ISDN 101
ISDN lines come in two varieties: Basic Rate Interface (BRI) and Primary Rate Interface (PRI). BRI lines are the kind we normally see in broadcast stations, as these are what is used with MPEG codecs. BRIs have a capability of one or two 64 kbps channels. The line from the central office is a single copper pair identical to a POTS line. When it arrives at the subscriber, this is called the "U" interface. The U interface converts to an "S/T" interface with a small box called an "NT-1." In the USA, NT-1 functionality is usually included in the terminal equipment. In Europe, the telephone company provides the NT-1. Only one NT-1 may be connected to a U interface. As many as eight terminals may be paralleled onto an S bus.

Professional equipment should usually provide access to the S interface, so that it is possible to parallel multiple terminals. You can use either an external NT-1, or the equipment may have an internal NT-1 with both U and S/T connectors.

PRI lines can have as many as 23 channels (31 in Europe). These travel over T1 (E1 in Europe) lines, which require two wire pairs. Normally, this service is intended for bulk-line connection to PBXs.

The B "Bearer" Channels are the 64 kbps paths which carry the voice audio. The D "Data" channel is the path between the central office and terminal equipment that is used for call set-up and status communication. It operates at 16 kbps.

Broadcast interfaces may use either type of line. A simple interface for the newsroom could use a single BRI. Even sophisticated multi-line systems could use BRIs, with enough of them to achieve the desired number of lines. While PRIs would seem to be a more technically appropriate solution for a multi-line system, BRIs may be more cost-effective, more readily available, and would serve to provide a measure of redundancy.

Data and Voice
ISDN lines may be used for voice signals encoded in standard fashion to allow interworking with analog phones, as proposed here, or may be used to transmit digital data streams. The latter mode is used for such applications as high-speed Internet access. It is also the mode used with MPEG codecs. In that case, the ISDN line may be carrying voice signals, but is doing so in a format which is not compatible with the POTS network.

The distinction is made in the "Setup" message which begins each call. Of course the ISDN call-in interface should always use the "voice" setup configuration.

(Incidentally, the voice mode may well be able to convey digital data–and some Telcos charge more money for data service. That is why some ISDN "modems" have a special data-over-voice option which fools the central office into billing for voice call, even when the payload is data.)

Send/Receive Separation
This is the traditional hybrid function provided by broadcast telephone interfaces. Despite the fact that ISDN lines naturally have two independent send and receive paths, there is still the need to provide additional functions to further reduce "leakage." The reason is that almost all calls will originate with telephone sets connected via two-wire analog lines, and so there will still be a mixing of both speech directions.

Acoustic Coupling Reduction
There is often an acoustic path between the received caller audio and the send audio signal. This results from having a loudspeaker in the studio that produces sound which couples into the microphones. When the talent use headphones for monitoring callers, this is not a problem. But sometimes it is not practical to convince guests to wear headphones, and television stations generally don’t want talk show talent to wear earplugs. In these cases, it is desirable to have a mechanism to reduce the coupling electronically. This can be accomplished with a combination of adaptive cancellation and dynamic gain reduction.

High-grade Digital-to-Analog Conversion
When an analog connection to studio equipment is required, pro-grade converters can be used to provide much better quality than the usual Telco conversion. At minimum, 16-bit parts should be used, but 18-20 bit parts may not be overkill given their current reasonable cost.

Sampling-rate Conversion
When the studio connection is via a digital AES/EBU channel, no analog-digital conversion is required, but it will be necessary to adapt the sampling rate of the telephone network to the studio rate. Telco sampling rate is 8 kHz and studio equipment will usually operate at 32, 44.1, or 48 kHz. A process is required to perform the required up-and-down sampling, while suppressing aliasing and reconstruction audio components.

Automatic Gain Control
This function should be provided on both the send and receive audio paths. On the send side, it is necessary to smooth the wide level variations which arise from usual studio practices. Unlike a telephone handset microphone which is placed fairly consistently from the mouth and can be relied upon to produce reasonably consistent levels toward the telephone network, studio microphones and mixing consoles produce a wide range of levels. Talent are used to having on-air processing take care of level variations and are generally not very careful at riding gain.

On the receive side, AGC is essential to deal with the very different levels that can result from the many types of phone sets and Telco analog network components. Our experience is that audio volume can vary as much as 30 dB from call-to-call on a given studio line. This degree of difference requires a careful approach to audio leveling. An AGC that maintains a constant compression ratio regardless of average gain reduction produces more consistency. Freeze gating is also important, so that gain does not increase during caller speech pauses.

Dynamic Equalization
The Telos Delta POTS hybrid interface has included for some time a feature which balances the frequency spectrum of caller audio. With phone sets having a very wide variety of microphone characteristics, this function helps callers to have a reasonable consistency, and has been proven valuable in the "real world." We have found a three-band dynamic equalization processor to have the right trade-off of enough power and not too much undesired audible shifting of frequency characteristics.

Caller "Ducking"
This is an "aesthetic" requirement of many talk hosts. The function reduces the level of the caller when the host talks, allowing her an automatic control over a caller who wants to "carry on." This is a matter of taste: some talents and programmers prefer no ducking so that hosts and callers can conduct heated exchanges without impediment, while others want to exercise control. Equipment therefore needs to have a control which can adjust the effect to the desired level.

The time constants of this operation should be carefully crafted so as to mask the audibility of the gain change with the talent’s voice.

Caller ID
ISDN naturally conveys caller ID information. This is transmitted instantly in the setup message, and is much faster than the 1200 baud modem method used in analog caller ID.

Conference Linking
With two B channels available on one BRI line, broadcast interfaces will be dual units, making possible high quality conferencing between the two potential callers. Some systems will probably support larger numbers of conferenced callers.

Conclusion

ISDN is now widely available, cost-effective, and offers many advantages for call-in talk systems. It is yet another example of digital technology enhancing broadcast operations.

•