Product Manual

BRIC-Link Manual

I. INtroduction 8
Major Changes in Firmware 4.0 8
Applications 8
Audio Coding 9
Transmission Modes and Delay 9
Switchboard Server 9
CrossLock 9
Additional Features 10

II. BRIC-Link Diagrams and Installation 11

Pinouts - Balanced Audio 12
Pinouts - AES3 13
Inputs 13
Outputs 13
Pinouts - Serial Port 13
Pinouts - Contact Closures 14

III. Setting Up the BRIC-Link 15

Ethernet 15
Audio Inputs 15
Sampling Rates 15

IV. Using the BRIC-Link Device Manager 16

About Locations 18

BRIC-Link • Nov 2017

 Advanced Ethernet Settings 18
Network Recovery mode 19

V. Controlling the BRIC-Link via The WEb-Based Interface 20

Login 20
Connections Tab 21
Media Statistics Tab 22
CrossLock Tab 23
Audio Metering Tab 23
Profiles Tab 24
Building a Profile 25
Local & Remote Settings 26
Advanced Profile Options 26
Advanced Channel 26
Advanced Channel Options 27
Advanced Encoder Options 27
Advanced Decoder Options 28
BRute ReLIABILIty settIngs 29
System Settings Tab 30
Global Options 31
BRIC Normal Settings 32
EBU3266/SIP Settings 32
Advanced System Settings 32
BRIC Normal Settings 32
Standard RTP Settings 33
EBU3266/SIP Settings 33
TCP Settings 33
VI. Making Connections on the BRIC-Link 35
Creating a Remote Connection 35
Connecting 36
Disconnecting 36
Password Filtering 37

VII. Making BRIC-Link connections with Switchboard 38

VIII. Using BRIC-Link with CrossLock 41

About CrossLock 41
Multi-network support 42
Security when using CrossLock 42
When CrossLock is used 42
Making and Breaking CrossLock connections 42
CrossLock Configuration Options 43
CrossLock Advanced Settings 44
CrossLock Tab 45
Utilization graph 46
Packet loss graph 46
Delay Slider 47
CrossLock Tab Detailed View 48

IX. Operating BRIC-Link in a 24/7 Environment 49

X. About the algorithms 51

AAC 51
HE-AAC 51
 HE-AACv2 51
Linear PCM 51
FLAC 52
G.711 52
G.722 52
Opus 52

XI. multi-streaming 53

XII. IP Multicast 55
Multicast Profiles 55
Setting up a Multicast Remote 56
Time-to-Live 56
Changing Port Numbers for Multicast 56

XIII.streaming server function 57

Decoding a BRIC-Link Stream 58
Simultaneously Connecting BRIC-Links and Streaming 58

XIV. Making EBU3326/SIP Compatible Connections 59

More about EBU3326 59
EBU3326 in BRIC-Link 59
EBU3326/SIP Modes 60
Unregistered Mode 60
Registered Mode 60
SIP Servers 60
SIP URIs 60
 Registering with a Server 61
Making SIP Registered Calls 62
Advanced EBU3326/SIP Topics 62
SIP Troubleshooting 63
Outgoing Call Issues 64
Incoming Call Issues 64
Solutions 64
STUNning Success 64
Fix of Last Resort 65

XV. License and Warranty Disclosures for Comrex BRIC-Link 66

Licenses 66
Warranty 67

XVI. Conformity and Regulatory Information 69

Suppliers’ Declaration of Conformity 69
EC Declaration of Conformity for R&TTE Directive 70

XVII. Appendix A - IP Compatibility 71

XVIII. Appendix B - Using BRIC-Link on Unidirectional Networks 73

Decode Side Settings Only 73
Encode Side Settings Only 73
Full-Time or Triggered Connections 73

XIX. Appendix c - Information for IT Managers 74

About Comrex
Comrex has been building reliable, high quality broadcast equipment since 1961. ur products are used daily in every
part of the world by networks, stations and program producers.

Every product we manufacture has been carefully designed to function flawlessly, under the harshest conditions,
over many years of use. Each unit we ship has been individually and thoroughly tested.

Comrex stands behind its products. We promise that if you call us for technical assistance, you will talk directly with
someone who knows about the equipment and will do everything possible to help you.

You can contact Comrex by phone at 978-784-1776. Our toll free number in North America is 800-237-1776.
Product information along with engineering notes and user reports are available on our website at www.comrex.
com. Our email address is [email protected].

Warranty and Disclaimer

All equipment manufactured by Comrex Corporation is warranted by Comrex against defects in material and
workmanship for one year from the date of original purchase, as verified by the return of the Warranty Registration
Card. During the warranty period, we will repair or, at our option, replace at no charge a product that proves to
be defective, provided you obtain return authorization from Comrex and return the product, shipping prepaid, to
Comrex Corporation, 19 Pine Road, Devens, MA 01434 USA. For return authorization, contact Comrex at 978-784-
1776 or fax 978-784-1717.

This Warranty does not apply if the product has been damaged by accident or misuse or as the result of service or
modification performed by anyone other than Comrex Corporation.

With the exception of the warranties set forth above, Comrex Corporation makes no other warranties, expressed
or implied or statutory, including but not limited to warranties of merchantability and fitness for a particular
purpose, which are hereby expressly disclaimed. In no event shall Comrex Corporation have any liability for indirect,
consequential or punitive damages resulting from the use of this product.
 I. INtroduction

The Comrex BRIC-Link is a low-cost, high-performance solution for audio-to-IP conversion. Leveraging many of the
core technical aspects of Comrex’s successful remote broadcast BRIC-Link product line, the BRIC-Link provides for
an elegant way of moving Linear or compressed audio with very low delay. BRIC-Link may be used over a range of IP
links, is very simple to use, and doesn’t require the expense of more full-featured codecs. While it carries an entry-
level cost, BRIC-Link maintains superb audio specifications and hardware reliability, making the system suitable for
STLs and other mission-critical functions.

BRIC-Link is contained in a small desktop package. Two BRIC-Links may be installed to occupy 1U of rack space.

Major Changes in Firmware 4.0

If you are updating a BRIC-Link from a previous firmware to 4.0 or higher, here are the major differences:
1 CrossLock - This is a new, optional reliability layer available for connection to other Comrex
devices running compatible firmware.
2 Ports - CrossLock makes connections on UDP port 9001. Legacy Comrex audio codec
connections are made on UDP port 9000. This may require that you open additional ports in
your router or firewall settings.
3 Toolbox - 4.0 firmware adds a web-based network configuration utility called Toolbox. This
allows you to change the Ethernet IP settings via a web page. Toolbox is accessed using the
Device Manager software and can also be accessed via a web browser at http://<ip_address>/
cfg/.

Previous versions of firmware used TCP port 8080 for XML commands used by the web interface and the Device
Manager. These connections are now made on TCP port 80 (along with all other web traffic). In some instances,
you may need to change the settings of Device Manager to properly interface with your BRIC-Link.

Applications
BRIC-Link is uniquely suited to point-to-point “nailed up” high-quality audio links over a variety of data networks,
like ISM band IP radios, T1s, satellite channels, WANs, and LANs. The robustness of the BRIC technology (Broadcast
Reliable Internet Codec) used in the box allows the system to perform well on the public Internet as well (using AAC
or Opus compression modes).

8
Audio Coding
For users concerned about delay and coding artifacts, the BRIC-Link offers a robust stereo or mono Linear mode
that does not compress audio. In addition, unique to real-time audio codecs, BRIC-Link offers FLAC lossless
compression, reducing network throughput by 30-40% with absolutely transparent coding and no tandem coding
concerns. For situations where more reduced bandwidth is desired, the BRIC-Link offers AAC/HE-AAC modes as
standard, allowing superb audio quality at dramatically reduced data rates. For compatibility with mobile phone
and web apps, BRIC-Link also implements Opus audio compression, along with VoIP standards G.722 and G.711.

Transmission Modes and Delay

BRIC-Link is a true codec, offering a full-duplex stereo encoder and decoder in each box. When two-way
transmission is not required, the reverse channel may be disabled. The BRIC technology incorporated includes a
jitter buffer manager that automatically balances delay and stability, dynamically increasing and decreasing delay
based on network performance. For networks where the QoS is known, these parameters may be set so that a
consistent level of jitter buffer is maintained.

End-to-end coding delay in Linear modes is less than 25mS. FLAC and Opus modes are less than 30mS. AAC modes
incorporate around 100mS total end-to-end delay, and HE-AAC modes deliver around 220mS.

In addition to coding delay, network propagation and jitter buffers will add delay to any IP link and are network
dependent.

Switchboard Server
Switchboard is an optional upgrade to BRIC-Link. It’s a feature that allows the codec to “sync” with a cloud-based
server. Switchboard allows for easy connections to be made between codecs without any knowledge of IP
addresses on either end of the link. It also provides presence and status information about all the Comrex codecs
in your fleet, and can help make some connections through routers and firewalls that might be difficult otherwise.
Switchboard is enabled by applying a license key (provided by Comrex) to the Web-based Interface of the BRIC-
Link.

CrossLock
Firmware version 4.0 and higher offers a reliability feature called CrossLock. This is an entirely new transport layer
that adds the following features:
• Error Correction (ARQ and FEC)
• Enhanced statistics and diagnostics

Use of CrossLock is optional, and requires a Comrex codec running firmware 4.0 or higher on each end of the link.
CrossLock connections can be made via the Comrex Switchboard function (see previous section) or manually. For
manual connections, CrossLock requires extra settings to assure connections are only made within your known
group of codecs. NOTE: CrossLock is not supported when using Linear PCM or FLAC algorithms.

9
Additional Features
BRIC-Link provides for four end-to-end contact closures to be delivered along with the audio stream in each
direction. Alternately, the contact closure inputs may be configured to initiate connections. An ancillary data stream
is available via RS232 along with the audio stream.

The system is capable of sending up to 6 one-way encode streams (using AAC, HE-AAC or Opus) to separate
decoders (requiring additional bandwidth) and multicasting on capable networks.

Finally, BRIC-Link can act as a streaming audio server. In this mode, BRIC-Link is capable of delivering many HE-AAC
streams that are compatible with computer-based media players like WinAmp and VLC.

10
II. BRIC-Link Diagrams and Installation

L R L R
INPUT OUTPUT ETHERNET CC SERIAL POWER

1 2 3 4 5 6 7 8

Figure 1 Rear
Panel Diagram and Descriptions

1 Left Audio/AES3 Input - Accepts professional level, balanced analog audio, or if configured,
AES3 stereo digital audio for input
2 Right Audio Input - Accepts professional level, balanced analog audio
3 Left Audio/AES3 Output - Delivers professional level, balanced analog audio, or if configured,
AES3 stereo digital audio for output
4 Right Audio Output - Delivers professional level, balanced analog audio
5 Ethernet 10/100baseT - Connection for network connections
6 Contact Closures - Provides for 4 contact closure triggers, and 4 open-drain style contact clo-
sure outputs
7 RS232 - Provides serial data I/O across the IP link. Data rate is configurable
8 Power - 4 Pin connector for attachment of Comrex approved DC power adapter.Requires 24V
DC @ 1A

11
 L R READY RESET
INPUT LEVEL

9 10 11 12 13

Figure 2 Panel Diagram and Descriptions

9 DIP Switches - 8 individual DIP switches allow selection of certain operational parameters (see
DIP Switch Settings on page 14)
10 Left/Mono I/O Level Indicator - Tri-color LED shows left channel input or output level

11 Right I/O Level Indicator - Tri-color LED shows right channel input or output level

12 Ready/Status LED Bi-color LED shows Ethernet carrier loss (Red) or valid connection state
(Green)

13 Reset Button - Issues a hardware reset to the system when pressed momentarily

Pinouts - Balanced Audio

Stereo professional level connections are available on the rear panel via XLR connectors for left and right input and
output. These connectors are wired as follows:

1 Ground
2 Balanced Audio +
3 Balanced Audio -

These connectors have a fixed level where a full scale signal represents +20dBu (22Vpp).

12
Pinouts - AES3
A nominal input level of 0dBu (2.2Vpp) is recommended. When configured via the DIP switches, the left input and
output become AES3 digital ports. The jacks connectors are wired as follows:

1 Ground
2 AES3 +
3 AES3 -

To use AES3, the front panel DIP switches must be set appropriately. AES3 input connections can be at 32, 44.1
or 48kHz. The front panel DIP switches must be set appropriately. On the BRIC-Link, the output audio sample rate
automatically locks to the input sampling rate. If analog input is used, (or AES3 input configured and unconnected),
AES3 output is always 48kHz.

Because BRIC-Link can encode and/or decode in stereo and mono modes, it’s important to understand how the
audio inputs and outputs are handled in each mode.

Inputs
In mono encode modes, BRIC-Link uses the left channel of the stereo input for delivery to the mono encoder.

