Professional AoIP for Broadcast - The Way Forward Introduction

Professional AoIP for Broadcast - The Way Forward Introduction
Professional AoIP for Broadcast The Way Forward
Tom Knowles, Product Manager - Broadcast Systems
Budgeting for an IP future
Today’s standard industrial IT infrastructure has already
overtaken the technology of AES/EBU, MADI and TDM routers in terms of performance, cost and flexibility. The rate of
development of IT systems, fuelled as it is by a multi-billion dollar industry many times the size of the broadcast
industry, is certain to widen this gap in the future. Over the
next few years IT infrastructure will replace current broadcast infrastructure, delivering additional flexibility, better
scalability and significantly lower costs. Broadcasters and
Systems Integrators can expect more choice in selecting
interoperable equipment and solutions from a range of
suppliers and will have more scope to manage and control
these systems.
It’s rare that a facility has the luxury and budget for a
ground up development and it’s likely that the thought of
moving to an IP future could be seen as a daunting proposition. The good news is the transition does not need to
be an all or nothing leap. In fact, SSL believes that IP and
TDM solutions are likely to coexist for some time to come.
For this reason, one of SSL’s first developments for our IP
systems product line is a ‘broadcast robust’ MADI to IP
bridge product.
Audio over IP (AoIP) has been embraced by the installed
sound, live and radio industries. Much has been learnt
from this but larger broadcast applications bring additional
constraints, especially channel count, synchronisation and
latency. Audio systems engineers will need to learn how to
set up, configure and manage IT networks; IT specialists
will have to understand why and when audio networks
need to be separate. Education will be key.
As with previous transitions, audio, with its lower data
rates, will be a pathfinder for developments that will follow
in the video domain. IT is already widely used in broadcast
for file transfer, but SMPTE 2022-6 (high data rate streamed
video over IP), which was published in 2012, is evidence
that SDI will eventually be replaced. Central to considerations about audio or video over IP is the value that
metadata brings to systems. Concepts such as discoverability and automatic configuration are key to delivering
powerful workflow benefits.
In this paper we describe our vision for the future of Audio
over IP for professional broadcast applications, explain
which technologies we are using, discuss the advantages
and challenges associated with AoIP and demystify the
As for installing a whole new physical infrastructure, the
likelihood is that the IT department may have done much
of the work already. A significant benefit of a good IP based
solution is that it uses ‘standard’ cabling and hardware to
provide the complete solution. This helps make
infrastructure planning straightforward, predictable and
easily costed. The required technology has a wide range
of price points and good quality solutions are available at
much lower cost than custom broadcast routing hardware.
Twisted pair copper and fibre, managed switches and
simple punch down termination all reduce overall cost and
make expansion straightforward as well as cost-effective.
It is useful to remember that the base wiring and switching
infrastructure for all IP based systems is the same. A key
benefit of standard industrial IT network based
infrastructure is that it is format agnostic and multiple
types of signal can coexist easily on the same hardware
and cabling so network infrastructure can be used to implement device control, as well as audio transport, further
reducing wiring and installation costs.
The benefits of an IP solution.
There are several significant benefits that are key reasons
to implement a network based routing solution:
Reduced Cost
As discussed, simple economics make an IP solution an
attractive proposition for audio routing.
High Channel Counts
Another significant benefit of a networked infrastructure is
capacity, and particularly capacity vs cost. A single gigabit
network connection is capable of carrying 512 channels at
48kHz, or 256 channels at 96kHz. AoIP capacity compares
very favourably when considering using network based
devices as your core audio router. With 512 audio channels
in each direction on a 1GB connection, a single 24 Port GB
Switch is capable of providing the equivalent audio routing
capacity to a 12,288 by 12,288 audio router. It should be
noted that this comparison is heavily based on a traditional
broadcast concept of a central core routing device. Depending on the physical building, existing network infrastructure
and system requirements, alternate network topologies
may be more advantageous and cost effective.
