isdn audio codecs at ISDNAUDIO.COM
 

ISDN Audio Codec sales for broadcasters and pro-audio users

 

 

 

 

 

B
R
O
A
D
C
A
S
T

 

I
S
D
N

 

G
U
I
D
E

 

a

n

d

 

F
A
Q

 

 

f
r
o
m

 

I
S
D
N
A
U
D
I
O
.
C
O
M

 

 

Welcome to the ISDN Audio Codec FAQ at ISDNAUDIO.COM

If you don't find the answer to your question here, try submitting your question in our discussion area Chances are that one of the moderators or another user will answer it for you.  

We'll continue building this FAQ with your new queries as they arise.

 

Index

Beginners section
What is ISDN?
What's a codec and how do they work?
What kind of quality can I expect?
How do I choose the right manufacturer/model for me?
What are the various 'protocols' used?

Main section
ISDN Worldwide Protocols & interoperability
Audio Coding Protocols & interoperability
What's Musicam? a protocol or a manufacturer
What about MPEG Layer III
I'm trying to connect to someone with a ....... codec and it doesn't work!
Live connections and coding delays
Sending audio to multiple locations
Installing a single codec for use in multiple studios

Beginners section

What is ISDN?

ISDN stands for Integrated Services Digital Network and is, put simply, is a digital equivalent to a telephone line.  ISDN services are available from most telecomms operators in most parts of the world.  What is descibed as 'an ISDN line' is actually made up of two different types of channels.  D-channels are used for signalling and call setup.  B-channels are the data channels and are used for the transfer of your data (and voice calls -  ISDN lines can still carry those too!)  Each B-channel can be used independently as a separate data channel (or telephone line), so you can have two separate connections (or calls) in progress at the same time.

Typically ISDN is available in two seperate forms.  Basic rate ISDN (also called BRI or Basic Rate Interface or ISDN-2 in Europe) consists of two B-channels and one D-channels.  ISDN audio codecs are generally connected to BRIs.
The second type of ISDN service is called Primary Rate Interface (PRI or ISDN-30 in Europe).  This is used by bigger installations as it provides one D-channel and up to 30 B-channels.  Devices designed for the BRI will not connect directly to PRI installations -  but will connect through the ISDN PABX's normally used with primary rate ISDN.

ISDN offers several key advantages over standard analogue phone-lines.  It has a higher bandwidth, so data can be sent faster.  It normally provides better call quality as the line is digital right throught o the point it reaches you.  ISDN lines can also use Multiple Subscriber Numbering (MSN's) to allow you to have several devices attached to your lines, each with their own unique telephone number.  So, for example, you could have five different telephones attached to a single ISDN BRI.  You still only have 2 B-channels, so only two can be in use at once, but all five could have their own phone number.

One current disadvantage to ISDN is that existing telephones, fax machines, modems etc. cannot be connected directly to it.  These Plain Old Telephony devices (POTS) need a Terminal adapter between them and the ISDN line.  ISDN audio codecs and other devices, which are designed for ISDN do connect directly (they have their own TA built in).

A single B-channel (and remember you have two, even with a BRI) can carry 64kbits of data per second -  that's twice as much as an analogue phone line.   ISDN lines are used for many different applications including: 

  • Broadcast and Professional Audio (ISDN audio codecs)

  • Newspapers -  for rapid worldwide distribution of photographs and stories

  • Internet Access -  up to four times the speed compared with connecting through an analogue phone line

  • Video-conferencing and PC based desktop versions

  • LANs for homeworking and interconnection or backup of computer networks

  • Group 4 fax -  which allows you to send a sheet of A4 in laser quality in seconds.

 

What is a codec and how do they work?

Whilst we've said above that you can use ISDN lines for ordinary telephone calls, and that the quality is better, the audio will still sound like a telephone call.   Typically ISDN telephony provides a bandwidth of 3.1kHz.

You can calculate the bandwidth you would need to carry live stereo CD quality audio as follows:  CD quality is two channels of 16-bit audio send at a sample rate of 44.1kHz.  This comes to 1.4 mbits per second.  Much more than the 128 kbits you have with the two B-channels available on BRI ISDN.  The codec's function is to compress and un-compress (or COde and DECode) the 1.4 mbits of data into the 128 kbits of space available over the ISDN line.

 

What kind of quality can I expect?

The manufacturers make many different claims here!  Basically, it depends on the way the audio is compressed -  but remember that it is just that -   compressed audio;  there's nothing cleverer than that going on. 

