second
was the combination of foresight and creativity applied by engineers involved in improvements
in the art (and in user operations as well) in anticipating the need for improved or new capabili-
ties, and then bringing them to pass.

Most of the advances made in the field could be classified either as further development or as

innovation of methods and techniques for meeting new requirements. Neither of the categories
predominated over the other in milestones. The modern digital audio workstation and digital
magnetic tape recorder represent the application of highly developed scientific technologies, the
result of many innovations and refinements since the invention of recording by Valdemar
Poulsen in 1898. Today, many technical and business disciplines depend on the audio recorder in
one form or another as an information storage device. The advancements in audio compression,
computer disk storage, recording media, heads, and signal-processing techniques have made it
possible to achieve storage densities that rival or exceed most other information-storage systems.

In This Section:

Chapter 9.1: Audio/Video Server Systems

9-5

Introduction

9-5

Basic Architecture

9-5

Server Design

9-7

Archiving Considerations

9-9

Audio/Video Server Storage Systems

9-10

Basic Drive Technology

9-11

RAID Levels

9-12

 RAID Level 0

9-12

Source: Standard Handbook of Audio and Radio Engineering

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2004 The McGraw-Hill Companies. All rights reserved.

Any use is subject to the Terms of Use as given at the website.