Easily Expandable
When the needs of the installation expand, the system
can grow much less expensively than a traditional TDM
solution. Adding a network switch is a much easier way to
expand capacity than having to replace a router with a
larger one. AoIP Network routing is not subject to the
square law growth of TDM routers i.e. doubling the size of
a TDM network typically quadruples the size (and possibly
cost) of the router.
Full Network Redundancy
Redundant operation is at the core of security in both the
broadcast and IT industries, therefore designing a system
to be secure shouldn’t be a concern. There are often
several redundancy strategies in network designs to create
fully resilient networks, these include Mesh networks,
Spanning Tree, Token Ring and Link Aggregation. These
designs are supported by many enterprise class network
switches and an AoIP network solution which uses
standard networking protocols should support any or all of
these solutions. As an example, SSL’s IP solution provides
inbuilt parallel redundancy allowing both main and
redundant networks to be always active, a device
simultaneously receives and transmits on both networks,
audio samples can exist on either or both networks
providing seamless, glitch free redundancy and failover.
Resilient Distributed Routing Control
As the signal routing in an IP network is end point based, a
catastrophic failure of a single device will have no effect on
any other devices in the network. With redundant routing
and/or connections, single points of failure are easily designed out of the system. In addition, the use of the
network for control information means routing control can
be achieved from an unlimited number of devices or
terminals on the network.
A History of Broadcast audio Routing
The earliest forms of signal routing in radio studios used mechanical patchbays inherited directly from the telephone
industry. With later evolution and more complex needs, telephone patchbays were augmented by ‘Uniselectors’,
electro-mechanical signal routers remotely controlled by electrical pulses. Companies like the BBC used these to
control routing within their broadcast stations, allowing flexible and efficient use of studios, equipment and people to
produce complex news and sports productions... Click here to read more
Professional AoIP for Broadcast The Way Forward
Tom Knowles, Product Manager - Broadcast Systems
These control benefits extend beyond routing control to
•Device parameters that can be configured from anywhere
on the network without additional cabling
by Audinate’s Dante technology. The Dante Protocol brings
some specific benefits that arise from their technology:
•Any other software control application can use the same
network infrastructure
A significant benefit of the Dante IP solution is the ability
of all Dante devices to support Auto Discovery and ‘Plug &
Play’ operation. This means that devices which are plugged
in to the Dante network announce their presence and can
be ‘found’ by other Dante devices on the network. This
makes it easy to move equipment from place to place and
extends to Dante products from any manufacturer.
•A single cable infrastructure can be used for configuration, control and audio e.g. Ember+, ProBel, SNMP, UI etc.
Looking to the Future - IP is already there!
With discussion about the future of broadcast audio
featuring Metadata and Object audio Transport, a network
infrastructure is already equipped to manage the technical
requirements for these complex applications. If we
consider ‘broadcasts’ where multichannel audio and
metadata are transmitted to the consumer to provide
personalised listening experiences, in future it is likely that
these ‘broadcasts’ will happen over IP to the consumer,
rather than over the air. The metadata requirements in a
production environment mean network based technology
is essential.
Selecting an IP Protocol
Many of the AoIP benefits discussed so far come with the
structure and technology inherent in using mature
solutions developed in the IT industry. For SSL a significant decision in the project to develop IP based solutions
was selecting which IP protocol to adopt. All options were
considered but in many ways, the decision was made easy
by the well-established and complete solution presented
The Evolution of audio over IP
In the early days of networking, research was driven by
the desire to overcome the inherent waste in a time
division multiplexed (TDM) connection. By turning signals
into packets and sending multiples of these along the..
Click here to read more
Auto Discovery and Plug & Play Device Connection
AES 67 Compatibility
Dante is compatible with the AES 67 standard, a “standard
for audio over IP interoperability” agreed in September
2013. AES67-2013 provides comprehensive interoperability
recommendations in the areas of synchronization, media
clock identification and network transport. It specifically
addresses high performance media networks that support
professional-quality audio (i.e., 16-bit, 44.1-kHz and
higher) with low latency (less than 10 milliseconds), and
at a level of network performance that can scale from local
area networks (LANs) to enterprise-level networks. This
means that Dante products will be able to exchange audio
over IP with other AES67 compliant products.