With a single Basic rate (ISDN-2) ISDN installation, and using both of the B-channels at the same time, you can expect to get FM broadcast quality stereo audio.  Using a single B-channel (either by choice, or because some cheaper codecs are only capable of using one channel for each connection), you will still get near broadcast quality speech.  This latter mode is often used for down-the-line interviews or commentary from sports matches etc.

Some codecs (like the Musicam range) allow you to use up to six B-channels simultaneously.  Of course this assumes that you have 3 sets of ISDN lines available.  Using all six together allows a data rate of 384 kbits to be achieved.   This means that the audio is less compressed, and so better quality.  Bear in mind that using any more than two B-channels at once does add significantly to the transmission delay of your audio.

 

How do I choose the correct manufacturer or model for me?

Take a look at our product range, and also ask your broadcast colleagues what they would recommend.  If you need a system that will connect with other people's, read the section on interoperability below.

 

What are the various protocols used?

There are two types of protocol for you to consider.  The first is the telecomms protocol.  The second is the audio coding protocol.  These are explained below.

 

Main Section

ISDN Worldwide Protocols & interoperability

ISDN is not quite the same all over the world.  Telecomms providers in different countries (and in the US, in different states) may use slightly different ISDN protocols.  When you buy a codec, you buy it with the correct type of Terminal Adapter installed.  Some TAs can be switched to work with different protocols, others are fixed to one or a selection of them. 

This is only a case of you having the right equipment to connect to your line (rather like having a 110v or 220v toaster -  you have to have the right power supply) it does not affect your ability to connect to other people who's equipment may be connected to another protocol in another country. 

Having said that:  there is one exception to this rule.  In some parts of the states, there is still an older version of ISDN in place, known as Switch56.   This offers B-channels that can only carry 56kbits each, rather than the usual 64kbits.  Musicam codecs will allow you to connect, with a little hope and prayer, to codecs on Switch56 lines.  Others normally don't.  One note on Switch56:  it's fair to say that a line that carries 14% less data will offer significantly poorer audio quality than 64k standard ISDN, hence few pro-audio users do actually continue to use this old system in any case.

 

Audio protocols and interoperability

There are four main methods or protocols which different types of codec use to COde and DECode the audio.  They are:  MPEG Layer II (MPEG-2), MPEG Layer III (MPEG-3), APTx and G.722.  For you to make a successful audio connection, both codecs must be using the same method.  Most codecs are capable of using one or two.   Only Glensound make codecs which are capable of three of the above.

G.722 by a cat's whisker, is the most universally compatible of all the audio coding protocols. MPEG-2 is the most widely used for broadcast audio, offering compatibility and quality.

G.722 uses just one B-channel to send decent quality speech over ISDN.  Baiscally, if two codecs are capable of using G.722 they will always work together.

MPEG Layer II codec interoperability can be more complex. In theory they should all be inter-compatible, but... For example, Musicam codecs, although they are fully MPEG-2 compatible, need to be set in a special mode called 'decoder independant' before they will connect to other MPEG-2 codecs.  The ISYS Pro system is the only codec currently available with an 'auto-detect' algorithm built in.  This queries any other codec with which you try and make a connection and automatically sets itself to 'speak the right language'.  ISYS Pro is quite new to the marketplace, and this feature has given it a real edge.

MPEG Layer II and MPEG Layer III are both open coding standards (i.e. manufacturers don't need a license to use them) designed by the Motion Picture Experts Group -  so called because they also define a standard for video compression.   MPEG Layer II is the best method for broadcast quality audio, and works over one or two B-channels.  FM broadcast quality over two channels, speech quality over just one.  MPEG Layer III is a newer standard developed by this group, which is in theory much better than MPEG-2.  However, most professional audio users still prefer MPEG-2 -  see What about MPEG Layer III -  later in this document.

APTx is something of a Betamax amongst coding standards.   Although it offers several key advantages over the other standards mentioned above, it's not one to go for if easy connection to other people's codecs is important to you.

 

What is Musicam?

It's Musicam USA's own version of MPEG-2 compression.

MPEG-2 is a standard defined by committee.  However, it's possible to change the performance of an encoder by changing the parameters it uses is analysing the signal.  A signal produced in this way is still decodable by other manufacturers' codecs.

As you'd expect, all manufacturers claim that their encoder works better than anyone else's.  Musicam have gone one stage further by giving their version of MPEG-2 its own name.

Philips are one of the biggest companies on the MPEG-2 committee;   their own commercial MPEG-2 coding technology is used in ISYS Pro.

 

What about MPEG Layer III

This is the newest coding method in the marketplace and is supported by a couple of manufacturers including Musicam USA.