Outputs
In stereo decoder modes, left and right channels are delivered to the output connectors separately. In mono
decoder modes, mono audio is delivered to both left and right output connectors.

Pinouts - Serial Port

The Serial Port is pinned to match serial connections on older Macintosh computers, so commercially available
adapter cables should have the proper pinning.

Pin # Function Direction

1
2
3 TX Data From BRIC-Link
4 Ground
5 RX Data To BRIC-Link
6
7
8 Ground

13
Pinouts - Contact Closures
Contact closures are available via the 9-pin mini-DIN connector on the rear panel of the BRIC-Link. Inputs are
triggered by shorting the respective input to Pin 9. Outputs consist of an open collector circuit which, when
inactive, will offer a high-impedance path to Pin 9 and, when active, will offer a low impedance path to Pin 9. These
outputs are capable of sinking up to 200mA at a voltage up to 12V.

Pin 1 Output #1
Pin 2 Output #2
Pin 3 Output #3
Pin 4 Output #4
Pin 5 Input #1
Pin 6 Input #2
Pin 7 Input #3
Pin 8 Input #4
Pin 9 Ground

DIP Switch Settings

BRIC-Link has a set of eight DIP switches used for audio and indicator configuration

DIP Switch # Function Default (Down)

1 Analog/AES3 Input Select Analog
2 Analog/AES3 Output Select Analog
3 Audio Loopback* Disabled
4 Level LEDs TX/RX Select TX
5 Headphone out select TX/ TX
RX
6-7 Future Use
8 Factory Reset

* This function connects the send and receive audio together before it touches the encoder - the audio is digitized
and converted back to analog, but not compressed or converted to a stream.

14
 III. Setting Up the BRIC-Link

Ethernet
At a minimum, BRIC-Link needs a source of power, an audio connection, and a network connection. The external
power supply delivers 24VDC at 1A.

The Ethernet connector is a standard 10/100baseT. A normal patch cord, such as one used for a computer, should
be connected here. In most ways, BRIC-Link will look like an ordinary computer to the network. In fact, BRIC-Link
contains an embedded computer with a Linux-based operating system and a full network protocol stack. BRIC-Link
is perfectly capable of working over most LANs. But there may be situations where a LAN is heavily firewalled,
subject to overloaded traffic conditions, or may have security concerns. If running over the public Internet, better
performance is possible if BRIC-Link has its own Internet connection. Often, it’s worth the trouble to install a
DSL line especially for BRIC-Link, especially if the cost is reasonable. Since there may be bandwidth, firewall,
and security concerns with installing BRIC-Link on a managed LAN, it is recommended that your IT manager be
consulted in these environments.

Audio Inputs
Audio inputs should be applied and levels checked with Dip Switch 4 down. If the audio indicators are showing red,
it indicates the level is approaching or reaching clipping stage. It is OK for audio levels to reach the yellow stage
often.

Sampling Rates
When utilizing analog audio I/O, BRIC-Link assumes an audio sampling rate of 48kHz for all encoders. When
utilizing AES3 I/O, the user has a choice of 32kHz, 44.1kHz, or 48kHz. The sampling rate changes have the following
limitations:

1 Whenever utilizing the AES3 input, the AES3 output sampling rate clock will be locked to the
input clock. This means it’s not possible to use a 32kHz input clock and a 48kHz output clock.
If the output is switched to analog, the clock rate of the D/A converter remains locked to the
input AES3 signal.
2 In AAC, HE-AAC and Opus modes, at AES3 rates other than 48kHz, the digital audio is sample-
rate-converted to a stream based on 48kHz on the network.
3 In FLAC and Linear modes, the input AES3 sampling rate is the same sampling rate used by the
network stream. If a different input sampling rate is used on each end of the link, the decod-
ers on each side will sample-rate-convert the stream to the same rate as the local input audio.

15
 IV. Using the BRIC-Link Device Manager
Initial IP configuration is handled using the BRIC-Link Device Manager software, which is a Windows and MAC
executable program. This program was provided on the disc with the BRIC-Link hardware, and can also be
downloaded from the Comrex website.

In order to configure BRIC-Link, the Device Manager must be run on a computer located on the same physical LAN
connection as the BRIC-Link hardware. If this is not possible, you may need to connect an Ethernet crossover cable
between the BRIC-Link and the computer for configuration.

Figure 3

As shown in Figure 3, running the Device Manager and clicking the “Scan” button will produce a list of all Comrex
devices found on your LAN. Device Manager will attempt to log in to each device using the default password. If the
default password has changed, Device Manager will prompt you to input the password for BRIC-Link after the scan.

Figure 3 shows the five tabs that appear on the right-hand pane after Device Manager has logged in. The fifth tab
is labeled Web Configuration. This will open a simplified setup interface on BRIC-Link called Toolbox. The Toolbox
interface allows you to configure several options including the Ethernet port. You will need to log in to Toolbox
separately with a user name (any) and password (default = comrex) to enter the Toolbox.

16
Once logged into Toolbox, choose the Network/Admin/CrossLock option and then choose Set up Ethernet. Choose
the Ethernet port that appears in the list. Figure 4 shows the Ethernet settings of Toolbox.

Figure 4

Figure 4 shows the Ethernet settings of Toolbox.

Generally, you will want to configure the Ethernet port of BRIC-Link for a static IP. This will allow you to more easily
access the Web-based Interface with a browser, and help configure any routers or firewalls if necessary. If BRIC-
Link is to be installed on a managed LAN, you should consult with your IT department about obtaining a static IP
address.

By default, BRIC-Link is configured for DHCP, or dynamic addressing. This means it tries to extract an unused IP
address from the network. To change your Ethernet addressing to static, select the “default” location at the bottom
of the list and change the “IP Type” to static. The system will prompt you for a list of static settings as shown in
Figure 5.

17
Figure 5

You will need to enter the static IP address, the Netmask, the Gateway Address, and at least one DNS Server
Address in the proper fields to set for static IP addressing. Once that information is entered, (and you’ve clicked the
“back” button) you can select Apply Changes to have BRIC-Link accept and activate your new Ethernet settings.
Note that if you’ve changed the IP settings of the Ethernet port, your connection to the Toolbox interface will no
longer work. You’ll need click the “Scan” button on Device Manager to resync with the new IP address.

About Locations
BRIC-Link includes the ability to have multiple “locations” programmed into it for different network settings. E.g.,
if you are moving BRIC-Link between venues, and want to store the static IP information for each venue, you
will define a new “location” (giving it a unique name) using the “Add Location” option shown in Figure 4. Once
multiple locations are defined, you can switch between them using the “Active Location” option in Figure 4.

Advanced Ethernet Settings

By choosing “Show Advanced” under Ethernet Settings, the following options appear:

Preserve after Reset - Normally, when BRIC-Link is set back to factory defaults (via Device Manager), all the
Ethernet settings are erased. By setting this option to “yes”, the settings for this Ethernet port will be preserved
after factory reset. Caution should be used, as it’s possible to “lock yourself out” of the BRIC-Link by setting the
Ethernet parameters incorrectly, which will require the use of Network Recovery Mode (see next section).

Use with CrossLock - On BRIC-Link, this setting should always remain “enabled”.

Broadcast Config - Normally enabled, this option instructs BRIC-Link not to respond to the “scan” function used by
Device Manager. Caution - without the “scan” function, Network Recovery Mode is disabled.

18
Network Recovery mode
Since the Ethernet settings are made with a web connection, keying in incorrect static IP information can result in
losing access to the Toolbox interface entirely. If this happens, Device Manager will still detect BRIC-Link via a scan,
but will be unable to log in. Device Manager has a network recovery tool to help with this.

For security reasons, Device Manager network recovery can only work during the first five minutes of BRIC-Link
operation. If five minutes have elapsed, the unit will need to be power cycled before network recovery can be
accomplished.

Figure 6

Figure 6 shows Network Recovery Mode. The “Scan” button has shown the presence of a BRIC-Link on the
network. On the “Device” tab on the right pane, the “Network Settings” button is activated and a countdown timer
is started. Selecting this will allow changing of the primary Ethernet settings in the same way as Toolbox.

Once you know the IP address (or have changed it) using the Device Manager, the rest of the setup
and operation of BRIC-Link is done via the built-in Web-based Interface, which is accessed directly via a web
browser as described below.

As a reminder, you can access the Toolbox interface on BRIC-Link at any time, without Device Manager, by adding
the suffix /cfg/ to the end of the BRIC-Link IP address (e.g. 192.168.0.25/cfg/) You will always need your BRIC-Link
system password to access Toolbox.

19
 V. Controlling the BRIC-Link via The WEb-Based Interface

Once your IP settings are configured and BRIC-Link has cleanly booted on your LAN, it’s time to take a look at the
BRIC-Link Web-based Interface. This is done by pointing a web browser on your LAN to the BRIC-Link IP address.
To do this, simply type the address into the URL bar of your browser. You will need Internet Explorer 6 or higher
or Mozilla Firefox 1.0 or higher with Adobe Flash Plug-in 7 or higher , or Google Chrome. Opera 8.5 works well
also. If you experience trouble connecting to the BRIC-Link, be sure you have the latest Flash Plug-in installed by
right-clicking your mouse in the main browser window and selecting “about Adobe Flash.” This will take you to the
Adobe website where you can download the latest free plug-in.

Login
Once you are connected to BRIC-Link, a login screen will appear (see Figure 7). Key in any username along with the
default password comrex (case sensitive) to get to the Main User Interface display. This display is optimized for
full-screen mode (F11 on most browsers) on a 1024x768 display.

Figure 7 Web Login

There are three main parts to the BRIC-Link Web-based Interface screen:

1 Main Audio Meter - The level meters are defaulted to off to conserve bandwidth and client
CPU, but when these are enabled this top bar gives an indication of audio levels.
2 Tabs - Use these tabs to control and obtain status of BRIC-Link. They are described in detail in
the next four sections.
3 Chat Window - Allows for a chat utility between any users that are logged into that particular
BRIC-Link web interface. In addition, when BRIC-Link is connected to a remote user, chat text
will appear from any users logged into the remote web interface.

20
Connections Tab

Figure 8 Connections Tab

The Connections tab is the default setting for the Web-based Interface as shown in Figure 8. In this tab, you can
program and save the names and addresses of any remote units you connect to. This allows custom programming
of policy parameters for each remote and allows point-and-click connect and disconnect. To add a remote BRIC-Link
to the list, simply click Store New Remote in the lower section. An input box will appear allowing you to enter a
user name (which can be anything) and the IP address of the unit. You will also need to choose a profile to use
when connections to that remote are initiated. To get started, simply choose one of the default profiles provided
(we’ll show you how to build your own later). You may remove any stored value simply by highlighting and clicking
Remove Stored Remote. Stored remote addresses are saved to system memory, where they will remain through
power cycles.

The Connections tab will also display IP and Status information of a remote BRIC-Link when it has initiated a
connection to you. Their information will only appear while the connection is active.

By default, three users appear on the list. You may use any of these to test different encoder modes.

1 Loopback - Allows for connection between encoder and decoder in the same system.
2 Comrex Lab Voice - This user provides a talk feed from the Comrex headquarters in
Massachusetts, USA for testing your connection.
3 Comrex Lab Music - This user provides a music feed from the Comrex headquarters in
Massachusetts, USA for testing your connection.

21
Media Statistics Tab
The Channel Statistics field (1 in Figure 9) delivers information on the total number of bits entering or leaving the
BRIC-Link (including multiple connections if applicable), IP , UDP and RTP packet headers and coded audio.

3
1

Figure 9 media Statistics Tab

The Active Connections box (2 in Figure 9) breaks this information down further. Because BRIC-Link is capable
of more than one simultaneous connection (in some modes), each connection is listed independently. The raw
Receive Rate and Transmit Rate are listed, along with an indication of how much overhead is required for the
various IP headers on each packet. Frame Loss is also listed as an individual figure for lost and late packets.

Finally, a Max Jitter figure is calculated to give an indication of the time difference between earliest and latest
received packets, followed by an indication of how much delay is being added to the decoder to compensate for
jitter.

Graphical representations of Jitter Buffer Manager activity and Frame Loss are also displayed (3 in Figure 9). The
light blue area in the upper graph represents the jitter values over time. The work of the Buffer Manager is shown
by the green line, which is the target buffer delay that the system is trying to achieve, based on measurements
done over the jitter window.

The lower graph displays a real time and historical representation of frame loss. If the decoder does not receive
packets in time, the chart will show a red line indicating percentage of lost packets over the one second interval.

22
CrossLock Tab
The Crosslock tab is only available during an active CrossLock connection to another codec—it is blank otherwise.