Guaranteed Interoperability
With Dante, devices from all manufacturers are guaranteed
to work together with a level of interoperability that goes
much further than the audio stream compatibility of AES67
compliant products. As proven at interoperability demonstration events, multiple broadcast products all appear
on the same routing matrix automatically. More than 100
manufacturers are already shipping over 350 different
Dante devices, with over 200 manufactures having
licensed the technology. With the majority of these devices
developed primarily for applications in the professional
public address and live sound markets, Dante also facilitates the specification of systems that encompass the
increasing demand for multi-purpose installations.
Standard IT Network Infrastructure
Dante is able to use existing, off the shelf network
switches, unlike AVB, which requires specialised,
expensive, less readily available switch hardware. In
addition, Dante uses established IEEE and IETF standards
so its network data can mix with traffic on any standard IP
With Dante being built on standard networking protocols,
moving beyond networks managed by your organisation
is possible. MPLS (Multiprotocol Label Switching) is often
used when high performance connections are needed
across an external telecommunication provider’s networks. Deploying AoIP on this type of leased network is
achievable, though as with any use of external services
the SLA (Service Level Agreement) is key to achieving the
performance required. A guaranteed bandwidth greater
than the required channel count (100MB for 64 channels)
and latency of <5ms stated in each direction is required to
use standard Dante settings. While uncompressed audio
across MPLS and other virtual network connections may
be new ground for the broadcast market, this is not new
ground for AoIP or Audinate. Examples such as the Sydney
Trains System’s networked public address system uses
Dante and MPLS extensively.
Rapid Cohesive Development
Although strict ‘standards’ based’ technology has its
merits, history demonstrates that it can take a long time to
be fully interoperable. Examples of licensed technologies,
such as products produced by Dolby Laboratories are
evidence that proprietary solutions can foster rapid
cohesive development. It means that manufacturers can
develop a wide range of niche products with a high
degree of confidence that they will integrate smoothly with
products from other manufacturers. The result for System
Engineers is a versatile, cost effective broadcast
technology ecosystem that contains all of the elements
they require.
DiffServ QoS (Quality of Service)
DiffServ specifies the mechanism for classifying and
managing network traffic and providing quality of service
(QoS) on IP networks. QoS must be enabled to allow Dante
to share network infrastructure with other types of data and
signals. In many cases QoS in installed network switches
may already be enabled, if not the network switches need
to have the Basic mode of QoS enabled, checking that
the switch is using DSCP (Differentiated Services Code
Point). Education and use of common language between
broadcast and networking engineers is key in successfully
leveraging the advantages in AoIP technology. The network
specialist may want to know more about the DSCP labels
Dante uses. Audinate publish the DSCP priority values for
Dante here.
Switching Capacity
To achieve high channel counts Ethernet switches must
meet specification requirements. In the same way TDM
routers are described as non-blocking, network switches
can also be described as non-blocking. By design and to
reduce cost some switches are not non-blocking, these are
often used as edge switches where full bandwidth
switching may not be required. The important figure is
‘switching capacity’ in Gbps, this is the max bandwidth
of switch backplane. For a switch to be non-blocking the
switching capacity needs to be greater than or equal to
the number of ports, times the speed of each port, times
2 (so in and out for each port is considered). For example:
a 24port 1GB switch would require a switching capacity of
Synchronisation and Latency
Dante uses the IEEE 1588 Precision Time Protocol (PTP), enabling all devices to synchronise their local Audio clock to
a master clock on the network with sample accurate alignment. No additional clock distribution is needed saving on
cabling and distribution amplifiers. Multiple sample rates
can coexist on the same network but external SRCs would
be needed to connect these signals.
Audio latency is deterministic and is defined on a per device basis at the receiving unit. The latency of connections
does not change when you add or remove devices from the
network. Device latency should be selected based on the
application and network size. Recommendations are from
125μs for networks with 3 switch hops, 1ms with 10 switch
hops, to 5ms for larger or leased networks.