It is a far better way of coding the signal.  However, many broadcast and pro-audio users prefer Layer II for the following reasons:

  • MPEG-2 offers a much shorter transmission (coding) delay than MPEG-3.   This is because one of the things that makes MPEG-3 better is that it looks further forward at what's about to happen in the audio than MPEG-2.  As a consequence, it's impossible to design an MPEG-3 codec with as short a delay as MPEG-2 systems.

  • Whilst MPEG-3 is a better technique for some types of audio, when it gets it wrong, it gets it more wrong!  Let us try to explain.  All codecs compress the signal.  Experienced ears can hear the imperfections in any compressed audio.  The most widespread use of audio codecs is in radio.  The imperfections in MPEG-3 audio, unfortunately, are worst in speech or audio including speech.  Many audio professionals will say that these imperfections are, in general, less audible in MPEG-2, whether for music or speech.

  • Most of the installed-base of codecs around the world are capable of Layer II.  A far smaller number of Layer III codecs have currently been installed.

  • MPEG-2 is the new standard which has been adopted for DAB (digital radio broadcasts), DVD and Digital television sound.  As a result, most of the new R&D money is going into further developing MPEG-2 not MPEG-3.  some manufacturers including Philips already boast that their MPEG-2 encoders offer better quality than the newer MPEG-3 devices.

  • Audio that has been passed over an MPEG-3 link sounds absolutely appalling (near un-broadcastable) if then transmitted or carried in a studio-transmmitter link over MPEG-2.  For this reason alone, pro-audio users, particularly in broadcast avoid MPEG-3 ISDN codecs.

 

I'm trying to connect to someone with a ....... codec and it doesn't work!

This really is one to try posting on the discussion board!   But, in brief:

Make sure that the codecs are compatible -  do they both have an audio coding protocol in common?

Try setting each codec manually to use exactly the same bit rate, sample rate, mono/stereo mode and audio coding protocol.  i.e. over-ride the 'auto' settings.

Try a different protocol.  The two almost universal ones are:   MPEG Layer II at 64 kbits mono and G.722.  If neither of these work, there's something very wrong.

If this is a regular problem for you, then you should take a look at our page on the ISYS Pro system, which features a unique 'auto-detect' algorithm.  As far as we've seen it in operation, if it's going to connect, ISYS Pro tends to get it right first time very nearly all of the time.


Live connections and coding delays

The shortest possible codec delay is obtained using G.722 or APTx.   MPEG-2 delays at FM broadcast quality are around 200ms each way.  ISYS Pro reduces that delay by around half.  However, the really clever thing to do is to split-code.  That is, you send 'cue' via G.722 and receive the audio via the higher-quality MPEG-2.  The very latest software update enables the Musicam codecs to do this.  ISYS Pro also offers this functionality.


Sending audio to multiple locations

The Musicam codecs really come into their own on this one.  A fully equiped CDQ Prima 2xx with three internal TAs installed is capable of sending a speech quality signal to six separate destinations simultaneously! 

Other units definitely worth a look for this type of application include the Glensound range.  They have a range of 2-inch wide 3U high codec modules -  they're so tiny that you can get over 20 of these into 3U of rack-space.   They also have a 1U rack system into which you can install two independent codecs/TAs.

Installing a single codec for use in multiple studios

Let's say you have one ISDN audio codec which is used in a transmission suite for part of the day, and also in a production suite.  If the codec only has to answer incoming calls, they you can simply take a spare feed of the audio send & return.

However, most of the time you also need to be able to use the front panel of the codec to dial & make settings.  So, there are two considerations here.  Firstly you have to be able to route the audio feeds.  Secondly you have to somehow route the control interface of the codec(s).

Assuming that you can take care of the audio yourself!  let's move on to the control interface.  CDQ Primas are remote controllable via an RS-232 interface, so you could install a custom script in your computer network server to control a Prima.  We also sell a Windows application which can remote control a CDQ Prima from a PC, and on request, we can supply (for free) software that extends this functionality over a PC network. 

ISYS Pro comes ready for this application - new and existing users should simply ask us for the remote control package which will allow you to use your ISYS system (or multiple systems) over your PC network. The host and guest PCs have 'dual-control' of one or more systems.  This is a very elegant solution for broadcast control rooms and comms areas.

 

That's the end of the current FAQs.  If you didn't find the answer you were looking for here, try our discussion message board, where one of the moderators or another user may be able to help!  If your enquiry is a pre-sales question, e-mail us at enquiries@isdnaudio.com   We'll continue to update this FAQ with the most frequent queries on this board.

 
Musicam   ISYSPro   GlenSound   EasyCorder   IntelliPRIMA
isdn audio codec ISDN AUDIO CODEC