CrossLock concepts are described in the CrossLock section of this manual. It’s an optional connect layer designed
for reliability. The Crosslock tab shows real-time statistics from the CrossLock layer, including error correction
activity. As shown in Figure 10, the data rate graph moves from right to left over a window of 60 seconds. Also,
CrossLock is able to provide these graphs for both directions in the link. The user can choose between transmit and
receive statistics on each codec.

Figure 10

In addition to the data rate, the lower half of the graph is dedicated to displaying the status of IP packets that
are late, lost or recovered with color-coded bars. It’s a visual illustration of the effectiveness of CrossLock’s error
protection in action. Deep red indicated lost packets that could not be recovered, bright red indicates packets that
were too late to be useful, and green indicates successfully corrected packets. The vertical axis of this graph is
percentage of affected packets in a one second window. If this graph shows lots of activity, it’s time to troubleshoot
your network for data loss.

The CrossLock tab has a delay slider bar in the lower half of the tab. This gives a visual indication of the delay
calculations being made by the CrossLock algorithm, while giving the user the ability to override them. Much more
detail on the delay slider is provided in the CrossLock section.

Audio Metering Tab

The Audio Metering tab, as shown in Figure 11, provides a representation of Input and Output audio levels in
several formats. Each of these meters (including the top section meters, which are always visible) may be turned On
and Off individually. All audio meters are defaulted to Off when BRIC-Link is first enabled. This is because transfer
of audio level information consumes bandwidth on the local network, as well as CPU cycles on the client computer.
Whenever BRIC-Link is connected to a data constrained network (e.g. wireless), it is strongly recommended that

23
these meters be Off, especially if the Web-based Interface on the constrained network will also be accessed via
the wireless network (e.g. from the studio end). The bandwidth requirements to drive the meters may affect
performance of the audio codec.

The Metering Quality option (which is defaulted to low) adjusts how often the meters are updated.

Note: Better networks can support higher quality settings

Figure 11 Audio Metering Tab

Profiles Tab
BRIC-Link provides a powerful set of controls to determine how it connects. The Profiles tab allows you to define
one or more profiles to assign to outgoing remote connections. It’s often not necessary to define any profiles since
BRIC-Link ships with a set of default profiles that cover most users, but this tab allows you to build custom profiles
to allow for different encoders in each direction and special options for jitter buffer management. Keep in mind that
these profiles are useful only for connections initiated from the local BRIC-Link. Incoming connections are defined
by the BRIC-Link at the other end.

Profile creation is segmented into commonly used and advanced options. In order to simplify the interface,
Advanced Options are normally hidden from the user.

Remember, building a profile doesn’t change how any remotes are connected until that profile is assigned to a
remote on the Connections tab. Once a profile is defined, it will be available on the Connections tab to be assigned
to any defined connection.

24
Figure 12 Profiles Tab

Building a Profile
To build a new profile, select Add New Profile (1 in Figure 13) and a new profile appears on the list labeled New
Profile. Select it and you’ll see the first set of options available in the General Profile Settings category (2 in Figure
13). Here you can rename the profile to something that will help you remember it. Under the Channel category
(3 in Figure 13), you can select whether this is a UDP IP connection (BRIC normal) or one of the other connection
modes offered by BRIC-Link. These other choices are defined in the Advanced Topics section. Note: It’s important to
define the channel of a profile before moving on to other options, since the choices in the subsequent sections will
vary in this choice. Make sure to press Apply in order to confirm your selection.

Figure 13 General Profile Settings

25
Local & Remote Settings
You’ll be presented with two categories of options: Local and Remote (4 in Figure 13). You’ll use the Local section
to determine how your BRIC-Link behaves, and the Remote section will determine how the BRIC-Link on the far end
behaves.

Each category lists identical options, so we’ll cover only the Local Settings:

Connection Timeout - Under normal circumstances, a connection will be terminated on one end and the other end
will drop the connection in turn. However, if a network failure occurs or a connection is ended abruptly (e.g. killing
power to a BRIC-Link), the system will drop the connection after a predetermined time. The default is 60 seconds,
but this can be shortened or lengthened here. If an indefinite connection is required, see Section VIII Operating
BRIC-Link in a 24/7 Environment for additional information.

Encoder - It’s not necessary to define any decoder types when using BRIC-Link because they automatically adapt to
the incoming stream. Using this menu, you can select the encoder used to send audio from this BRIC-Link (local) as
well as the encoder used to return audio to this BRIC-Link (remote). The default value of the remote encoder is to
follow the local encoder (i.e., it will send exactly the same codec mode it receives). This is defined as Follow Mode
in the remote encoder selection table. See About the Algorithms section for more info on selecting encoders.
Transmit On/Off –-This option determines whether the selected encoder (local or remote) is actually sending any
data. By default, all encoders are turned on, but there may be circumstances where one-way operation is desired
(e.g. multi-streaming, as described in Section IX). Turning off the local encoder disables outgoing audio streaming,
and disabling the remote encoder disables incoming audio streaming.

Advanced Profile Options

The options available in the default mode should provide good performance for most users, but in some
circumstances it may become important to fine tune some of the more obscure parameters that make BRIC-Link
work. By clicking the Advanced Options box in the lower left of the Profile Settings screen, the following Advanced
Options will be available:

Advanced Channel

In addition to BRIC Normal, BRIC-Link provides the ability to set up several other channel types. The Advanced
menu gives the option to use a different channel rather than the normal UDP/RTP created in BRIC Normal mode.

Some explanation:

Internet IP packets come in two flavors: TCP and UDP. Most web browsing, email and other computer-based
functions travel over the TCP protocol, which inherently assures retransmission if a packet is lost, and is therefore
reliable. UDP is optimized for real-time applications, and does not offer any guarantee of packet delivery.
Retransmission typically causes extra delay in an IP network, and BRIC-Link is optimized to conceal an occasional
lost packet, so it makes more sense for BRIC-Link to use UDP for transmission under most circumstances. But there
are occasions where a network will treat UDP packets poorly.

26
Some examples are:
• Networks with high packet loss (rather than jitter)
• Networks with very high security firewalls
• Networks trying to discourage the use of VoIP functions

In these circumstances it makes more sense to enable a TCP channel. The result will usually be a more robust audio
channel with a delay several magnitudes higher than an equivalent UDP channel. Channel overhead is also raised so
you will utilize a higher network bandwidth.

In addition to TCP, there are several other advanced channel modes:

HTTP - BRIC-Link has the ability to act as a streaming server, delivering AAC and HE-AAC to compatible PC based
media players. Normally in this mode, connections are requested on an incoming basis so no outgoing profile
setup is required. But BRIC-Link also has the ability to initiate a stream to a Shoutcast-compatible server in order to
distribute the stream to users. Only in this instance should a profile be set for HTTP.

Multicast - Should only be used to initiate IP Multicast connections (not for use on the Internet). See Section 13 for
more on Multicast connections.

Standard RTP - This setting is used in the unusual scenario where the network is viable in only one direction.
Standard RTP has the ability to send and receive streams without any status information being relayed between the
codecs.

Advanced Channel Options

When designating Local and Remote options for a normal BRIC or TCP channel, several new categories will appear.
Some of them address the encoder and some address the decoder.

Most of the Advanced Encoder options alter the relationship between frames and packets. In this context, a frame
is the smallest chunk of encoded audio that can be extracted from the encoder. For the lowest possible delay, this
frame is wrapped into its own packet and sent into the network.

Advanced Encoder Options

The following advanced options affect the encoder:

Frames per Packet - This function allows the encoder to wait for “X” number of frames to exist before sending a
packet. This option differs from FEC because each frame is only sent once. Setting this value to a number higher
than one can reduce network usage, at the expense of delay. This is because packet overhead bits like IP and UDP
headers are sent less often.

Log Statistics - This function is used in factory diagnostics and should be left disabled unless instructed by Comrex
support.

27
Advanced Decoder Options

Advanced Decoder options have to do with how the jitter buffer manager performs. This is the algorithm that
determines, based on network performance, how much delay to install in front of the decoder to achieve
uninterrupted audio. It does this by creating a statistical analysis of the amount of jitter experienced over a fixed
interval of time (the window) and making a judgment based on other parameters like the decoder’s resiliency to
errors. This is actually a very complex decision-making process involving many variables, and most of the time the
default parameters should work well. The Advanced Decoder options are a means to override these defaults, and
changing them should be done with care.

The following advanced options affect the Decoder:

Retransmit Squelch - These options are used to determine how the buffer manager reacts to typical data dropouts
like those seen on wireless networks. Some explanation:
Many wireless networks have their own layer of data protection riding on top of any other data layer, providing
packet retransmissions in the event of signal fade. The symptom from the network standpoint is that data will come
to a stop for some period of time while the signal is faded, and the network will buffer all packets during this time.
Once the wireless link is restored, all the buffered packets will appear to the decoder as if they were simply very
late. In essence, the protection layer will “fight” the buffer manager. The effect will be that the buffer manager will
expand the buffer, increasing delay dramatically without any benefit.
The Retransmit Squelch allows the decoder to detect these events and avoid having the buffer manager react. The
squelch has several user adjustable parameters with good default settings. These should normally be left where
they are, but there may be unusual circumstances where they should be changed.

Retransmit Squelch Trigger - Determines the amount of time the decoder must experience 100% packet loss before
the Retransmit Squelch function is triggered. Default is one second.

Retransmit Squelch Max - The longest period of data loss during which the squelch function is active (default is two
seconds). During the squelch period, the buffer manager ignored the relative jitter experienced and does not adjust
buffer size to compensate.

Jitter Window - This parameter defines the amount of time (in minutes) that historical network performance is
analyzed in order to make the rest of the calculations. As an example, if the Jitter Window is set to the default
of five minutes, and if a dramatic network event happens and the buffer manager reacts (perhaps by increasing
the buffer), the event will be included in the manager’s calculations for the next five minutes. If the network
experiences improved performance over this period, the manager may choose to wind the buffer back down after
the five minutes has passed.

Loss Cushion - Packets may arrive at the decoder displaying a range of statistical properties. They may arrive in
reasonably good timing and in order, or half may arrive quickly with the other half delayed significantly. In some
cases, most of the packets arrive in a timely manner, but a small percentage of them may be extremely late. In this
case, it’s usually preferable to allow these late packets to be left out, and keep the delay lower. The decoder error
concealment does a very good job of hiding these losses. The Loss Cushion parameter instructs the buffer manager
to ignore a certain percentage of late packets in its calculation. The default value is 5%. Applications that are not at

28
all delay sensitive may wish to reduce this value to zero, while extremely delay sensitive applications may prefer to
have this closer to 25%.

Delay Cushion - The jitter buffer manager usually works very hard to keep absolute delay to a minimum. Some
applications are not delay sensitive and would rather not have the manager working that hard. The Delay Cushion
setting is a way to instruct the manager not to attempt to drive the delay below a certain value. E.g., if the delay
cushion is set to 500mS, this amount of fixed delay will be added to the buffer. If the jitter manager needs to
increase the buffer it will do so, but will not fall below the ½ second level.

Delay Limit - The inverse of the Delay Cushion, this parameter instructs the manager not to wind the buffer out
beyond a certain delay value, regardless of how many packets are lost. This is useful in applications where staying
below a certain delay figure is essential, but use of the delay limit can result in very poor performance if the
network jitter dramatically exceeds the limit.

Fixed Delay - This option simply sets the Delay Cushion and Delay Limit at a similar value, so that the delay buffer is
defined to the chosen value and will not increase or decrease significantly.

Buffer Management On/Off - This is a diagnostic setting used to troubleshoot buffer manager performance by the
factory. For usage, it should always remain “on”.

CrossLock Managed Delay - To understand this setting, you need to understand that there are two ways BRIC-Link
can calculate its target delay (and therefore how much decoder buffer to add). The first is the BRIC-Normal way, and
is the default for non-CrossLock connections. Buffer size is set based on a histogram of past jitter performance. This
will incur the shortest delay possible.

For CrossLock connections, the buffer is increased to allow the use of error correction, so the buffer is based on a
combination of the jitter histogram, and the round-trip-delay as measured by the system. This will usually result in
bigger decode buffers (and higher delays).
Because it is lower, the default setting is to use the jitter histogram for all connections. This setting allows the profile
user to use the CrossLock “error correction friendly” setting, for instances where delay is not of prime importance.

Log Statistics - Function is not supported in BRIC-Link.

BRute ReLIABILIty settIngs

BRUTE Reliability settings are a legacy reliability enhancement compatible with firmware versions 3.0 and lower. For
firmware 4.0 and higher, better performance is achieved using CrossLock modes. More information on CrossLock is
available in the CrossLock section.

Three options are available to help transmissions that are suffering from poor network performance. There are
encoder treatment options, which are applied to the “local” encoder, the “remote” encoder, or both. BRUTE
options require 2.7 or higher software on both ends of the link.