Professional AoIP for Broadcast The Way Forward
Tom Knowles, Product Manager - Broadcast Systems
Device Control
Routing Control and Security
One of the most active topics of discussion in the various
standards bodies and around broadcast conventions is the
subject of control. For many applications the advantages of
AoIP seem to be tied with the holy grail of a single standard
control protocol for everything, with ultimate flexibility.
While this goal or aspiration for a common control protocol
across many manufacturers may ultimately have advantages, this thinking may miss certain aspects of the
advantages of a network; the infrastructure is agnostic to
the signal, control data and metadata. Some control
scenarios lend themselves to device to device control,
while some are more suited to central (or perhaps distributed) control and monitoring systems. As with audio
transport, basing these control protocols on standard and
widely adopted underlying networking protocols is the key.
Integrated control is a significant benefit of a networked
solution and the Dante solution to this is well defined.
Audinate provide a simple X/Y routing interface called
Dante Controller which can be run on any suitable
computer connected to the network. The software provides
an elegant approach to basic routing. It is important to
understand that routing and configuration for Dante
devices is end point stored and routes are based on device
and channel names. A receiving device subscribes to
signals from other devices that transmit. If a device is
moved to an alternate location on the network audio
connections are re-established automatically and the
channel names appear as they were before. This can be
used to great advantage when moving devices around a
building, venue or anywhere with a network connection.
This also means that there is no need for Dante Controller
or another controller based on the DAPI to be running for
audio to pass, controller software is required only to make
Audinate’s DAPI (Dante Application Programming
Interface) includes its own control and monitoring protocol,
ConMon. This allows software to be developed that will be
able to perform routes, plus configure Dante and network
settings on any manufacturers’ Dante enabled device, with
true interoperability. In addition to the Audinate defined
parameters, ConMon includes the ability to add vendor
specific messages into the same transport layer. By
definition all Dante devices include ConMon. For device
to device control, e.g. Mixing console to Mic Pre, this is an
elegant path to control interoperability beyond what
Audinate have already defined.
As with traditional broadcast audio routing, control
systems will play a key part in AoIP migration or ground
up broadcast IP installations. Routing arbitration is one
area that remains the subject of ongoing discussion. If we
take arbitration to be the locking of a destination when a
route is made so this target route cannot be unintentionally
overwritten, then we need to question where to arbitrate
this. Within an AoIP system where routing is stored on the
end points, this question is similar in scope to arbitration
across multiple hardware TDM routers. Previous technology
experiences have shown arbitration from a control system
considered the master is often the safest way to achieve
this without conflict. This is achievable today with third
party control systems implementing DAPI. In this scenario
security of the network becomes important so that only
clients of the main control system would alter routes. There
are many approaches to achieve this such as;
authentication based access controls (e.g. 802.1X), ACL
(Access Control Lists), segmenting switches with VLANs
or even mirroring what many people do with traditional
broadcast technology in preventing access to the network
via locked rooms, locked cabinets and rights managed
To facilitate using Dante to augment existing MADI based
infrastructure SSL supports MADI control data tunnelling
with its Network I/O: MADI-Bridge.
Control Development
As discussed Control is an active topic of debate, SSL are
actively working with other manufacturers to further
develop control definitions applicable to the Broadcast
industry, with a vision not only of audio interoperability
between manufacturers’ solutions, but also control.
AoIP in action - some examples.
While many Broadcast AoIP discussions are about future
large scale infrastructure projects there are many scenarios that can benefit from AoIP deployment today. SSL and
other manufacturers’ Dante implementations are allowing
these use cases to be realised with a guarantee of future
IO everywhere
In many scenarios audio inputs and outputs are needed at
many locations across a large physical area. Fig 2. Shows a
typical Golf Setup, with connection nodes at each hole.
Network connections by their nature allow efficient resource connectivity regardless of channel count. IO devices
can be located where needed without running multiple
analogue cables to a 64 channel capacity location as you
would with MADI. The advantages when broadcasting from
an event spread over a large area are huge. Again with a
protocol such as Dante these connections can be used for
more than just the audio.
Remote Production
It is possible to use a network connection to a remote
location where a previously laid fibre TDM connection is in
place (e.g. Fibre MADI). A fibre cable previously used for
a point to point 64 channel MADI (125 Mbps) connection
could very easily be repurposed to carry up to 512 audio
channels with MADI bridges and network switches. Alternatively the connection could simultaneously be used for
other network connectivity alongside the audio.
Professional AoIP for Broadcast The Way Forward
Tom Knowles, Product Manager - Broadcast Systems
A network switch as an audio router
Both of the above examples are scenarios that could use AoIP to enhance existing equipment and infrastructure. For a ground up
installation any existing TDM based system can be replicated and improved upon. A deceptively simple design ethos applies: Ring,
Star or Mesh network configurations can be used for facility wide network design. Placement of network switches locally within control rooms, galleries, studio floors, remote locations etc enables connection of any combination of audio and control
devices at each location.
Ring configuration
In a ring configuration data can flow
either way around the ring therefore a
broken cable does not disconnect any
device. A ring configuration can rely on
spanning tree to prevent packets from
indefinitely passing around the loop.
Mesh configuration
A mesh configuraton is extremely
robust. There are no central switches.
All points are connected to all other
points. The managed switches are
configured so that packets do not flood
the network.
Star configuration
A traditional TDM style star configuration with a central pair of redundant
switches distributing to each individual location. Any cable breaks are
resolved by switching to the parrallel network.
Professional AoIP for Broadcast The Way Forward
Tom Knowles, Product Manager - Broadcast Systems
A History of Broadcast audio Routing
The Evolution of audio over IP
In the early days of networking, research was driven by the desire to overcome the inherent waste in a Time Division
Multiplexed (TDM) connection. By turning signals into packets and sending multiple packets along the same cable, the
capacity of that cable quickly rose to thousands of channels. In addition, those data packets could travel with addresses,
meaning they could traverse connections without needing the connection to be explicitly established.
The earliest forms of signal routing in radio studios used mechanical patchbays inherited directly from the telephone industry.
With later evolution and more complex needs, telephone patchbays were augmented by ‘Uniselectors’ - electro-mechanical
signal routers remotely controlled by electrical pulses. Companies like the BBC used these to control routing within their
broadcast stations, allowing flexible and efficient use of studios, equipment and people to produce complex news and sports
Complexity and Density
As the handful of signals that needed to be managed for a production gradually became more complex, audio production
consoles needed to handle an ever expanding number of sources and outputs and the signal routing requirements increased
exponentially. This led to the increasing density of local signal routing patchbays to work in conjunction with the facility
routing, expanding the number of wires and the complexity of installation. At the same time, the use of FET electronic
switching to replace costly relays increased the ability to route many signals with a single button press. This all increased the
complexity and cost of typical broadcast audio installations.
Digital Signals - from Analogue to TDM
With the development of digital protocols and processing in the late ‘70s/early ‘80s, audio broadcast signal routing quickly
became the realm of Time Division Multiplexed (TDM) digital signals. A single cable could carry more than one channel and
the broadcast industry adopted firstly stereo digital audio (AES3) and then later multi-channel digital audio (MADI) to
transport signals. With these digital signals came larger and larger capacity digital audio routers. In this (and later!)
technological routing developments, audio preceded video in the digital transition. TDM connections are inherently
inefficient, interconnections are often carrying no useful information. Many broadcast infrastructures today use large
scale TDM routing solutions and both AES and MADI digital audio to interconnect facilities.