29
Congestion Avoidance - Enabling this option allows the encoder to dynamically change the number of frames per
packet sent, thereby reducing total data requirements. In addition, in most encode modes, enabling congestion
avoidance provides the system a license to step down to a lower encode data rate if desired. This will happen
automatically and with no audio interruption. Step down congestion avoidance is not enabled in the Linear PCM
mode.

UDP Reliability – UDP, the Internet protocol used by BRIC Normal connections, does not have any inherent
error correction capability. UDP reliability adds an intelligent algorithm that requests packet resends only when
appropriate. UDP reliability can be useful on some wireless connections that have unsatisfactory performance due
to packet loss.

UDP Reliability Max Retransmissions – This parameter allows you to set an upper limit on how much additional
bandwidth is utilized by the BRUTE UDP reliability layer. The default setting is 100, which allows the error correction
layer to use the same amount of bandwidth as the audio stream. As an example, if your audio stream is consuming
80 kb/s of network bandwidth, and UDP Max Retransmissions is set at 50%, up to 40kb/s additional network
bandwidth may be used for error correction.

System Settings Tab

The System Settings tab defines parameters that are not specific to a particular remote connection. Examples are
how incoming calls are handled, codec name and how the contact closures are assigned. The System Settings tab is
shown in Figure 14.

The Systems Settings tab has several categories: Global Settings, Aux Serial Settings, Security Settings, BRIC
Normal Settings, and EBU3266 SIP Settings. As with the Profiles tab, basic options are shown by default. Less used
options are hidden until the Show Advanced Options box is clicked.

Figure 14 System Settings Tab

30
Global Options

Several options are available under the Global Settings section.

Unit Name – Users are encouraged to name their codecs here. The default name of a codec is the unique MAC
address of the Ethernet port. By changing this to something familiar and unique (e.g. roving reporter, weather guy,
etc.) this name is reflected in several places:

1 In the browser used to show the remote control page

2 In Comrex provided utility software such as Remote Control and Device Manager
3 In Switchboard TS Buddy lists

Always Connect to Remote - This field is available to designate a remote for always on operation. This is useful in
“nailed up” environments, where a signal is required across the link 24 hours a day. To assign an always on remote,
simply pull down the menu and select which remote to designate as Always On. A connection will be made and
sustained to the chosen remote. Remote connections must be created in the Connections tab before they can be
assigned to this function.
The next four fields define auto connect rules for remotes to be triggered by the four external triggers available on
the rear panel of the BRIC-Link. Note: These inputs are shared with the end-to-end contact closure signals, so if a
remote is designated as Auto Connect on a closure, that closure signal is sacrificed in the direction from this BRIC-
Link.

To assign a remote connection to a contact closure, simply pull down the menu box next to the desired closure
and select the proper remote. A connection attempt will be made whenever the contact is triggered, and will
disconnect whenever the contact is released.

CC Connect Status - Contact Closure Input 4 alters the performance of output contact closure #4. Under normal
circumstances, the signal indicates a trigger of the corresponding contact closure input on the far end of the
connection. If this box is selected, that function is no longer available, and the signal follows the BRIC-Link front
panel Ready light. This signal will be valid (closed) when a valid connection is present, and invalid (open) when no
connection is present.

Aux Serial Settings

This allows you to set the parameters of the auxiliary serial data port provided on the BRIC-Link. This port is always
active during an IP connection and allows serial data transfer along the same path used for the audio data. It does
not remove any audio data; the serial data is added to the packets and bandwidth is increased to support the
additional data. For this reason, heavy use of serial data can affect overall codec performance. Settings are available
for BAUD RATE, DATA BITS, PARITY, STOP BITS and FLOW CONTROL. Most users will leave the defaults of 9600, 8, 1,
None for Flow Control and None for Parity.

31
Security Settings
Connection Password – This allows you to define a password that must be attached to all incoming connections
before they are accepted. Units placing outgoing connections to you must know this password and apply it to their
outgoing stream. Leaving the field blank will disable this function.

GUI Password - This allows you to define a password for the web page screen and firmware updater. The default
password is comrex (lower case). You can disable the remote control and firmware updating functionality
completely by disabling the Remote Control option.

Enable Remote SSH Access - This provides the ability for Comrex support to connect to this unit using the SSH
protocol in order to troubleshoot. We recommend leaving this option enabled, since SSH access requires a key
value that is not disclosed by Comrex, generic SSH requests are rejected.

BRIC Normal Settings

Accept Incoming Connections - This determines if this BRIC-Link is to used for incoming normal IP connections. If
this function is not enabled, BRIC-Link will only support outgoing calls using BRIC Normal Mode.

EBU3266/SIP Settings
Accept Incoming Connections - This determines whether incoming calls are accepted in EBU3326/SIP format (used
for compatibility with other manufacturers who follow this protocol).

Use SIP Proxy - This option determines whether the SIP function is “registered” to a SIP server in the cloud. If “Use
SIP Proxy” is enabled, you will need to add the address, user name, and password for the proxy in the relevant
fields.

SIP Proxy Address - IP address or URL of the SIP proxy used.

Username/Password - Log in credentials for the SIP server, provided by your service provider

CrossLock VPN Settings - See Crosslock section for these settings.

Advanced System Settings

When the Advanced System Settings box is checked, a few additional options are enabled.

BRIC Normal Settings

IP Port - This option allows you to define the incoming UDP port: the number to be used for incoming IP
connections. The default is 9000. Note that since most BRIC-Link codecs attempt a connection on this port number,
changing it can mean the BRIC-Link in the field must dial specifically to your new port number in order to connect.
An outgoing call must be made to a specific port number in the form of IPADDRESS:PORT e.g. dialing port 5004 on
the Comrex test line is formatted 70.22.155.131:5004

32
Standard RTP Settings

These settings offer several modes that allow compatibility with specific IP coding devices. For complete details,
please review the IP Compatibility appendix of this manual.

EBU3266/SIP Settings

IP Port – The port used by the SIP negotiation channel when using EBU3326/SIP Mode. If this port is changed, it’s
likely to break compatibility with other manufacturer’s codecs

VIP QC Password - For legacy purposes with the VIP QC app, which has been deprecated.

RTP IP Port – The port used for audio transfer during EBU3326/SIP mode. Since this port info is transferred during
the negotiation process, it can be changed without breaking compatibility. Note that RTSP data is always sent and
received on the port one address higher than this.

Public IP Override – Enable this in an environment where ports have been forwarded through a router to the BRIC-
Link, and a EBU3326/SIP connection is desired. The SIP protocol is assuming no ports are forwarded and may have
trouble connecting without this function enabled.

Use STUN Server - Determines whether or not to use the STUN derived address in the outgoing fields. BRIC-Link has
alternate NAT Traversal ability so this is off by default.

SIP Proxy Keepalive - Allows you to change how often the SIP proxy handshake happens when no call is present.

SIP Domain - When registering with some services, a separate domain entry is required. If not populated, the
domain of the SIP proxy entry is used.

SIP Auth Username - When registering for some services, a separate Auth Username is required. Do not populate
unless a specifc entry is required by your provider.

Send RTP to source port - A NAT Traversal function used with smartphone apps. Enabled by default.

SIP Routing - Specifically required by some SIP servers (OpenSIPS). Modifies the behavior of the route header.

TCP Settings

BRIC-Link performs best when using UDP for connections but there are some rare circumstances when the system
may need to be switched over to TCP operation. This advanced option defines how incoming TCP calls are handled.

Outgoing calls are defined as TCP when their profile is configured. BRIC-Link normally listens for incoming calls on
both TCP and UDP ports, and chooses the first to arrive. If a TCP call is detected, BRIC-Link will attempt to use the
same TCP link to transmit in the reverse direction.

33
Accept Incoming Connections - This allows you to turn TCP Auto Answer on and off. Disabling this function means
only outgoing TCP calls can be established.

IP Port - You have the option of setting the incoming TCP port number, which can be different than the UDP port
number.

Note: Warnings given above about changing port numbers — calls with mismatched port numbers will fail

34
 VI. Making Connections on the BRIC-Link
Creating a Remote Connection
So now it’s time to make a connection on BRIC-Link. We will assume that the network and audio connections have
been made. Before you can establish an outgoing connection on BRIC-Link, you must enter the info about remote
connection into the Connections tab. This acts like a phone book, saving the names and numbers of everyone to
whom you connect.

As shown in Figure 15, BRIC-Link comes pre-programmed with three connections. Loopback is chosen when you
wish to test BRIC-Link by connecting the local encoder and decoder together. The other two entries are connections
to Comrex in Massachusetts, and these may be used for your testing (when they’re not busy with other users). We
maintain two CD players on these codecs, feeding voice and music audio respectively.

Figure 15 Connections Tab

To create your own outgoing connection, click Store New Remote (1 in Figure 12) to get the entry pop-up. Choose a
name for the remote, (e.g.. WXYZ), followed by the IP address or phone number of the remote.

The MAC address field is only required when using the CrossLock function without the Switchboard server. See the
CrossLock section for more information.

The next field is optional. If the remote has password filtering enabled for incoming calls, you will need to enter
that password into the next field (case sensitive) in order to make a connection to it. If no password is required,
leave this blank.

35
Finally, you will need to choose a profile to use when making these connections. BRIC-Link includes several common
default profiles to choose from, each of which enable a simple full-duplex link using one of the available algorithms.
If you wish for a more complex feature set when making this connection, you will need to click over to the Profiles
tab and set up a specific profile using your custom parameters. Custom options can include one-way transmission,
different encoders in each direction, specialized packet arrangement, etc. Once defined on the Profiles tab, the new
profiles will be available in the Profile Select window and they can be assigned to a remote connection.

Figure 16 Store New Remote

Connecting
Once your remote connection entry is correct, it’s simply a matter of pointing and clicking to connect and
disconnect a remote. When a connection is attempted, the Current State value in the connection table will change
to reflect the progress of the connection. If the connection fails, the reason for failure will be shown in the Last
State category. If it succeeds, the encoder and decoder mode will be reflected in the Transmit Status and Receive
Status columns.

Disconnecting
Disconnecting is just as simple. Highlight the desired connection and click DISCONNECT to end the connection.

36
Password Filtering
The CONNECTION PASSWORD function can be used to filter incoming connections. Using this function, attempted
incoming connections will be rejected if they do not know the proper case-sensitive password. For outgoing
connections, the password is entered when the remote connection is created on the Store New Remote menu. For
incoming connections, the password is set on the System Settings tab.

There is no way to retrieve a forgotten password; it must simply be changed in each BRIC-Link.

37
 VII. Making BRIC-Link connections with Switchboard

Note: Use of the Switchboard Server (formerly known as BRIC Traversal Server or TS) is an optional upgrade for
BRIC-Link. In order to use the Switchboard function of the BRIC-Link, a license key must be applied to the product
using the Device Manager utility. The license may also have been applied at the factory. You can check for License
status of a BRIC-Link on the System Settings Tab by selecting the Software Licensing button.

Comrex maintains a Switchboard Server (ts.comrex.com) and a NAT Traversal server (stun.comrex.com) which is
available for use by BRIC-Link users. An account can be provided on Switchboard by Comrex support staff, or has
already been included in your Switchboard upgrade purchase.

For information on setting up and maintaining your Switchboard account see, the technote Setting up SwitchBoard
for ACCESS and BRIC-Link codecs on the Comrex web site.

Once Switchboard is active on your BRIC-Link, you’ll see its status reflected on the Connection tab in the Status box
as shown in Figure 17.

Figure 17

38
Figure 18

The System Settings tab has a few Switchboard Configuration options available under Switchboard settings. As
shown in Figure 18, the function can be enabled or disabled with the “enable” option. The two address fields
reflect the locations of the Directory part of the server (Switchboard) and the location of the portion of the server
that aids in NAT Traversal (STUN). If you host your own Switchboard, either of the fields can be changed to the IP
address or URL of your private server. Both addresses can be the same.

Switchboard will show you a dynamic list of all other Comrex codecs in your account that are subscribed. The last
option determines the scope of this list. By default, all other codecs in your account are always displayed, but you
can limit this list to units that are currently “synced up” to Switchboard, excluding any units offline or with their
SwitchBoard function disabled.

39
Figure 19

Figure 19 shows the Connection tab of a BRIC-Link with a properly operating Switchboard service. The listings
with a gray background are those that appear via Switchboard. These can generally be connected without any
knowledge of IP address with the help of Switchboard. A default profile is assigned to Switchboard connections
(Opus mono) but can be changed by editing the entry in the usual way.

40
 VIII. Using BRIC-Link with CrossLock
Firmware 4.0 for BRIC-Link adds an entirely new reliability layer, which is optional to use. In order to place
connections via CrossLock, both ends of the link need to be a BRIC-Link, BRIC-Link II, ACCESS Rack or Portable
running firmware 4.0 or higher.