ATM - From Cables to Packets
In the late 90’s, the first non-synchronous, network broadcast audio transport solutions were explored using ATM
(Asynchronous Transfer Mode). ATM is a network technology based on transferring data in cells or packets of a fixed size
developed by the telecoms industry as a way to improve on the inefficiencies of TDM. By using small, constant packet sizes
ATM equipment could transmit video, audio, and computer data over the same network. This early network transport solution
was explored by several companies and broadcasters, including the BBC. Although in many ways suited to media data, the
cost and availability of ATM technology was significantly impacted by the increasing use of Internet Protocol (IP). The crucial
difference between ATM and IP is that ATM is connection-oriented while IP is connectionless. This means that the establishment of a connection between two endpoints in ATM defines the route all packets (cells) related to that connection must
travel. In principle, IP is connectionless, so each IP packet carries a full destination address so there is no concept of a
connection at the IP level. Fundamentally, support for ATM packets and connections have largely disappeared from most
network switches, while IP has become a de-facto standard.
Click here to return to Page 1
The first AoIP solutions used existing IT networks to deliver a limited number of channels through a standard 10 or 100Mbit
network infrastructure. Other solutions used standards that were limited to lower OSI layers in order to connect with
greater efficiency, but at the price of needing specific network hardware and being unable to be mixed with other Ethernet or
IP network traffic. With the use of network technologies to build more ‘real time’ applications, networking standards evolved
to add prioritisation of traffic so that interactive audio applications were possible. Voice over IP (VoIP) requirements for IP
based telephone systems introduced many of the AoIP standards that have been adopted and developed.
AES 67
AES 67 is a standard aiming for interoperability of audio transport between multiple existing and future real time AoIP
protocols. It is intended to further the possibilities for sending audio between different manufacturers’ devices. However AES
67 does not define details of device discovery and configuration, manufacturers will need to agree to define these in the same
way to achieve a fully integrated interoperable solution. To interface with AES 67 connections, Dante systems will be able to
produce AES 67 streams so that audio can be interchanged with devices that support AES 67.
Audio Video Bridging (AVB) is a common name for the set of technical standards developed by the IEEE Audio Video Bridging
Task Group of the IEEE 802.1 standards committee. This is different to AES 67 in that AVB introduces a change to Ethernet to
redefine how a network deals with real time audio traffic. It aims to slightly reduce latency and puts a reservation on audio
bandwidth so the network knows when audio connection cannot be made, this potentially makes the network design simpler.
AVB is a change to Ethernet and needs new switches for it to work. A limited number of switches exist, none of these are from
the market leaders in the manufacture of network switches (Cisco & HP) and at the time of writing, AVB is layer 2 only.
Audinate’s Dante
Audinate were one of the first companies to build a high capacity, low latency AoIP solution built on the IEEE and IETF standards. Audinate provide OEM hardware, firmware images and a full software API to allow manufacturers under licence to
produce various different products with guaranteed interoperability. They continue to evolve their solution to deliver increasing benefits of capacity and speed, while retaining compatibility with standard networking components and interoperability
between components. This reduces issues with installation and increases choice for the installer. A key benefit to the Dante
solution is the integration of audio transport, device discovery and device control which is key to the provision of the most
interoperable system possible.
Click here to return to Page 2
Professional AoIP for Broadcast The Way Forward
Tom Knowles, Product Manager - Broadcast Systems
Useful Glossary of terms
Network layers and the OSI
The Open Systems Interconnect model is a standardised
model which defines the functions of a communication system by partitioning it into seven abstraction layers. It was
published in 1984 in ISO 7498 and also in parallel by the
ITU in the X200 standard. If a network device e.g. a switch
is “Layer n managed”, it can be assumed that it is capable
of performing functions that relate to the given OSI layer
as shown below. It may also be assumed that if a switch
is layer 3 managed, it is also capable of managing layer 2
Layer 1 – Physical Layer
• Defines the electrical and physical specifications of the
data connection. It defines the relationship between a
device and a physical transmission medium (e.g., a copper
or fibre optic cable).
Layer 2 – Data Link Layer
• The data link layer provides a reliable link between two
directly connected nodes. It detects and possibly corrects
errors that may occur in the physical layer. VLANs are layer
2 constructs.
Layer 3 – Network Layer
• The network layer provides the means of transferring
datagrams. Datagram delivery at the network layer is not
guaranteed to be reliable. A number of layer-management
protocols belong to the network layer. These include routing protocols and multicast group management.
Layer 4 – Transport Layer
• The transport layer provides reliable transmission of data
packets between nodes (with addresses) located on a
Layer 5 - Session Layer
• Establishes, manages and terminates the connections
between the local and remote application.
Layer 6 - Presentation Layer
• Transforms data into the form that the application
Layer 7 – Application Layer
is possible to have multiple subnets on one VLAN but not
multiple VLANs on one subnet.
Classification of traffic
Subnets divide a large network into n smaller networks
for performance and security reasons. subnetting involves
the separation of the network and subnet portion of an IP
address from the host identifier. Devices on one subnet
cannot communicate with devices on a different subnet
without passing through a gateway.
MAC address
• Interacts directly with software applications running on a
host e.g. determining identity and availability.
Unicast packets are sent from one source to one
destination, identified by its IP address. In unicast
communications, every copy of a signal – even identical
signals – is a point to point connection with its own
bandwidth requirement.
Multicast packets are sent from one source to many
destinations on a network subnet. Multicast communication allows a saving in network bandwidth to be made
when one source is being sent to many destinations. Using
multicast communication, a sending device requires only
one unit of bandwidth per discrete signal sent. E.g. a
one-to-many distribution system. Replication and
distribution of a multicast signal is performed by network
switches or routers. Devices may subscribe or unsubscribe
from receiving multicast traffic.
Broadcast packets are automatically forwarded to all
devices on the subnet to which they are connected.
Broadcast traffic is usually reserved for discovery and DHCP
services which, by their nature, must be able to contact
every device on a network segment without initially
knowing their address.
In computer networking, a single network may be
partitioned to create multiple distinct domains, which are
mutually isolated so that packets can only pass between
them via one or more routers. These domains are referred
to as Virtual Local Area Networks, or VLANs. VLANs are layer
2 constructs, compared with subnets, which are layer 3. It
A Mac address is a globally unique hardware
identifier assigned usually by the hardware vendor and
stored in ROM. It is notated as a six digit Hex number, e.g.
Network hardware
Ethernet hub:
A multi-port repeater operating on OSI Layer 1. A hub
forwards any signal received to all its ports except the
originating port. It contains no memory or routing logic but
may send jam signals if a collision (more than one device
transmitting at once) is detected.
Ethernet switch:
A ‘smart’ repeater. Buffers, processes and forwards packets
only to devices who request or require them at a link speed
which has been predefined or negotiated. A Switch may
be managed and perform security or traffic management
A gateway is a type of router which allows data to be sent
from one subnet to a different subnet. If a device sends
data to a device outside its own subnet, it will be
automatically routed to the gateway (assuming IP settings
on the originating device are set correctly).
A router may also direct traffic between networks, but is
responsible for finding the best physical route for traffic to
take to reach its destination. Routers prioritise traffic based
upon type, network load and policy.
A self-contained, independent packet of data containing
enough information for it to be routed from the source to
the destination device without reliance on earlier interaction between the source and destination device and the
transporting network. The source device need not establish
a direct connection with the destination device, and can
send each data packet through any available route.
Useful Protocols
UDP - (User Datagram Protocol)
UDP is a simple, message-based, connectionless protocol
which does not set up a dedicated end-to-end
connection. Communication is achieved by transmitting
information in one direction from source to destination
without verifying the readiness or state of the receiver. The
primary benefit of UDP over TCP is the application to audio
over IP (AoIP) where latency and jitter are the primary concerns. UDP is usually a fast, efficient and reliable protocol
within a wholly wired local area network, with enough
network capacity, it needs careful management for bridged
networks, and/or wireless networks, hence the need for
TCP/IP and its use in wireless/wide area networks.