About CrossLock

Figure 20

As shown in Figure 20, CrossLock is a VPN (Virtual Private Network) that gets established between your codec
hardware before a connection is established. Using this VPN, your codecs can transfer much more information in
each direction of the link. This information includes network status and packet loss statistics in each direction, error
correction parameters, and information needed to set encoder throttle rates.

The CrossLock VPN gets created immediately when a new connection is initiated, and lingers for a short time after
the connection. Options are available to have the CrossLock VPN established always, with only the media streams
and statistics being make/break.

NOTE: CrossLock is not supported when using Linear PCM or FLAC algorithms.

41
Multi-network support
Due to the limitations of the hardware, multi-networking is not possible with BRIC-Link. However, the BRIC-Link will
accept connections from equipment utilizing the multi-network feature (BRIC-Link II, ACCESS Rackmount or ACCESS
2USB.)

Security when using CrossLock

Because CrossLock creates an actual Virtual Private network (VPN) between your codecs, it is configured to be
closed to codecs not known to you. This is done by limiting connection to those contained in a “list” of approved
unit IDs. The unit ID of a codec is the Primary Ethernet port MAC address.

So to connect via CrossLock, each unit, in addition to knowing the other’s IP address, must know each other’s
Ethernet MAC address.

The easiest way to connect codecs using CrossLock is to use the Switchboard server function. Switchboard is
specially designed to transfer all needed information about your codec fleet to all other codecs in the fleet. Unit IDs
are distributed among the units in the contact list and connections can be made without any further input on either
end.

When CrossLock is used

BRIC-Link will make a CrossLock connection when the following conditions are met:
1 Both ends of the link have Firmware 4.0 or higher.
2 There is at least one active network attached at each end.
3 The connection is established via Switchboard OR The MAC address of each unit is added to
the outgoing and incoming remote entry.

If these conditions are not met, a legacy-style “BRIC Normal” connection without CrossLock will be attempted.

Making and Breaking CrossLock connections

If using Switchboard as described above, CrossLock connections will happen automatically if both codecs are
capable. The easiest way to tell whether a connection is using CrossLock is that the word “CrossLock” is displayed in
the status entry as shown in Figure 21.

Figure 21

42
In the case of non-Switchboard based connections (e.g. closed networks or STLs), you will need to know the unit ID
(Primary Ethernet MAC address) of the unit to which you wish to connect.

If not using Switchboard, the unit ID must be applied to the outgoing remote entry. In addition, a remote entry on
the incoming codec must be made in the same way, with the Unit ID of the calling unit populated in the same field
(even if this entry is never selected for outgoing calls). The IP address field of the incoming entry does not need to
be populated.

Figure 22

As shown in Figure 22, the unit ID is input to the “Create New Remote” pop-up in the “MAC Address” field.

In addition, the codec receiving the connection must have a similar entry made, with the MAC Address
of the calling unit populated.

CrossLock Configuration Options

As shown in Figure 21, there is a CrossLock VPN settings option on the System Settings tab of the Web-based
Interface. Here are the basic options:

Enable - By turning off the enable function, CrossLock connections can not be received or established.

Redundant Mode - Not available for BRIC-Link “Classic.”

43
Throttling - One key element of CrossLock is the ability to determine whether the data rate of the encoder should
be reduced to compensate for network contention. CrossLock does this by default. If BRIC-Link is being used on a
managed private network, or if multiple CrossLock sessions are employed (Multistreaming) then this function can
be defeated here.

CrossLock Advanced Settings

Selecting the Advanced option under CrossLock VPN settings reveals the following CrossLock Advanced settings:

UDP Port - By default, CrossLock uses UDP port 9001 for connections. For best results, this port should be open
for incoming data on at least one of the codecs in the link. This means that unless the BRIC-Link is an an “open”
Internet connections (no firewalls or routers used) the port will need to be forwarded to it. In instances where
more than one codec will be attached to the same public IP address, you may need to change the default incoming
port. It can be changed here. If this port is changed and Switchboard is used to establish connections, no further
changes are required. In the case of connections without Switchboard, the port change will need to be noted in the
outgoing address on the calling unit.

Permissive - Enabling Permissive mode removes the unit ID filter entirely. CrossLock connections can be made
without regard to unit ID. Note the far end unit must know this codec’s unit ID, or must also have permissive mode
available. Recommended for closed networks without security concerns.

Authentication - BRIC-Link firmware 4.0 and higher uses security certificates assigned to the codec hardware to
authenticate it as a Comrex product. This option determines whether connections will be made to codecs without
these certificates. Certificates are assigned to codecs by the Switchboard server after an upgrade to firmware 4.0 or
higher (Switchboard upgrade not necessary). Because some codecs may be firewalled and not received certificates
after upgrades, this option is defaulted off.

Encryption/Protection - BRIC-Link has the ability to prevent interception of streams (Encryption) and alteration
of streams (Protection). The CPU requirements of these modes are large, and therefore it is not recommended to
apply these options to streams when not required. They are set to off by default to conserve CPU.

Maximum Delay - CrossLock operates by choosing a “Target Delay” figure based on jitter performance of its various
networks over a time window. To prevent excessive delay in the case of one extremely laggy network, it has a
maximum delay setting here.

FEC - CrossLock has a powerful Forward Error Correction algorithm that is enabled in the presence of multiple
networks, using any excess network bandwidth to add in the parity required. Use of FEC is recommended, but it can
be disabled here.

FEC Delay - Controls both how much delay is introduced into the system by FEC , and also to an extent how effective
the recovery is (which is dependent greatly on the packet rate). The default of 100mS should only be altered on
recommendation of Comrex support.

44
Base FEC - This parameter applies a constant rate of FEC targeting recovery of the specified expected loss rate. It is
measured in percent packet loss to be corrected. This is useful when retransmission is not effective (e.g., high delay
network) and auto-FEC is not working as desired. It should be used on recommendation of Comrex support.

Retransmit - In addition to FEC, CrossLock utilizes an ARQ style algorithm to allow retransmission of lost packets
when time permits. This mode is recommended but can be disabled here.

Header Compression - The nature of Internet packets sometimes results in IP overhead (RTP headers and other
info) actually using nearly as much bandwidth as the payload. CrossLock, by default, compresses some of these
headers to conserve network bandwidth. In instances where the network rejects this, or packet inspection is
required, this compression can be disabled.

Always Connect - This option provides for CrossLock to be always connected to a destination. By it’s nature,
CrossLock uses very little data so the network utilization of this mode (when idle) is very small. If you only connect
to one destination, having CrossLock always connected makes media connections faster, and provides an indication
of network status between the devices (“ready” light or CrossLock status). Most users should leave this setting off.

CrossLock Tab

Figure 23

NOTE: This figure shows 2 networks being utilized. BRIC-Link does not have the ability to use more than one
network

When a CrossLock connection is active, the CrossLock tab becomes active. Otherwise it’s blank. The CrossLock tab
is shown in Figure 23. It provides a large amount of real-time information about network performance.

45
The CrossLock tab is similar to the information available on the Statistics tab, which shows streaming performance
without regard to the CrossLock layer. The CrossLock tab shows finer details about network performance in both
directions than can be obtained through the Statistics tab.

In addition, the CrossLock tab contains the Delay Slider Bar. This bar gives a visual indication of the target (delay
the system thinks is required) and actual delay of the link. The default of the slider bar is “automatic” operation, but
manual mode can be engaged in circumstances where desired.

As shown in Figure 23, you can choose between seeing graphs for the transmit (outgoing) or receive (incoming)
side of the link using the pull down option in the upper left side.

Utilization graph
The top section contains a graph of the outgoing (or incoming) utilization of the network. The bars indicate the
average data rate used by the system during each one second window. It is expected that the size of these bars
will vary because CrossLock has control over data rate through a technique called “throttling”. Based on network
feedback statistics, CrossLock will reduce and increase the utilization dynamically.

If more than one network device is in use, the utilization graph will be color coded, indicating the relative utilization
of each network device. The color code key for each network device appears on the right side of the graph.

Overlayed on the network utilization graph are two colored lines:

1 Encoder target - This reflects the bitrate chosen in the profile used in the connection. This
is treated as a maximum value, so utilization should mostly remain below this line. Because
these values are averaged, there may be moments where the utilization moves above the
target line momentarily.
2 Throttle rate - When CrossLock throttling is enabled (the default state) this line indicates how
much throttling will be applied. You will see utilization stay mostly below the throttling line.

Packet loss graph

The bottom graph indicates, in percentage terms, what’s gone wrong on the network during each one-second
window. Three different color-coded entries appear here:

1 Packet Loss (dark red) - The system has detected a packet has been completely dropped by
the network and was never received by the decoder.
2 Packet Late (bright red) - The system received the packet, but it was too late for decoding and
playout.
3 Packet recovered (green) - The packet was either lost or late, but was recovered either by the
Forward Error Correction (FEC) or Automatic Repeat Query (ARQ) error correction built into
CrossLock.

46
Delay Slider
The most powerful way to stabilize any streaming connection is to have the decoder add a delay buffer to the
connection. This compensates for changes in the rate packets are received - known as jitter in Internet speak.
CrossLock uses a combination of decode delay buffering and error correction to keep connections stable. When
CrossLock is active, the activity of the delay buffer is illustrated and controlled via the delay slider on the CrossLock
tab.

Figure 24

Figure 24 shows the Delay Slider. In this figure, the slider is in “Auto Delay” mode and the information on the slider
is purely for informational purposes. Clicking off the “Auto Delay” box sets the system to Manual Delay mode and
allows the slider to be moved with a mouse.

The entire slider is scalable, and the range of it from left to right will vary from one hundred milliseconds to several
seconds depending on the range of delays currently being addressed. In either Auto or Manual mode, a series of
color bars are overlayed on the slider, to signify delay “zones” of safety. Furthest left is the red zone, which indicates
a buffer level that is too low for stable transmission. The yellow zone indicates a delay buffer that may have stability
issues, and the green zone indicates a buffer level that should provide stability. These “zones” scale, increase and
decrease in size based on the history of jitter experienced by CrossLock on the network.

In “Auto Delay” mode, two other elements are of interest. The downward arrow signifies the “Target Delay”, which
is the best compromise value calculated by the system to balance stability and delay. The “bubble” indicates current
delay. The Current Delay bubble will attempt to track with the Target Delay Arrow, but may at times fall outside it.
This usually indicates that the system is actively reducingor increasing the buffer.

Figure 25

Figure 25 shows the slider in “Manual Mode”. Here, the arrow becomes a moveable cursor, and it’s up to the
operator to determine how best to set the compromise between delay and stability.

47
CrossLock Tab Detailed View
A powerful feature of the CrossLock Link tab is to allow the user to “drill down” to the statistics for
each network interface, and view the performance of that network alone. By clicking the button
that appears for each interface on the right side of the tab, a new window opens that shows
graphs similar to the “overview” graph, but limited to the performance of that particular network.

48
 IX. Operating BRIC-Link in a 24/7 Environment
BRIC-Link can be easily set up for “always on” operation. It will be helpful to describe a little bit about the BRIC-Link
data transfer protocol before describing how to set the system up.

In BRIC Normal mode, the default mode of operation, BRIC-Link transfers all its audio data via the UDP protocol.
This is in contrast to most web-based connections like browsing and email, which use the TCP protocol. UDP, unlike
TCP, is not “connection oriented”, (i.e., no virtual connection actually exists in this protocol layer between the
devices). In UDP, the transmitter simply launches packets into the network with the correct address, hoping the
network will make its best effort to deliver the packets in a timely fashion. If a packet is delayed or lost, no error
message is sent and no packets are retransmitted. It is up to the receiver to cover up any lost data, if it can. This
allows the Internet to deliver packets with the smallest amount of overhead and delay. Since there is no intelligent
connection built between the codecs, there isn’t actually any connection to break in the event of network failure.
The encoder simply launches packets into the network, regardless of whether they arrive or not. If the network fails
and is later restored, the packets stream will be restored to the decoder.

For most applications like remote broadcasting, it’s useful to simulate a connection-oriented stream, so BRIC-Link
uses a low-bandwidth sub channel to deliver information back to the encoder about overall connection status. It
does this in its “application layer”, rather than the “transport layer” where UDP exists. By default, it monitors the
health of a connection and if no data is detected as received by the decoder for 60 seconds (this is a user adjustable
timeout), it “tears down” this connection and goes back to idle state. This can give an indication to the user that the
network has failed and it’s time to look at the problem.

The good thing about having the connection protocol in the application layer is that its use is optional. For 24/7
operation, there’s no advantage to having the connection end if no data is received for a timeout interval.

So to set BRIC-Link for 24/7 operation, several parameters are changed:

1 The timeout value is set to infinity; the connection will never be torn down regardless of data
status.
2 BRIC-Link is configured to re-establish the connection in the event of a power-up.
3 The local Disconnect control is disabled. The Disconnect function on the receiving side is still
enabled, but will result in an immediate reconnection by the initiating side.

In the System Settings tab, the field labeled ALWAYS CONNECT TO REMOTE offers a pull-down menu of all available
connections. Setting this value to one of your pre-defined connections results in configuring the unit for 24/7
operation to that remote. No configuration is necessary on the remote side.

BRIC-Link has another option for persistent connections. When building a remote entry a field is available for
backup options, one of those options is Keep Retrying This Remote mode. In a similar fashion, using this mode
will allow the unit to disregard the timeout value and keep a persistent connection. The difference is that the

49
Disconnect function still works and the connection will not be reinitiated on a power-up. This mode is meant for
users who are making temporary connections, but do not want the system to time out and disconnect in the event
of network failure.

BRIC-Link has the capability of automatically making a backup IP connection in the event of failure of the main
connection. This is called Fallback, and is an option chosen after defining a new Remote connection.

Figure 26 Backup Remote

As shown in Figure 26, highlight an existing connection (This will be your primary connection) and choose Change
Remote Settings. In the pop up window, a pull-down box is available to allow you to choose a fallback connection
from the list of existing remotes.

After connection, if data is stopped on the primary connection for the length of the timeout value (set in the
connection’s profile), a connection will be attempted and maintained to the fallback remote.

There is also a box in the Change Remote Settings tab labeled Automatically fall forward. If this box is checked,
the system will constantly attempt to reconnect the primary remote while connected to the fallback remote. If
connection is successful, the connection to the fallback will be terminated.

50
 X. About the algorithms
When building a profile, there are several choices of encoders to use for each direction of the link. Here’s a
description of the choices:

AAC
This algorithm is a highly regarded standard for compressing audio to critical listening standards. It has been judged
to produce “near transparent” audio at a coding rate of 128 kbps stereo. The standard is a collaborative of several
audio companies best efforts, and has become popular as the default audio codec of the Apple™ iTunes™ program.
AAC should be considered the highest quality codec in BRIC-Link - Enhancements like HE-AAC and attempt to
maintain a similar quality and reduced bandwidth and delay.

HE-AAC
This is a newer version of AAC defined for increased efficiency. The goal of the algorithm is to produce AAC
comparable quality at a lower bit rate. It does this by encoding lower frequencies to AAC, and higher frequencies
using Spectral Band Replication (SBR), a technique that partially synthesizes these high frequencies. HE-AAC is
trademarked by other companies as AACPlus™. HE-AAC (and close derivatives) are often used as the main audio
codec for digital radio and satellite networks.

HE-AACv2
This algorithm further increases the efficiency of HE-AAC by adding intensity stereo coding. This results in a lower
bit rate for stereo signals. We also cluster a very reduced rate HE-AAC mono into this category, although technically
it does not contain v2 coding.

Linear PCM
This encoder does not compress audio at all. It uses a 48 kHz sampling rate (using analog inputs or 48kHz AES3)
and simply applies small frames of linear audio to IP packets. This mode is only useful on high bandwidth LAN or
managed WAN environments. Mono Mode requires a network capacity of 804 kbps while Stereo (Dual Mono)
Mode requires a network bandwidth over 1.56 Mb/s.

In Linear PCM, if the input AES3 sampling rate is 32kHz or 44.1kHz, the network stream will also run at this rate and
required bandwidth will be lower.

51
FLAC
This encoder compresses the audio data using a lossless algorithm. This means that the audio extracted from the
decoder is identical to the audio input to the encoder, with no coding artifacts. FLAC typically removes 30-40% of
the network data compared to Linear PCM, but the actual data rate is variable and is based on the complexity of
the coded audio.

Using FLAC over Linear PCM typically results in a slightly higher (5ms) overall delay.

G.711
G.711 (µ-law and a-law) - These are the coding algorithms used by standard digital POTS calls, and provide about
3kHz (telephone quality) audio. µ-law is utilized in North America, while a-law is prevalent in Europe. These
algorithms are provided for compatibility with SIP-style VoIP phones, but don’t provide much benefit over standard
telephony in audio terms.

G.722
G.722 - This is a well known 7kHz (medium fidelity) algorithm used in some VoIP telephones and codecs. It is
provided for compatibility purposes, but is not considered a superior algorithm for audio codecs.

Opus
Opus - A newer offering that combines low delay and low network utilization. Opus is included primarily for
compatibility with softphone apps, and Internet connections using WebRTC (see Technotes about WebRTC on
the Comrex website). Special CBR modes are offered for compatibility with Tieline products - avoid these in other
applications.

*Linear PCM and FLAC are not supported when utilizing CrossLock.

52
 XI. multi-streaming

BRIC-Link supports the ability to run one encoder per box, but this single encoder stream may be sent to up to
three destinations simultaneously. We call this capability multi-streaming, since the encoder creates a separate but
identical outgoing stream to each decoder. Note: Your Internet connection must be able to support these streams.
For example, if your encoder runs at 35 kbps network utilization, sending to two locations will require 70 kbps
upload speed from your network.

Multi-streaming should not be confused with IP Multicast, which is described in the next section. CrossLock should
be turned off for Multi-streaming connections.

Each BRIC-Link can also run only one decoder. So it’s important that in a multi-stream environment, a maximum of
one stream is sent in the reverse direction. This means that users interested in hearing a multi-stream must turn off
their encoders.

This can be a bit confusing because multi-streams can be initiated from either end of the link.

Figure 27 shows an BRIC-Link multi-stream arrangement. BRIC-Link A is the multi-streamer, with BRIC-Link B, C and
D listening to the same audio. In order to set up a multi-stream scenario, you will need to know how to turn BRIC-
Link encoders Off. This must be done by building a profile with either the Local or Return Transmitter mode set to
Off, as shown in Figure 28.

INPUT OUTPUT L R READY RESET

INPUT LEVEL

User B - Sending and Receiving AAC Mono

INPUT OUTPUT L
INPUT LEVEL
R READY RESET
Internet
INPUT OUTPUT L R READY RESET

User A - AAC Mono INPUT LEVEL

User C - Receiving AAC Mono

INPUT OUTPUT L R READY RESET

INPUT LEVEL

User D - Receiving AAC Mono

Figure 27 Multi-Streaming Arrangement

We’ll give two examples of multi-steaming scenarios. The first is an environment where the BRIC-Link that is serving
the multi-stream initiates the calls, and in the second the serving BRIC-Link accepts all its incoming connections.

53
Figure 28 Transmit On/Off

In the “multi-streamer as caller” model, two different profiles will be built on BRIC-Link A. The first profile, labeled
“Multi-Duplex”, will be defined as a normal, full-duplex BRIC-Link connection. The encoder to be used will be
selected in the Local Encoder section, and the stream desired in return will be defined in the Remote Encoder
section.

The second profile is called “Multi-Simplex” and in this profile the Remote Transmitter is turned Off. Most other
selections in this profile are irrelevant.

User A will define remote connections for BRIC-Link B, C, and D. He will assign the “Multi-Duplex” profile to BRIC-
Link B, and “Multi-Simplex” profile to the others. He will then establish a connection with BRIC-Link B first, followed
by C and D.

In model number 2 where the serving BRIC-Link accepts all incoming connections, all the profiles are built on the
Remote Receivers. BRIC-Link B will use a simple profile by defining the encoders in each direction, and assign it to
BRIC-Link A. BRIC-Link C and D will each define a profile with their Local Encoders turned Off, and assign them to
A. BRIC-Link B should connect first. When C and D connect, they will hear the same stream as B, regardless of how
their Remote Encoders are set in their profiles.

In a multi-streaming environment, the first man wins. For example, the first connection made between units will
determine the encoders used for all others. After the first full-duplex connection is made, all other attempts at
full-duplex connections to either end will be rejected.

54
 XII. IP Multicast

IP Multicast is an efficient way of delivering BRIC-Link digital audio streams to multiple locations. This involves
relying on the network to distribute the stream to the locations that require it, rather than creating an independent
stream for each user.

IP Multicast requires the use of an IP Multicast-capable network. The commercial Internet, with few exceptions, is
not capable of supporting IP Multicast. Some private LANs and WANs are IP Multicast capable.

IP Multicast supports only a single direction stream. An IP Multicast encoder can not receive input streams.
CrossLock should be turned off for IP Multicast connections.

In this manual, we assume that IP Multicast users will be familiar with the basic concepts of setup and operation of
the network, so we will focus on how to configure BRIC-Link for Multicast mode.

Multicast Profiles
To set any remotes to Multicast, you must first create a profile for either a Multicast Sender or a Multicast Receiver
on the Profiles tab.

Figure 29 Multicast Settings

55
As shown in Figure 29, when you define a new profile, you have the option to choose Multicast as the profile type.
Multicast profiles have fewer options than other profile types, and some of the available options will have no effect
(e.g. setting an encoder type on a Multicast receiver has no effect). The important settings for Multicast are:

• Sender/Receiver - Determines whether this particular BRIC-Link is designed to generate the IP

Multicast stream (send) or decode one (receive).
• Encoder Type - Determines the type of stream to be used by the Multicast encoder - not
relevant for decoders.

In addition to the basic options for IP Multicast profiles, clicking the Advanced box will allow setting of the same
Advanced Options available for Normal BRIC (Unicast) profiles. See the Profiles tab section for more information.

Setting up a Multicast Remote

All Multicast connections are outgoing connections - A Multicast Sender must initiate an outgoing stream, and a
Multicast Receiver must initiate an incoming one. These remotes are configured within a special address range
known as a Multicast Block, typically 224.0.0.0 to 239.255.255.255. To establish a Multicast connection, simply
define a remote as having an address within the IP Multicast Block, use an IP Multicast profile, and press Connect.

Time-to-Live
Time-to-Live (TTL) is a variable set by Multicast encoders to determine how long a packet is processed before it is
dropped by the network. The default value of TTL in BRIC-Link is 0, which limits its use to within a LAN environment.
TTL may be manually changed on a Multicast Sender remote by configuring the IP address followed by a “/”,
followed by the TTL value. An example remote Multicast encoder could be set for the address 224.0.2.4/255, which
would signify an address with the Multicast Block with a TTL of 255 (which is the max value available).

Changing Port Numbers for Multicast

The default port of UDP 9000 may also be changed on Multicast remotes. The port number is assigned in the usual
way, directly after the IP address, preceded by “:”, followed by the TTL. As an example, the IP address of a Multicast
Sender on port 443 with a TTL of 100 would read:

224.0.2.4:443/100

56
 XIII. streaming server function

BRIC-Link has the ability to act as a streaming server, delivering AAC and HE-AAC to compatible PC based media
players. Currently tested media players include WinAmp, VLC and Windows Media Player 12 and up.

By default, streaming server functionality is turned off. To enable it, go to the System Settings tab of the User
Interface and choose HTTP Settings option. Under the first option, set Accept Incoming Connections to Enabled.
This will allow outside users to initiate “pull” connections to your codec.

The default port to serve streams from is 8000. If you want to change this, it can be done in the HTTP settings under
IP Port. Note you will need to reference the specific server port in the URL you provide to listeners (see below).

Figure 30 HTTP STREAMING ENCODER

Next you will need to choose an encoder for use by the streaming server. Only the encoder choices that are
compatible with the players listed are shown in this menu. Choices span between a mono audio feed at 18kb/s,
up to a stereo feed at 128kb/s. Keep in mind, multiple streams will require this bandwidth along with around 25%
overhead for each stream.

The Genre, Info URL and Public options may be set for anything, or left alone. These options, if applied, will be
embedded into the stream.

57
Decoding a BRIC-Link Stream
To decode a stream, open one of the supported players and find the option to open a URL-based stream.

In Winamp and VLC, input the address of the BRIC-Link in the following format:

http://192.168.0.75:8000
(using the actual IP address, and the actual port if not changed from default 8000)

In Windows media, input the address like this:

http://192.168.1.75:8000/stream.asx
(using the actual IP address, of course)

Simultaneously Connecting BRIC-Links and Streaming

BRIC-Link can stream while connected to another BRIC-Link in Normal mode. If the BRIC connection is using an AAC
algorithm supported by players, when a stream is requested it will be delivered using the same encoder as the BRIC
connection, regardless of the HTTP settings. If the BRIC-Link encoder is Linear or FLAC, the stream request will be
rejected.

58
 XIV. Making EBU3326/SIP Compatible Connections

Comrex codecs (and many other brands) have a set of protocols that allow easy IP connections between units. In
general, when connecting between Comrex hardware, it’s best to use these proprietary modes to take the most
advantage of the features of the product.

However, many users are concerned about getting “locked in” to a certain codec brand. Because of this, an
international committee was formed by the European Broadcast Union called N/ACIP to hammer out a common
protocol to interconnect codec brands. This committee resulted in the establishment of EBU3326, a technical
document describing how best to achieve this goal.

EBU3326 by and large establishes a set of features each codec should support, then leaves most of the heavy lifting
to other, previously established standards like SIP (IETF RFC 3261). Topics not covered (yet) by EBU3326 include
things like carrying ancillary data and contact closures from end-to-end, codec remote control and monitoring, and
complex NAT traversal, which at this point are still left to the individual manufacturer’s discretion. So if these topics
are important to your application, it’s best to stick to a single codec vendor and their proprietary protocols.

More about EBU3326

The Tech 3326 document defines several mandatory encoding algorithms, and the transport layer that could be
used on them for compatibility. But the most complex part of the standard was the decision on how to arrange
Session Initialization, which is the handshake that takes place at the start of an IP codec call. The most commonly
used protocol is called SIP, which is used extensively by VoIP phones and therefore was a logical choice. SIP carries
the advantage of making BRIC-Link compatible with a range of other non-broadcast products, like VoIP hardware,
software, and even mobile phone apps.

EBU3326 in BRIC-Link
BRIC-Link does not fully comply with EBU3326, as it does not feature the mandatory MPEG Layer II codec. Aside
from this, BRIC-Link has been tested to be compatible with several other manufacturer’s devices using encoders
supported by both products. When using EBU3326/SIP Compatible mode (this is what how the user interface
describes EBU3326), ancillary data, contact closures, Switchboard TS, Multi-streaming and Multicasting are not
supported. Outgoing call profiles built with the NACIP/SIP channel may lack some advanced options, and can not be
set for different encoders in each direction (i.e. EBU3326/SIP calls are always symmetrical).

59
EBU3326/SIP Modes
A function of placing a SIP-style call is the ability to register with a SIP server. This is a server that exists somewhere
on the network, usually maintained by a service provider. Several free servers exist that can offer registration like
GetOnsip.com.

BRIC-Link allows EBU3326/SIP calls to be placed or received with or without registration on a SIP server. If
registration is not enabled, connections are made directly to the compatible device by dialing its IP address, just like
in BRIC Normal mode.

Unregistered Mode
Placing a call in Unregistered EBU3326/SIP mode is simple--just build a profile, but instead of choosing BRIC
Normal channel, choose EBU3326/SIP. This will make sure the call is initiated on the proper ports and with the
proper signaling. The majority of system settings relating to EBU3326/SIP relate to Registered mode.

Registered Mode
Registering with a SIP server in EBU3326/SIP mode can have some advantages. When using a SIP server:

• The server can be used to help make connections between codecs through routers
• The remote codec can be dialed by its SIP URI instead of IP address
• The SIP server can be used to find codecs on dynamic IP addresses

SIP Servers
A SIP server exists in a domain. This domain is represented by a web-style URL like sipphone.com or iptel.org. A SIP
server or proxy generally handles IP connections within its domain.

SIP URIs
The SIP server assigns a fixed alphanumeric name to each subscribed account. For example, an Iptel user may be
assigned the user name comrex_user. URIs consist of a SIP user name, followed by a domain, delineated with the @
symbol, like an email address. Our Iptel user’s URI would be [email protected]. Comrex devices do not use the
designation “sip:” before a SIP address.

If a connection is to be made exclusively within a domain, the domain name can be left off. As an example, to make
a call to this codec from another Iptel registered codec, the dialing string can simply be comrex_user (with the
domain being assumed).

60
Registering with a Server
At a minimum, you will need the following information when registering ACCESS with a SIP server:

1 The Internet address of your SIP proxy/server (e.g. proxy01.sipphone.com)

2 The user name on the SIP account (this is usually the dialing address)
3 The password on the SIP account

Figure 31 shows where this information can be applied in the systems setting section. You will also need to enable
the Use SIP Proxy option in that menu.

Figure 31 EBU3326/SIP Settings

Figure 32 Sip status

61
Once this information is correctly entered, a new field appears in the “Registration Status” box located on the
Connections tab (see Figure 32).

The status will reflect the progress of the registration process. When complete, this will display Online. If the box
does not display Online after a short time, it means that registration likely failed. It’s best to go back and carefully
check the registration info. It might also be useful to be sure the registration information is valid by configuring a
VoIP phone or softphone with it and see if that registers.

SIP registration can be very simple with some servers, and others can require more advanced settings. There are
several advanced settings available for use with SIP and they are described in the Advanced Topic sections.

Making SIP Registered Calls

When registered, calls made using an EBU3326/SIP profile behave differently than normal. The address field,
regardless of whether it is a SIP URI or an IP address, is forwarded to the server. No connection attempt is made
until the server responds.

If the server accepts the address, the call will be attempted. If not, an error message will appear in the status line.
Reasons for call rejection by a server are many. Some examples:

• The server does not support direct connection to IP addresses (if the address is in this format)
• The server does not recognize the address
• The server does not forward calls beyond its own domain
• The server does not support the chosen codec
• The called device does not support the chosen codec
• The address is a POTS telephone number, and POTS interworking is not supported
• The address is a POTS telephone number, and no credit is available (most services charge for
this)

Advanced EBU3326/SIP Topics

The basic entries provided will allow support for the vast majority of EBU3326/SIP based applications. However,
there are inevitably situations where the defaults don’t work. We’ve provided some advanced options that
can help. As always, these options are located in the Systems Settings and can be made visible by selecting the
Advanced box.

1 IP Port - Universally, SIP connections are supposed to use UDP port 5060 to negotiate calls
between devices (and between servers and devices). Note this is only the negotiation channel
- actual audio data is passed on the RTP ports. Changing this port number will change which
incoming ports are used to initiate connections and to which ports connection requests are
sent. Obviously, the change must be made on both devices, and this change will essentially
make your codec incompatible with industry-standard VoIP devices.

62
 2 RTP Port - This is one of two port numbers used for audio data transfer (the port number
directly above this is used as well). Because this port number is negotiated at the beginning of
a call (over the IP port), this port may be changed without breaking compatibility. Note that
many SIP standard devices use port 5004 for this function. Due to the negotiation, it is not
important that these numbers match on each end. Changing this port to 5004 can actually
have an adverse effect, since 5004 is the default port for other services on Comrex codecs.
3 Public IP Override - See the next section, SIP Troubleshooting, for more information on this
option.
4 Use STUN Server - See the next section, SIP Troubleshooting, for more information on this
option.
5 SIP Proxy Keepalive - Only applies to Registered mode. This variable determines how often
the codec “phones home” if registered with a SIP server. It’s important that the codec
periodically “ping” the server, so the server can find the codec for incoming calls. It can be
adjusted primarily to compensate for firewall routers that have shorter or longer binding
timings, i.e., the router may have a tendency to “forget” that the codec is ready to accept
incoming calls and block them.
6 SIP Domain - [only applies to Registered mode]. This is the name of the network controlled
by the SIP server. This parameter must be passed by the codec to the server. Under most
circumstances, this is the same as the server/proxy address, and if this field is not populated,
that is the default. If, for some reason, the domain is different than the server/proxy address,
then this field is used.

SIP Troubleshooting

In a nutshell, SIP establishes a communication channel from the calling device to the called device (or server)
on port 5060. All handshaking takes place over this channel, and a separate pair of channels is opened between
the devices: one to handle the audio and the other to handle call control. The original communication channel is
terminated once the handshaking is complete. Note that firewalls must have all three ports open to allow calls to
be established correctly. Also, port forwarding may be required to accept calls if your codec is behind a router.

The main area where SIP complicates matters is in how an audio channel gets established once the handshake
channel is defined. In the common sense world, the call would be initiated to the destination IP address, then the
called codec would extract the source IP address from the incoming data and return a channel to that address. In
fact, that’s how the default mode of Comrex codecs work, and it works well.

But SIP includes a separate “forward address” or “return address” field, and requires that a codec negotiating a call
send to that address only. This is important in the case of having an intermediate server. And this works fine as long
as each codec knows what its public IP address is.

63
Outgoing Call Issues

A unit making an outgoing call must populate the ”return address” field. But any codec sitting behind a router has
a private IP address, and has no idea what the public address is. So, naturally, it will put its private IP address (e.g.
192.168.x.x style) address into that “return address” field. The called codec will dutifully attempt to connect to that
address and undoubtedly fail, since that can’t be reached from the Internet at large.

Incoming Call Issues

Incoming calls to codecs behind routers are complicated by the fact that ports on the router must be forwarded to
the codec. In the case of SIP, this must be three discrete ports (For Comrex codecs these are UDP 5060, 5014 and
5015)<6014 and 6015 with 3.0 firmware>. And since even the “forward address” is negotiated in SIP, the incoming
unit is likely to populate the “forward address” field with its private address as well.

Solutions

Many times the “return address” field issue is fixed by the SIP server (in Registered mode) and no compensation
measures are necessary. Often, in fact, the server insists on acting as a “proxy” and handles all the traffic itself--
outgoing and incoming streams are relayed directly by the server, solving any router issues.

In point-to-point connections, this isn’t possible. All is not lost here, since we can find some hacks to make this
work. The first place to look is your router, since many modern routers are aware of this issue and have taken steps
to relieve the pain. If your router supports a SIP Application Layer Gateway (ALG), then enabling this option can fix
the issue. Essentially, the router will get smart enough to read your SIP handshake, find the outgoing address field,
and replace it with your public IP. This is a pretty slick solution, but there may be environments when you are not
aware whether this option is supported on your router, or have the ability to enable it. So on to solution two:

STUNning Success

Another technique for working around the SIP-Router issue is by using a protocol called STUN. This can be enabled
in Comrex codecs in the Advanced EBU3326/SIP options and essentially allows for the codec to learn what its
public IP address is. It does this by contacting a STUN server out on the Internet (the default one is maintained by
Comrex) and simply asking. If this option is enabled, the codec itself will handle the address switching.

Be aware of the dreaded “battling workarounds” issue. In our simple description, we left out the fact that ports are
being translated by the router as well as IP addresses. If the ALG-enabled router receives an unexpected result in
the SIP address field (as it might if using STUN), it may not translate ports as expected, and it’s likely that the call
will fail. When in doubt, the best technique is to try a SIP call with STUN turned off, and if the return channel fails,
try enabling STUN.

64
Fix of Last Resort

Finally, there’s a brute-force option available on Comrex Codecs when STUN ports are blocked by a firewall, or it
can’t be used for some other reason. Under Advanced System Settings, a field is available called Public IP Override.
Any address put into that field will be pasted into the address SIP field. So if you know what your public IP address
is (you can obtain it from many websites via a browser) you can manually paste it here. Keep in mind, this is often
subject to change over time (and obviously if you use a different network) so it’s important to remember this
change has been made on your codec.

65
 XV. License and Warranty Disclosures for Comrex BRIC-Link

Licenses
MPEG-4 audio coding technology licensed by Fraunhofer IIS
http://www.iis.fraunhofer.de/amm/

BRIC-Link uses proprietary and open-source software programs. Some of the open-source programs are
licensed under the Gnu Public License (GPL). For more information on GPL see http://www.gnu.org.

As per the GPL, source code for this software is available on request from Comrex on CD-ROM or other
electronic format. To obtain this software please contact our support department at +1 978 784 1776. We
retain the right to charge a small handling fee for distribution of this software.

BRIC-Link makes use of open-source and/or free software with the following copyright restrictions:

ncurses
Copyright (c) 1998,1999,2000,2001 Free Software Foundation, Inc.
See further Copyright notice below

dropbear
Copyright (c) 2002-2004 Matt Johnston
Portions copyright (c) 2004 Mihnea Stoenescu
All rights reserved.
See further Copyright notice below

libxml2
Copyright (C) 1998-2003 Daniel Veillard. All Rights Reserved.
See Further Copyright notice below

Import code in keyimport.c is modified from PuTTY’s import.c, licensed as follows:

PuTTY is copyright 1997-2003 Simon Tatham

Portions copyright Robert de Bath, Joris van Rantwijk, Delian Delchev, Andreas Schultz, Jeroen Massar,
Wez Furlong, Nicolas Barry, Justin Bradford, and CORE SDI S.A.

Further copyright notice for ncurses, dropbear PuTTY and libxml2

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and
associated documentation files (the “Software”), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the
following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial

66
portions of the Software.
Libpcap
tcpdump
Copyright © 1988, 1989, 1991, 1994, 1995, 1996, 1997
The Regents of the University of California. All rights reserved.

Redistribution and use in source and binary forms, with or without modification, are permitted provided
that the following conditions are met:

1. Redistributions of source code must retain the above copyright notice, this list of conditions and the
following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and
the following disclaimer in the documentation and/or other materials provided with the distribution.
3. The names of the authors may not be used to endorse or promote products derived from this software
without specific prior written permission.

THIS SOFTWARE IS PROVIDED “AS IS” AND WITHOUT ANY EXPRESS OR

IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

Warranty
All Equipment manufactured by Comrex Corporation is warranted by Comrex against defects in material
and workmanship for one year from the date of original purchase, as verified by the return of the
warranty registration card. During the warranty period, we will repair or, at our option, replace at no
charge a product that proves to be defective, provided you obtain a return authorization from Comrex
and return the product, shipping prepaid to Comrex Corporation, 19 Pine Rd, Devens MA 01434 USA. For
return authorization, contact Comrex at 800-237-1776 or 978-784-1776 or email [email protected].
This warranty does not apply if the product has been damaged by accident or misuse or as a result of
service or modification performed by anyone other than Comrex Corporation.

The next two paragraphs apply to all software contained in this product:

WITH THE EXCEPTION OF THE WARRANTIES SET FORTH ABOVE, THE PRODUCT (MEANS
COLLECTIVELY THE HARDWARE AND SOFTWARE COMPONENTS) IS PROVIDED STRICTLY “AS-IS.”
COMREX CORPORATION AND ITS SUPPLIERS MAKE NO WARRANTY, EITHER EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO ANY IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR
A PARTICULAR PURPOSE OR WARRANTY AGAINST LATENT DEFECTS. COMREX CORPORATION AND
ITS SUPPLIERS DO NOT WARRANT THAT THE PRODUCT IS ERROR-FREE, THAT ALL ERRORS MAY BE
DETECTED OR CORRECTED, OR THAT THE USE OF THE PRODUCT WILL BE UNINTERRUPTED. IN NO
EVENT WILL COMREX CORPORATION AND ITS SUPPLIERS BE LIABLE FOR INDIRECT, INCIDENTAL,
SPECIAL OR CONSEQUENTIAL DAMAGE RESULTING FROM THE USE OF THE PRODUCT INCLUDING
LOSS OF PROFITS, LOSS OF SAVINGS, LOSS OF USE OR INTERRUPTION OF BUSINESS EVEN IF
COMREX CORPORATION OR ANY OF ITS SUPPLIERS HAS BEEN ADVISED OF THE POSSIBILITY OF
SAME. IN NO EVENT SHALL COMREX CORPORATION AND/OR ITS SUPPLIERS’ TOTAL LIABILITY
TO YOU REGARDLESS OF THE FORM OF ACTION EXCEED THE AMOUNT YOU PAID AS PART OF
THE PURCHASE PRICE OF THIS PRODUCT. COMREX CORPORATION AND ITS SUPPLIERS MAKE
NO WARRANTY, EITHER EXPRESSED OR IMPLIED, THAT ANY USE OF THE PRODUCT WILL BE FREE
FROM INFRINGEMENT OF PATENTS, COPYRIGHTS, OR ANY OTHER THIRD PARTY’S INTELLECTUAL
PROPERTY RIGHTS.

67
THE SOFTWARE OWNED BY COMREX CORPORATION OR BY ITS SUPPLIERS RESIDING IN OR
OTHERWISE ASSOCIATED WITH THIS PRODUCT ARE PROTECTED UNDER COPYRIGHT LAW AND
INTERNATIONAL TREATIES. UNAUTHORIZED REVERSE ENGINEERING, REPRODUCTION AND/OR
DISTRIBUTION OF THE PRODUCT OR ANY PORTION THEREOF, IS STRICTLY PROHIBITED AND MAY
RESULT IN CIVIL AND CRIMINAL SANCTIONS, AND WILL BE PROSECUTED TO THE FULL EXTENT OF
THE LAW. COMREX CORPORATION AND ITS SUPPLIERS OWNS AND SHALL RETAIN ALL RIGHT, TITLE
AND INTEREST IN AND TO ANY SOFTWARE SUPPLIED TO YOU IN AND AS PART OF THE PRODUCT
AND ALL INTELLECTUAL PROPERTY RIGHTS RELATED THERETO. THE SALE OF THE PRODUCT SHALL
NOT BE CONSTRUED IN ANY MANNER AS TRANSFERRING ANY RIGHT OF OWNERSHIP IN ANY SUCH
SOFTWARE.

68
 XVI. Conformity and Regulatory Information

Suppliers’ Declaration of Conformity

Place of Issue: Devens, Massachusetts

Date of Issue: April 2, 2009

Equipment: Comrex BRIC-Link

This equipment has been tested and found to comply with the limits for a Class A digital device, pursuant
to part 15 of the FCC Rules. These limits are designed to provide reasonable protection against harmful
interference when the equipment is operated in a commercial environment. This equipment generates,
uses, and can radiate radio frequency energy and, if not installed and used in accordance with the instruc-
tion manual, may cause harmful interference to radio communications. Operation of this equipment in a
residential area is likely to cause harmful interference in which case the user will be required to correct
the interference at his own expense.

Thomas O. Hartnett, Vice President, Comrex Corporation

69
EC Declaration of Conformity for R&TTE Directive

We:

Manufacturer’s Name: Comrex Corporation

Manufacturer’s Address: 19 Pine Road

Devens, MA 01434

hereby declare on our sole responsibility that the product:

Comrex BRIC-Link
Digital Audio Codec

to which this declaration relates is in conformity with the essential requirements and other relevant re-
quirements of the R&TTE Directive (1999/5/EC). This product is compliant with the following standards
and other normative documents:

European EMC Directive (89/336/EEC)

EN 55022:1998/A1:2000, Class A Conducted and Radiated Emissions
EN55024: 1998/A1:2001/A2:2003 (Immunity, ITE Equipment)
Low Voltage Directive (2006/95/EEC)
EN 60950-1: 2001

Contact person: Thomas O. Hartnett, V.P., Engineering

Signed:

Date: 02 April 2009

70
 XVII. Appendix A - IP Compatibility

The BRIC-Link is capable of encoding and decoding a choice of three different types of non-BRIC-Link streams:
Standard RTP, Luci Live and Zephyr Xstream. The choice is exclusive i.e. you must set the BRIC-Link specifically for
the type of stream you wish to be compatible with, and you will remain incompatible with the other two types until
you change it. This setting has no effect on normal BRIC-Link functions, which continue to operate as before.

1 Luci Live - This PDA/PC-based software allows real-time streaming over IP links. As of version
1.2, Luci Live includes AAC and HE-AAC in addition to the default MP2 algorithm. BRIC-Link
can communicate with Luci Live only in Luci’s AAC modes. Note: The free demo available from
Luci does not incorporate the AAC functions; you must have a licensed and registered copy to
use AAC.

To communicate with a Luci Live device:

a Initial Setup - This will define all Standard RTP connections to be Luci Compatible
b BRIC-Link - On the System Settings tab, open the Standard RTP Settings option and
choose RTP Compatibility Mode. On the pull-down box, choose Luci Live.
c Incoming Connections - Luci Live sends either an AAC or HE-AAC stream to the BRIC-
Link on UDP port 5004. These streams will be automatically decoded. By default, a
return channel of AAC 56kb/s mono is returned to the Luci Live product. The return
channel may be altered to any Luci-compatible mode in the Systems Setting section.
d Outgoing Connections - Build a profile using the Profile Manager on the BRIC-Link and
select a Channel Mode of Standard RTP. Then choose a Luci-compatible encoder for
the outgoing call. The Luci software will control what type of stream, if any, is returned
to the BRIC-Link.

2 Zephyr Xstream - Xstream Firmware version 3.2.0 and higher support an “RTP Push” function
that is compatible with BRIC-Link in some modes. BRIC-Link is not currently compatible with
the Xstream’s HTTP and SIP streaming functions. There are several limitations imposed by the
Xstream when using the RTP Push function:
a On the Xstream, only AAC and MP3 coding are available in this mode, and BRIC-Link is
only compatible with the AAC mode.
b The Xstream uses downsampling in modes below 96Kb/s, which is not supported by
BRIC-Link.
c In order for an Xstream to decode an BRIC-Link stream, the default decoder setting
must be changed from <Auto> to <AAC> in the codec menu of the Xstream.

71
To communicate with a Zephyr Xstream:
a Initial Setup - This will define all Standard RTP connections to be Xstream Compatible.
b BRIC-Link - On the System Settings tab, open the Standard RTP Settings option and
choose RTP Compatibility Mode. On the pull-down box, select Zephyr Xstream.
c Incoming Connections - Zephyr Xstream sends an AAC stream to the BRIC-Link on UDP
port 9150. These streams will be automatically decoded. By default, a return channel
of AAC 96kb/s mono is returned to the Xstream. The return channel may be altered to
any Xstream-compatible mode in the Systems Setting section.
d Outgoing Connections - Build a profile using the Profile Manager on the BRIC-Link and
select a Channel Mode of Standard RTP. Then choose an Xstream-compatible encoder
for the outgoing call. The Xstream will control what type of stream, if any, is returned
to the BRIC-Link.

3 Standard RTP - This mode is set to receive a basic, unformatted AAC stream within a standard
RTP/UDP structure. At present, this mode does not offer compatibility with other industry
devices.

72
 XVIII. Appendix B - Using BRIC-Link on Unidirectional Networks

Under most circumstances, BRIC-Link and ACCESS require an IP path in both directions for successful connections,
even when audio is being sent only one-way. For networks that provide data only in one direction, it is possible to
use Standard RTP mode to establish and maintain these links. This section describes how to set that up.

The following setting applies to both codecs in the link (encoder and decoder):

The codec has several compatibility modes under the Standard RTP channel mode. The units default to a mode that
is compatible with the Luci Live PC-based encoder. This must be changed on both codecs.

1 On the BRIC-Link or ACCESS Rack, enter the Web-based Interface and choose the System
Settings tab. On the ACCESS Portable choose Configure > System Settings
2 Find the Advanced tick-box and check it
3 Find Standard RTP Settings and choose to edit the RTP Compatibility mode
4 Change this setting to Standard and click Apply (or Save on ACCESS Portable)

Decode Side Settings Only

Also under Advanced Standard RTP Settings, find the Return Channel Enable entry. Disable the return channel and
click Apply (or Save on ACCESS Portable). This will make sure that no channel will be set up in the direction to the
encoder.

Encode Side Settings Only

Obviously, connections of this type must be established from the encoding side of the link. So you’ll need to build
a new Profile that uses the Standard RTP channel mode under the Profile Editor. Choose your outgoing encoder
along with any other special attributes in the profile editor. Name the Profile something descriptive like “Simplex”.

Next, create your outgoing remote entry in the address book. Apply the new profile to that entry. Any connection
made with that entry will connect in a unidirectional fashion.

Full-Time or Triggered Connections

A remote entry using a unidirectional profile can still utilize the tools required for automatic connection.

To set up a connection to be “always active” (i.e. reconnect in the case of power outage or network failure), choose
that connection on the System Settings tab as the Always Connect To location.

To trigger the connection when an external contact is closed, choose the connection under one of the Contact
Closure settings on the System Settings tab.

73
 XIX. appendix c-Information for IT Managers
The purpose of this appendix is to describe all open ports and services available on the Comrex BRIC-Link.

The Comrex BRIC-Link is a device designed to move real-time, wideband audio over IP networks. The main network
interface is 1000BaseT-Ethernet. The device contains an optimized version of Linux kernel. The IP parameters are
set by a computer on the local LAN using a proprietary broadcast UDP protocol.

Comrex provides a Windows or MAC application (Device Manager) on the included cd or available on our website
at www.comrex.com, to perform this function on the local computer. Once the unit is powered on your BRIC-Link,
you have five minutes before this function is disabled.

IP parameters can also be changed online using the password protected Toolbox interface at <ip-address>/cfg.
Updates to the system are provided by a custom online updater utility. This update process is password protected
and requires access to TCP 80 and TCP 8081. In addition to the password protection, the update data itself must
have a valid cryptographic signature from Comrex, or else it is rejected.

Incoming Services

Port Service Default

TCP 22 SSH* On (Off on products shipped after 1 July 2017)
TCP 80 HTTP control On
TCP 443 TLS protected HTTP control On
TCP 8081 Firmware upload Open only during upgrade process
UDP 9000 BRIC Normal Media On
UDP 9001 CrossLock Media On
UDP 5060, 6014, 6015 SIP Off
UDP 5004, 5005 Standard RTP On (Off on products shipped after 1 July 2017)
TCP 9000 BRIC Normal/TCP Off
TCP 8000 HTTP Media Off

*Only SSH clients with an authorized DSA key can access SSH services on the device. Other forms of authentication
are disabled. This key is kept confidentially by Comrex for factory diagnostics only. SSH services may be disabled
completely via the user interface.

Outgoing services

Service Destination
NTP 0.comrex.pool.ntp.org:123 (UDP)
Switchboard switchboard.comrex.com:8090, switchboard.comrex.com:8081 (secondary) (TCP)
STUN stun.comrex.com:3478 (UDP)
DNS Lookup DNS Server:53 (TCP and UDP)
74