TCP/IP – (Transmission Control
Protocol/Internet Protocol)
Transmission Control Protocol is a connection-oriented
protocol, meaning that it requires handshaking to set up
end-to-end communications. Only once a connection is
established may user data may be sent bi-directionally
over the connection. TCP is Reliable. It manages message
acknowledgment, retransmission and timeout. Multiple
attempts to deliver the message are made. If it gets lost
along the way, the server will re-request the lost part. In
TCP, there’s either no missing data, or, in case of multiple
timeouts, the connection is dropped. TCP is Ordered – if
two messages are sent over a connection in sequence,
the first message will reach the receiving application first.
When data segments arrive in the wrong order, TCP buffers
delay the out-of-order data until all data can be properly
re-ordered and delivered to the application.
Professional AoIP for Broadcast The Way Forward
Tom Knowles, Product Manager - Broadcast Systems
DHCP - (Dynamic Host Configuration Protocol)
DHCP is a protocol where a server automatically provides
an Internet Protocol (IP) address and other related configuration information such as the subnet mask and default
gateway when a host is connected to its network.
NTP - (Network Time Protocol)
Network Time Protocol (NTP) is a networking protocol for
clock synchronization between computer systems over
packet-switched, variable-latency networks. NTP is intended to synchronize all participating computers to within a
few milliseconds of Coordinated Universal Time (UTC).
PTP - (Precision Time Protocol)
Precision Time Protocol (PTP) is a protocol used to synchronize clocks throughout a computer network. On a local area
network, it achieves clock accuracy in the sub-microsecond
range. In audio Over IP systems, each device has its own
highly accurate internal clock, whose drift relative to the
master clock in the system is controlled by PTP messages.
Using NTP vs PTP for network/system timing all comes
down to the accuracy needed. If the system accuracy
needed is measured in microseconds or nanoseconds then
PTP (IEEE 1588) is required. If the accuracy needed is only
required to milliseconds or seconds, then NTP is sufficiently accurate.
Why is PTP so accurate? Because hardware timestamping is
commonly implemented in PTP technology, but not in NTP.
Hardware timestamping is allowed in the client and server
devices which are running NTP, but not many devices implement this. The largest source of error in network timing
is often due to the variations in queuing time in switches
and routers. NTP does not have a solution for this, PTP
QOS - (Quality Of Service)
QOS is an industry-wide set of standards and mechanisms
for ensuring high-quality network performance for critical
applications. By using QoS, network administrators can
prioritise allocation of existing resources efficiently and
ensure the required level of service for defined classes of
network traffic at the expense of less time-critical services.
QoS must be enabled to allow Dante to share network
infrastructure with other types of data and signals. In many
cases QoS in installed network switches may already be
enabled, if not the network switches need to have the Basic
mode of QoS enabled, checking that the switch is using
DSCP (Differentiated Services Code Point). Education and
use of common language between broadcast and networking engineers is key in successfully leveraging the advantages in AoIP technology. The network specialist may want
to know more about the DSCP labels Dante uses. Audinate
publish the DSCP priority values for Dante.
IGMP - (Internet Group Management Protocol)
IGMP is a network protocol which provides a way for a
network device to report its multicast group membership
to adjacent switches and routers. Multicasting allows one
computer on the Internet to send content to multiple other
computers that have identified themselves as interested in
receiving the originating computer’s content. If it is desirable to send the same data to many devices at the same
time, a single multicast stream provides significant savings
in network bandwidth over using multiple unicast streams
all sending the same data.
Flows are a construct of Dante and can be visualised as
a bundle of audio channels streaming across a network.
Routing audio in a Dante network automatically creates
flows, which carry one or more channels of audio from a
transmitting device to one or more receiving devices. Flows
may be either unicast or multicast.
Unicast routing creates flows to a single receiving device;
a unicast flow typically assigns space for 4 channels of
audio. Multicast routing creates flows that can be received
by multiple receivers. Multicast flows are assigned IDs,
enabling them to be identified in Dante Controller.
More Information
Details of the SSL Network I/O range of Dante based audio
interfaces is available on the SSL web site here.
The SSL Systems Team bring decades of
broadcast systems design to the challenges of
leveraging AoIP tehcnology for broadcast. They can offer
expert guidance on the design of
broadcast infrastructure using AoIP technology.
Click here for help and advice.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF